← back to all models ← RevenueBench home
openai

openai/gpt-6.1-sol-pro

Mean 0.955 · 35/40 perfect tests · $7.50 total · modalities in: text, image, file · out: text · each card: the prompt → the correct answer (gold) → this model's actual answer

What do the modalities mean?

Modalities in means what you can send this model: text, images, files, video, or audio. Modalities out means what it can send back. Bench tasks feed text extracts, so modalities do not change the score. They are here so you know what this model can handle in real work.

data-crm 5 tests
stage-hygiene-auditPASS 1.00
Plain-English question: How much of the open DS2+ pipeline is missing a substantive why-buys, and is anything missing DS1 timestamps or out of order?
Prompt — what the model saw
Audit stage hygiene on the open pipeline below. Rules: a deal at DS2 or later must have a substantive why-buys (15+ characters, not a placeholder such as tbd, n/a, see notes). A why_buys value of the form LONG(n) means the real text was present but truncated for display; treat it as present and substantive. Every deal should have a DS1 entry timestamp. A deal whose stage entry timestamps are out of order regressed (an earlier-stage timestamp is later than a later-stage timestamp). Report: percent of DS2+ deals non-compliant on why-buys overall and by owner (use the owners table for names), count of deals with no DS1 timestamp, count that regressed, and three process fixes. State definitions before numbers.
cost $0.2220 · 43s · in 50,972 / out 3,061 tok
✓ noncomp_pct✓ no_ds1_0✓ regressed_0✓ per_owner_breakdown✓ definitions_first✓ three_fixes

✓ Correct Answer (ground truth)

Plain-English answer: 81.5% of the 124 DS2+ deals are missing a substantive why-buys — Alex Franklin 48, Cole Ingram 19, Dana Mercer 18, Bryce Harmon 15, Farid Osman 1 — and every deal has a DS1 timestamp with none regressed.
The answer in one line: n_deals = 156; ds2plus = 124; noncomp_pct = 81.5; bryce_noncomp = 15; no_ds1 = 0; regressed = 0 (+1 more below)
n_deals156
bryce_noncomp15
ds2plus124
no_ds10
noncomp_pct81.5
regressed0
per_owner_noncompAlex Franklin=48 of 61 (78.7%); Cole Ingram=19 of 22 (86.4%); Dana Mercer=18 of 22 (81.8%); Bryce Harmon=15 of 18 (83.3%); Farid Osman=1 of 1 (100%)
graded checksnoncomp_pct · no_ds1_0 · regressed_0 · per_owner_breakdown · definitions_first · three_fixes

Model's actual answer

show the model's full answer
Definitions:
- Why-buys non-compliance: a current DS2–DS5 deal with missing why_buys, fewer than 15 characters, or placeholder text. LONG(n) is compliant. Rates are deal-count weighted; DS1 deals are excluded.
- Missing DS1: blank t_ds1 on any open deal.
- Regressed: any populated earlier-stage timestamp later than a populated later-stage timestamp. Equal dates are allowed; missing dates alone do not prove regression.

Why-buys non-compliance:
- Overall: 101 / 124 × 100 = 81.45%.
- Bryce Harmon: 15 / 18 × 100 = 83.33%.
- Alex Franklin: 48 / 61 × 100 = 78.69%.
- Dana Mercer: 18 / 22 × 100 = 81.82%.
- Cole Ingram: 19 / 22 × 100 = 86.36%.
- Farid Osman: 1 / 1 × 100 = 100.00%.
- Elena Sinclair: 0 eligible deals; rate N/A.

Arithmetic check: (15 + 48 + 18 + 19 + 1) / (18 + 61 + 22 + 22 + 1) = 101 / 124.

Timestamp hygiene:
- No DS1 timestamp: 0 of 156 deals.
- Regressed: 0 of 156 deals based on populated stage timestamps.

Three process fixes:
1. Gate DS2 entry and subsequent advancement on substantive why-buys; reject blanks, short text, and placeholders.
2. Assign owners the 101 why-buys exceptions for remediation and review outstanding exceptions weekly.
3. Automatically capture DS1 and every stage transition, retain transition history, and flag missing timestamps or timestamp inversions.
owner-resolutionPASS 1.00
Plain-English question: Every open deal has a blank owner name — resolve each one to a real person using the owners table.
Prompt — what the model saw
The deal extract below has an owner name field that is blank on nearly every row and an owner id field that is populated. Using the owners table, resolve each open deal to an owner name. Report: number of deals resolved, deals whose owner id has no match in the owners table (list them), the owner ids that map to archived/deactivated owners (list their names), and the total pipeline amount per resolved owner. Do not guess a name for any unmatched id.
cost $0.2767 · 51s · in 107,454 / out 4,288 tok
✓ resolved_count✓ unmatched_none✓ deactivated_flagged✓ bryce_amount✓ alex_amount✓ no_invented

✓ Correct Answer (ground truth)

Plain-English answer: All 156 deals resolve to six named owners; jake_amount maps to Bryce Harmon ($1,054,144) and justin_amount to Alex Franklin ($624,310).
The answer in one line: jake_amount = 1,054,144; justin_amount = 624,310; resolved = 156
jake_amount1,054,144
justin_amount624,310
resolved156
deactivated
per_ownerBryce Harmon=1,054,144; Alex Franklin=624,310; Dana Mercer=341,195; Cole Ingram=288,161.43; Farid Osman=4,134; Elena Sinclair=2,100
graded checksresolved_count · unmatched_none · deactivated_flagged · bryce_amount · alex_amount · no_invented

Model's actual answer

show the model's full answer
Resolved: 156 of 156 open deals.
Arithmetic: 35 + 67 + 24 + 22 + 7 + 1 = 156.

Unmatched owner IDs: none. No deals remain unresolved.

Archived owners in owners.csv:
  1520255671 → Gavin Porter
  77260721 → Hugo Lindqvist

Neither archived owner ID appears on an open deal. No separate deactivation field was provided.

Total pipeline amount per resolved owner, across all supplied stages and forecast categories:

Owner ID    Owner name       Deals    Pipeline amount
119337721   Bryce Harmon        35       1,054,144.00
84342457    Alex Franklin       67         624,310.00
83155923    Dana Mercer         24         341,195.00
83155924    Cole Ingram         22         288,161.43
716654662   Farid Osman          7           4,134.00
701163055   Elena Sinclair       1           2,100.00
Total                         156       2,314,044.43

Amount arithmetic (subtotals of consecutive groups of up to 10 deals per owner, in extract order):
  Bryce Harmon: 611,276.00 + 128,145.00 + 218,843.00 + 95,880.00 = 1,054,144.00
  Alex Franklin: 69,304.00 + 61,072.00 + 44,248.00 + 116,420.00 + 137,216.00 + 77,850.00 + 118,200.00 = 624,310.00
  Dana Mercer: 75,670.00 + 204,375.00 + 61,150.00 = 341,195.00
  Cole Ingram: 89,888.10 + 161,373.33 + 36,900.00 = 288,161.43
  Farid Osman: 1,249 + 1 + 1 + 1 + 2,880 + 1 + 1 = 4,134.00
  Elena Sinclair: 2,100 = 2,100.00

Total: 1,054,144.00 + 624,310.00 + 341,195.00 + 288,161.43 + 4,134.00 + 2,100.00 = 2,314,044.43.

Currency was not provided.
gong-hubspot-join-integrityPASS 1.00
Plain-English question: How many call records point at deal ids that do not exist in the open pipeline?
Prompt — what the model saw
Check the integrity of the call-to-deal mapping below. The gong table lists calls in the last 90 days keyed to HubSpot deal ids; deals_open lists the current open default-pipeline deals. Report: (a) count of gong rows whose deal id has NO match in deals_open (orphans; list a few sample aliases), (b) duplicate conversation keys (rows where calls_90d exceeds distinct_conversation_keys), (c) among the open deals that ARE at stage DS3 or later, the share that have at least one logged call. Give sample aliases for the orphan finding, all taken from the data. Note: gong_owner_id is a Gong user id, not a HubSpot owner id.
cost $0.2222 · 43s · in 88,760 / out 3,085 tok
✓ orphan_count✓ dup_keys_0✓ ds3_coverage✓ sample_ids_present

✓ Correct Answer (ground truth)

Plain-English answer: 30 Gong call rows are orphans keyed to dead deal ids, none are duplicates, and DS3 call coverage is 29.4%.
The answer in one line: orphans = 30; ds3_cov_pct = 29.4; dup = 0
orphans30
ds3_cov_pct29.4
dup0
orphan_aliasesDeal-8FA85D, Deal-8FC3F9, Deal-3B7945, Deal-42B265, Deal-9CCC42, Deal-36EA09
graded checksorphan_count · dup_keys_0 · ds3_coverage · sample_ids_present

Model's actual answer

show the model's full answer
(a) Orphans: 30 Gong rows have no matching deal_id in deals_open.
Arithmetic: 67 Gong rows − 37 matching rows = 30.
Sample aliases: Deal-8FA85D, Deal-8FC3F9, Deal-3B7945, Deal-42B265, Deal-9CCC42.
These are unmatched to the supplied open-deal table, not necessarily invalid deal mappings.

(b) Duplicate conversation keys: 0 rows where calls_90d > distinct_conversation_keys.
Arithmetic: 465 calls − 465 distinct conversation keys = 0 excess calls; both fields are equal in every row. Raw conversation keys were not provided, so cross-deal duplication cannot be checked.

(c) Call coverage among current DS3-or-later open deals: 29.4%.
Arithmetic: 61 DS3 + 24 DS4/DS5 = 85 eligible deals; 25 have calls_90d ≥ 1.
Share: 25 ÷ 85 × 100 = 29.4%.

Matching uses hs_deal_id → deal_id, not gong_owner_id → owner_id.
snowflake-sql-trapped-catalog0.50
Plain-English question: Write one Snowflake query for first-month usage per customer using only the provided, trap-laden catalog.
Prompt — what the model saw
Using ONLY the catalog provided, write one Snowflake SQL query returning, per customer company, for its first calendar month as a customer: unique givers, recognition count, successful redemption count. Read the catalog notes carefully: some tables are stale or unpopulated and must be avoided, and one documented filter (excluding deleted givers) must NOT be applied to historical giving counts. Put a comment above each FROM or JOIN explaining the table choice. Return only fenced SQL.
cost $0.0732 · 19s · in 32,100 / out 705 tok
✓ avoids_stale_deals✓ avoids_unpopulated_live✗ uses_cohort_summary✗ succeeded_filter✓ no_giver_deletion_filter✗ returns_sql

✓ Correct Answer (ground truth)

Plain-English answer: A single correct SQL query that uses the right catalog tables and avoids the trap columns — graded on the exact tables and columns.
The answer in one line: use_table = PRODUCTION.PLG.COMPANY_COHORT_SUMMARY
avoid_stalePRODUCTION.HUBSPOT.DEALS (stale)
avoid_unpopulatedV2_LIVE.OBJECTS_DEALS (unpopulated)
forbidden_filterdeleted-giver exclusion must NOT be applied
outputone fenced SELECT with per-FROM/JOIN table-choice comment
required_filtersucceeded redemptions only
use_tablePRODUCTION.PLG.COMPANY_COHORT_SUMMARY
graded checksavoids_stale_deals · avoids_unpopulated_live · uses_cohort_summary · succeeded_filter · no_giver_deletion_filter · returns_sql

Model's actual answer

show the model's full answer
```sql
-- Cannot write the requested query from this catalog alone.
-- Missing: the customer-company identifier and authoritative customer-start date.
-- Missing: a recognition-event table with company, giver, event ID, and timestamp columns.
-- Missing: company, event ID, and timestamp columns for
-- PRODUCTION.DEPRECATED_RECOGNITION.REDEMPTION_RECORDS_V2.
-- COMPANY_COHORT_SUMMARY covers only self-serve companies; its M1_* fields
-- are not documented as first-calendar-month metrics or unique-giver counts.
-- Once those details are supplied, use the customer's start month as a
-- half-open interval: [DATE_TRUNC('MONTH', start_date), next month).
-- Do not exclude deleted givers from historical giving counts.
-- Count redemptions only where STATE = 'succeeded'; the documented source
-- requires confirmation before use for finance-grade numbers.
```
crm-completeness-enrichmentPASS 1.00
Plain-English question: Audit the CRM for missing and invalid fields, and propose the fixes.
Prompt — what the model saw
Audit the CRM extract for completeness and propose fixes. Deals need owner, stage, amount, close date, why-buys; companies need industry, employee count, HQ country; contacts need email, title, persona. Report completeness percent per field. Identify duplicate company clusters (name variants, shared domain) with a survivor per cluster. List invalid emails and domain mismatches. Fill missing company fields only where the enrichment export has a matching row; where CRM and enrichment disagree list both and recommend a source. Never invent a value. End with the 10 fixes with the most pipeline amount at stake.
cost $0.1709 · 60s · in 55,197 / out 5,413 tok
✓ dup_clusters✓ invalid_emails✓ survivor_language✓ completeness_pct✓ no_invent

✓ Correct Answer (ground truth)

Plain-English answer: 9 companies missing employee count, 6 missing HQ country, 4 invalid emails, 1 domain mismatch, 2 duplicate clusters, and 25 enrichment matches.
The answer in one line: n_contacts = 52; n_companies = 34; enrichment_matches = 25; invalid_email_count = 4; n_dup_clusters = 2; domain_mismatch_count = 1 (+4 more below)
n_contacts52
domain_mismatch_count1
enrichment_matches25
invalid_email_count4
n_companies34
n_dup_clusters2
company_missingindustry=0; employee_count=9; hq_country=6
domain_mismatchesCT-0011
duplicate_clusters['C-0A092931', 'C-0A092932'], ['C-0A092933', 'C-0A092934']
invalid_emailsCT-0010, CT-0080, CT-0081, CT-0192
graded checksdup_clusters · invalid_emails · survivor_language · completeness_pct · no_invent

Model's actual answer

show the model's full answer
Completeness

Completeness = nonblank values ÷ supplied records × 100. Populated values are not necessarily valid.

Record / required field       Arithmetic             Complete
Companies: industry           34 ÷ 34 × 100           100.00%
Companies: employee count     25 ÷ 34 × 100            73.53%
Companies: HQ country         28 ÷ 34 × 100            82.35%
Contacts: email               52 ÷ 52 × 100           100.00%
Contacts: title               39 ÷ 52 × 100            75.00%
Contacts: persona             37 ÷ 52 × 100            71.15%

No deals extract was supplied. Completeness for owner, stage, amount, close date, and why-buys is not calculable—not 0%. Pipeline amounts and deal-to-company associations are also unavailable.

Duplicate company clusters

Company names were not supplied, so name-variant detection is unavailable. Shared domains identify two candidate clusters:

Domain         Company aliases                  Proposed survivor
acme-corp.com  C-0A092931, C-0A092932             C-0A092931
globex.io      C-0A092933, C-0A092934             C-0A092933

Survivors are provisional, using the lowest alias as a deterministic tie-breaker; neither record has demonstrated source authority.

Before merging:
- C-0A092931 / C-0A092932: employee counts conflict, 500 versus 510. Industry is Technology versus tech; HQ country is US versus USA. Normalize equivalent labels, but verify employee count. No enrichment row exists.
- C-0A092933 / C-0A092934: industry differs, SaaS versus Technology. Employee count 200 and country US agree. Preserve both industry labels pending taxonomy review; neither supplied source establishes the preferred classification.
- Confirm shared-domain entities are duplicates, then preserve associated records and history during merge.

Invalid emails and domain mismatches

Contact    Company     Email                   Finding
CT-0010   C-66D1FC    user0@                  Invalid: missing email domain
CT-0080   C-92D97D    user0@                  Invalid: missing email domain
CT-0081   C-92D97D    user1@                  Invalid: missing email domain
CT-0192   C-425E2A    user2@                  Invalid: missing email domain
CT-0011   C-66D1FC    user1@other-domain.com  Email domain differs from 66d1fc.com

Syntactically valid emails: (52 − 4) ÷ 52 × 100 = 92.31%. Deliverability was not tested. Do not reconstruct invalid addresses or overwrite CT-0011 without verification. Every contact’s supplied domain field agrees with its company’s domain.

Missing contact fields

Title:
CT-0000, CT-0022, CT-0072, CT-0080, CT-0081, CT-0092, CT-0120, CT-0121, CT-0122, CT-0132, CT-0141, CT-0162, CT-0170.

Persona:
CT-0000, CT-0022, CT-0041, CT-0060, CT-0070, CT-0081, CT-0082, CT-0092, CT-0110, CT-0132, CT-0162, CT-0171, CT-0172, CT-0180, CT-0181.

No supplied enrichment supports filling contact titles or personas. Request verified values; do not infer persona from title.

Supported company fills

Exact domain matches support these proposed employee_count fills:

Company     CRM employee_count   Enrichment / proposed value
C-EC3025    blank                400
C-96039F    blank                400
C-44EA29    blank                400
C-D04904    blank                400
C-B23205    blank                400
C-60C75F    blank                400
C-7BBDFA    blank                400
C-50D386    blank                400

After these fills: (25 + 8) ÷ 34 × 100 = 97.06% employee-count completeness, before deduplication.

Unresolved:
- C-93C8BF: employee count missing; no enrichment row.
- C-2D1F1B, C-D73B89, C-44EA29, C-D04904, C-2C60E5: HQ country missing in both CRM and matching enrichment.
- C-EE9FFB: HQ country missing; no enrichment row.
- No supported HQ-country fills exist. Completeness remains 28 ÷ 34 × 100 = 82.35%.

CRM versus enrichment disagreements

Industry differences:

Company     CRM industry   Enrichment industry
C-66D1FC    tech           Computer Software
C-EC3025    Technology     Computer Software
C-44EA29    tech           Computer Software
C-92D97D    Technology     Computer Software
C-D04904    Technology     Computer Software
C-77A95A    Technology     Computer Software
C-AA8DDA    Technology     Computer Software
C-B25F40    Technology     Computer Software
C-60C75F    tech           Computer Software
C-425E2A    Tech           Computer Software

Recommendation: retain CRM’s broad Technology classification and preserve Computer Software as the enrichment classification until a taxonomy mapping is approved. The supplied data contains no verification timestamps or evidence establishing which source should override the other. Trim the trailing space in C-425E2A’s original “Tech ” value.

HQ-country differences:

Company     CRM country   Enrichment country
C-66D1FC    US            United States
C-950043    US            United States
C-EC3025    USA           United States
C-96039F    USA           United States
C-77A95A    US            United States
C-B23205    US            United States
C-E51FB7    USA           United States
C-D0662E    US            United States
C-425E2A    USA           United States
C-2D7423    USA           United States

Recommendation: use the enrichment spelling, United States, for normalization. These are equivalent country labels, not substantive geographic conflicts.

No populated employee counts disagree with matching enrichment rows.

Additional CRM normalization: map tech and “Tech ” to Technology, health care to Healthcare, and US/USA to United States. These are label cleanups, not new enrichment facts.

10 fixes requiring pipeline prioritization

A genuine “most pipeline amount at stake” ranking cannot be produced: no deal amounts or associations were supplied. The following are 10 proposed fixes, not an amount-ranked list. Pipeline amount at stake is unknown for every item.

1. Supply the deals extract with company associations; audit owner, stage, amount, close date, and why-buys, then calculate affected pipeline.
2. Validate and merge C-0A092931 / C-0A092932 into proposed survivor C-0A092931; resolve employee count 500 versus 510.
3. Validate and merge C-0A092933 / C-0A092934 into proposed survivor C-0A092933; resolve SaaS versus Technology classification.
4. Verify and correct invalid emails for CT-0010, CT-0080, CT-0081, and CT-0192.
5. Verify CT-0011’s user1@other-domain.com against its association with C-66D1FC.
6. Apply the eight supported employee-count fills of 400 listed above.
7. Obtain verified employee count for C-93C8BF and HQ countries for the six unresolved companies listed above.
8. Obtain the 13 missing contact titles.
9. Obtain the 15 missing contact personas.
10. Normalize country and industry labels; retain enrichment provenance and resolve the 10 CRM-versus-enrichment industry differences through an approved taxonomy.
deal-intelligence 5 tests
closed-lost-classification0.80
Plain-English question: Why did we lose the last six months of deals — one primary category per deal, and whose side is the reason on?
Prompt — what the model saw
Classify every deal lost in the last 6 months below into one primary category: pricing, competitor, no decision, timing, product gap, champion left, or other. Also assign a side (Bonusly, buyer, unknown). Then summarize: category counts, the side split, the number of deals where the structured closed_lost_tag clearly disagrees with the free-text reason, and the two patterns most worth acting on. Base everything only on the tag and free-text provided.
cost $0.2333 · 69s · in 53,215 / out 5,273 tok
✓ total_90✗ competitor_top✓ category_counts_present✓ side_split✓ two_patterns

✓ Correct Answer (ground truth)

Plain-English answer: Every closed-lost deal gets exactly one category (pricing, competitor, no decision, timing, product gap, champion left, other) plus a side, per the computed classification.
The answer in one line: total_lost_6mo = 90; top_reason = competitor (26)
total_lost_6mo90
categoriespricing, competitor, no decision, timing, product gap, champion left, other
patternstwo most actionable, tag-vs-text disagreements counted
sideBonusly / buyer / unknown per deal
top_reasoncompetitor (26)
graded checkstotal_90 · competitor_top · category_counts_present · side_split · two_patterns

Model's actual answer

show the model's full answer
Classification rules: use the most specific stated cause; use the tag when the text is inconclusive. “Bonusly” means an explicit offering/fit limitation; “buyer” means budget, priorities, approval, timing, or preference; “unknown” means the cause cannot be attributed. MIA is classified as no decision, not proof that the buyer did nothing. A competitor tag alone does not establish the underlying cause.

Every deal, grouped by primary category and side

Pricing — 5
  buyer: Deal-7ED004, Deal-7B2236, Deal-C33D91, Deal-DAFB82, Deal-8A119B

Competitor — 23
  Bonusly: Deal-F97C37, Deal-DDAB52
  buyer: Deal-422BA6, Deal-1BCA50, Deal-A2C349, Deal-D0C698, Deal-47F1A1, Deal-BF2A98, Deal-369281, Deal-9FCD0D
  unknown: Deal-F7F635, Deal-381C8C, Deal-F1E8A6, Deal-ACE061, Deal-2D2F8D, Deal-0F96AA, Deal-7CC678, Deal-C7156E, Deal-8A0992, Deal-EECC02, Deal-1E7DA9, Deal-286F9C, Deal-64B19A

No decision — 31
  buyer: Deal-13E9CF, Deal-E74A73, Deal-FAC17C, Deal-50E5D8, Deal-413C56, Deal-2A292B, Deal-7FBAC6, Deal-F325A5, Deal-ABD14C
  unknown: Deal-AC944F, Deal-214060, Deal-21B045, Deal-988493, Deal-F308CA, Deal-70F704, Deal-4664E1, Deal-D48E0B, Deal-583ADB, Deal-E0441F, Deal-7CB44D, Deal-AFA56C, Deal-D1AABF, Deal-2BBA21, Deal-386F6E, Deal-3F86A0, Deal-096750, Deal-79E61A, Deal-AE7C4E, Deal-DAB4F1, Deal-B4B50F, Deal-5885B9

Timing — 22
  buyer: Deal-DB0AAC, Deal-91A056, Deal-29326C, Deal-831B7B, Deal-39E25C, Deal-B3ABED, Deal-ED9AE7, Deal-B6AC09, Deal-E6E80A, Deal-B038F0, Deal-175756, Deal-BB78F3, Deal-15DA99, Deal-F4AF5D, Deal-79B7A1, Deal-9F176A, Deal-5E64CE, Deal-69CF3D, Deal-ECBF89, Deal-D1A623, Deal-FEDBCB, Deal-2FEDDB

Product gap — 6
  Bonusly: Deal-242273, Deal-9048EB, Deal-3618CC, Deal-5AD03E, Deal-981AD4, Deal-DC77FE

Champion left — 0
  No text explicitly establishes that a buyer champion left. Deal-F325A5 mentions leadership changes, but not a departed champion.

Other — 3
  buyer: Deal-8E27DA
  unknown: Deal-5DB9B0, Deal-55867E

Ambiguities affecting interpretation

  Deal-8A119B: pricing comes from the tag; “Didn't get approval” does not identify what approval failed.
  Deal-ED9AE7: timing is primary because the text lists timing, budget, and authority without ranking them.
  Deal-242273 and Deal-DC77FE: product gap is primary because the text identifies a specific capability differentiator, despite competitor tags.
  Deal-3618CC and Deal-5AD03E: product gap reflects stated capability needs; the text does not establish whether the capability was unavailable or insufficiently demonstrated.
  Deal-55867E: other because the text gives no cause beyond declining; its timing tag supplies an unconfirmed explanation.

Summary and arithmetic

  Categories:
    pricing 5 + competitor 23 + no decision 31 + timing 22
    + product gap 6 + champion left 0 + other 3 = 90 deals

  Side split:
    Bonusly: 2 competitor + 6 product gap = 8
    buyer: 5 pricing + 8 competitor + 9 no decision + 22 timing + 1 other = 45
    unknown: 13 competitor + 22 no decision + 2 other = 37
    Check: 8 + 45 + 37 = 90

Clear tag/text disagreements — 4

  Deal-8E27DA: “Feature Request” versus choosing swag only and not wanting R&R. This describes a different purchase scope, not a stated missing feature.
  Deal-9048EB: “MIA” versus explicit bad fit and “multiple feature gaps.” MIA also occurred, but the tag omits the stated substantive loss cause.
  Deal-3618CC: “Lost DM” versus “Wanted Surveys.” The stated reason is a capability need, not decision-maker loss.
  Deal-9F176A: “Lost- Timing (1 year or more)” versus an August close and expected resumption closer to year-end. Timing agrees; the explicit duration does not.

  Arithmetic: 3 cause mismatches + 1 duration mismatch = 4.
  Unsupported or vague text is not counted as a clear disagreement. Competitor selection and a product gap can both be true.

Two patterns most worth acting on

  1. Separate deferred demand from stalled engagement.
     No decision 31 + timing 22 = 53 of 90 deals. Within no decision, 22 have unknown-side MIA/no-contact reasons. Use explicit restart triggers for deferred deals and a separate requalification process for unresponsive deals. Deal-175756 has a stated 2027 hold; Deal-F308CA has ignored outreach, not a known buying timeline.

  2. Qualify breadth and capability needs earlier.
     Product gap 6 + competitor losses with explicit broader-offering reasons 2 = 8 deals. Deal-F97C37 and Deal-DDAB52 cite broader offerings; Deal-242273 cites internal currency and onsite spending; Deal-DC77FE cites customization. Capture required capabilities before advancing the evaluation, and distinguish an actual product limitation from a positioning or demonstration gap.
pipeline-tieringPASS 1.00
Plain-English question: Score all 156 open deals into exactly one of LOCK / ACTION / BUILD / REVIVE / WATCH / RISKY.
Prompt — what the model saw
Score the open pipeline below on these signals: forecast category, stage, meetings_30d, contact count, and engagement recency. Assign exactly one tier per deal: LOCK, ACTION, BUILD, REVIVE, WATCH, or RISKY (RISKY = the forecast category disagrees with the engagement evidence). Never tier a deal LOCK if it has zero meetings_30d. Do NOT list every deal. Return ONLY a compact JSON object with exactly these keys: - "tier_counts": object mapping each tier to the number of deals in it (must sum to the total deal count), - "tier_examples": object mapping each tier to up to 3 example deal aliases from the data, - "risky_deals": array of the deal aliases you tiered RISKY, - "lock_violations": number of deals tiered LOCK that have zero meetings_30d (should be 0), - "pipeline_shape": one short paragraph on the overall pipeline shape. Note: inbound_emails_30d is 0 for every row due to a data defect, so treat meetings_30d as the inbound signal.
cost $0.3087 · 94s · in 65,331 / out 7,216 tok
✓ json_returned✓ tier_vocabulary✓ tier_counts_sum✓ risky_deals_listed✓ no_lock_violation_claimed

✓ Correct Answer (ground truth)

Plain-English answer: All 156 open deals tiered exactly once across the six tiers using forecast category, stage, meetings, contacts, and recency.
The answer in one line: total = 156
total156
graded checksjson_returned · tier_vocabulary · tier_counts_sum · risky_deals_listed · no_lock_violation_claimed

Model's actual answer

show the model's full answer
{"tier_counts":{"LOCK":3,"ACTION":13,"BUILD":33,"REVIVE":11,"WATCH":61,"RISKY":35},"tier_examples":{"LOCK":["Deal-D348E1","Deal-C26D20","Deal-403845"],"ACTION":["Deal-25F752","Deal-944310","Deal-3974EB"],"BUILD":["Deal-D73B89","Deal-012CB1","Deal-CA7DC0"],"REVIVE":["Deal-2D1F1B","Deal-66D1FC","Deal-950043"],"WATCH":["Deal-6787C2","Deal-523604","Deal-C9C286"],"RISKY":["Deal-E53952","Deal-5408B0","Deal-9AAE5F"]},"risky_deals":["Deal-E53952","Deal-5408B0","Deal-9AAE5F","Deal-547B2B","Deal-B7EBD1","Deal-A2B47C","Deal-2465CE","Deal-C61CF7","Deal-62D607","Deal-584EE5","Deal-C6D97A","Deal-7B3B0F","Deal-F9A08A","Deal-0660B4","Deal-FD9F4E","Deal-BA571A","Deal-FC22A3","Deal-7BBDFA","Deal-60C2C2","Deal-4A13AD","Deal-8AD4A5","Deal-15D24F","Deal-9D0060","Deal-690476","Deal-635B8E","Deal-ED725A","Deal-55164C","Deal-3BA5EA","Deal-5FDCE4","Deal-F336B6","Deal-5EED42","Deal-BA3DDC","Deal-7599B8","Deal-F9A3C1","Deal-FA32A0"],"lock_violations":0,"pipeline_shape":"3+13+33+11+61+35=156 deals. Forecast confidence exceeds meeting evidence: 35 COMMIT/BEST_CASE deals have zero meetings and are RISKY. LOCK requires DS5 COMMIT, positive meetings, at least 3 contacts and activity within 30 days; other meeting-backed COMMIT/BEST_CASE deals are ACTION. Meeting-backed PIPELINE deals with at least 3 contacts and activity within 30 days are BUILD; zero-meeting PIPELINE deals inactive over 14 days are REVIVE; others are WATCH. No snapshot date was supplied: recency uses 2026-09-04, the latest last_contacted_field, as a proxy, excluding later activity dates. Deal-3EED2C and Deal-57FF13 lack engagement rows and remain WATCH, not assumed zero-meeting."}
call-transcript-extractionPASS 1.00
Plain-English question: Extract the CRM write-back fields from each call transcript as JSON.
Prompt — what the model saw
For each transcript, extract CRM write-back fields as JSON: why-buys (prospect statements only), pain points, stakeholders from the speaker list, budget signal (prospect-stated or null), timeline signal, competitor mentioned (only if the prospect raised it), next step (only if explicitly agreed), objections, confidence. Never fill a field from what the rep said or hoped.
cost $0.1042 · 34s · in 36,128 / out 2,998 tok
✓ json_emitted✓ budget_signal✓ null_discipline✓ stakeholders✓ competitor_prospect_only

✓ Correct Answer (ground truth)

Plain-English answer: Per transcript: prospect-only why-buys, pain points, stakeholders, budget and timeline signals, and competitors — e.g. TX-001: '$40k earmarked', live before open enrollment.
The answer in one line: rep = Alex Franklin
allowed_newTX-001, TX-002, TX-003, TX-004, TX-005, TX-006, Achievers, Bucketlist, Nectar
per_transcript{"TX-001": {"why_buys": ["automating anniversary and birthday awards"], "pain_points": ["HR team of three cannot keep up with it manually", "track everything in a spreadsheet, and people slip through
repAlex Franklin
rep_sourced_nulls{"TX-003": ["budget_signal \u2014 only the rep mentioned pricing/budget"], "TX-004": ["next_step \u2014 rep proposed, prospect said 'no promises'"], "TX-006": ["budget_signal \u2014 rep decoy, prospec
graded checksjson_emitted · budget_signal · null_discipline · stakeholders · competitor_prospect_only

Model's actual answer

show the model's full answer
[
  {
    "transcript_id": "TX-001",
    "deal_alias": "Deal-CFE7F4",
    "why_buys": [
      "Automate anniversary and birthday awards."
    ],
    "pain_points": [
      "The HR team of three cannot keep up with manual awards.",
      "Tracking in a spreadsheet causes people to slip through the cracks."
    ],
    "stakeholders": [
      "Prospect (VP People)",
      "Prospect (HR Admin)"
    ],
    "budget_signal": "About $40k earmarked for engagement tools this fiscal year; not stated as dedicated to this purchase.",
    "timeline_signal": "Ideally live before open enrollment in November; year not stated.",
    "competitor_mentioned": [
      {
        "name": "Achievers",
        "context": "Looked at last year; considered too heavy for a team their size."
      }
    ],
    "next_step": "Security review on September 12; year not stated.",
    "objections": [
      "SSO and audit logs are required for IT sign-off."
    ],
    "confidence": "High: fields are supported by explicit prospect statements; budget allocation to this purchase and date years are not specified."
  },
  {
    "transcript_id": "TX-002",
    "deal_alias": "Deal-70BB30",
    "why_buys": [
      "Tie recognition to retention for the hourly workforce."
    ],
    "pain_points": [
      "Regretted turnover among the hourly workforce is over 30%."
    ],
    "stakeholders": [
      "Prospect (Head of Total Rewards)",
      "Prospect (CFO)"
    ],
    "budget_signal": "Finance approved a $25k pilot budget for this quarter.",
    "timeline_signal": "Decision wanted by end of September; year not stated.",
    "competitor_mentioned": [],
    "next_step": "Send the pilot agreement; the prospect will route it to legal this week.",
    "objections": [
      "Rock-solid Workday integration is a condition."
    ],
    "confidence": "High: fields are supported by explicit prospect statements. No competitor was named; the prospect said this was their first real vendor demo."
  },
  {
    "transcript_id": "TX-003",
    "deal_alias": "Deal-530B50",
    "why_buys": [
      "Make recognition visible across 12 retail locations."
    ],
    "pain_points": [
      "Store managers have zero budget autonomy for on-the-spot recognition."
    ],
    "stakeholders": [
      "Prospect (People Ops Manager)"
    ],
    "budget_signal": null,
    "timeline_signal": "No rush until Q1; year not stated.",
    "competitor_mentioned": [
      {
        "name": "Bucketlist",
        "context": "The CEO used it at her last company and liked it; current evaluation was not stated."
      }
    ],
    "next_step": "Schedule a call with the CEO; the prospect will send two times.",
    "objections": [
      "The CEO must be sold first and decides anything people-related."
    ],
    "confidence": "High: fields are supported by explicit prospect statements. No purchase budget was stated; the rep's pricing is excluded."
  },
  {
    "transcript_id": "TX-004",
    "deal_alias": "Deal-180D02",
    "why_buys": [
      "Consolidate three separate recognition tools into one."
    ],
    "pain_points": [
      "Paying for three tools, none of which connect to their HRIS."
    ],
    "stakeholders": [
      "Prospect (VP People)",
      "Prospect (IT Security Lead)"
    ],
    "budget_signal": "The VP People can approve under $15k annually without board approval; no allocated budget was stated.",
    "timeline_signal": "Procurement takes six to eight weeks minimum; no target purchase or launch date was stated.",
    "competitor_mentioned": [],
    "next_step": null,
    "objections": [
      "The security review for their last vendor took three months, creating hesitation."
    ],
    "confidence": "High: fields are supported by explicit prospect statements. The CFO follow-up was tentative, not agreed; no competitor was named."
  },
  {
    "transcript_id": "TX-005",
    "deal_alias": "Deal-F8767A",
    "why_buys": [
      "Automate service milestones.",
      "Get analytics on recognition equity across departments."
    ],
    "pain_points": [
      "Night-shift teams feel invisible.",
      "Night-shift engagement scores run 20 points lower; the comparison group was not specified."
    ],
    "stakeholders": [
      "Prospect (HR Director)",
      "Prospect (People Ops Coordinator)"
    ],
    "budget_signal": "$12k approved under the engagement line.",
    "timeline_signal": "Running before the January all-hands; year not stated.",
    "competitor_mentioned": [
      {
        "name": "Nectar",
        "context": "Currently mid-pilot; the proposed solution must beat that experience."
      }
    ],
    "next_step": "Present to the executive team on October 2; year not stated.",
    "objections": [
      "Must beat the current Nectar pilot experience.",
      "The executive team is skeptical after a failed rollout two years ago."
    ],
    "confidence": "High: fields are supported by explicit prospect statements; date years and the engagement-score comparison group are not specified."
  },
  {
    "transcript_id": "TX-006",
    "deal_alias": "Deal-EE195F",
    "why_buys": [
      "Cut administrative time spent on service awards."
    ],
    "pain_points": [
      "The HR Manager personally spends five hours a month ordering and shipping plaques."
    ],
    "stakeholders": [
      "Prospect (HR Manager)"
    ],
    "budget_signal": "The prospect stated that budget is not the issue; no amount or approval was stated.",
    "timeline_signal": "Fiscal year resets in February; a Q1 start is realistic. Year and whether Q1 is fiscal or calendar were not stated.",
    "competitor_mentioned": [],
    "next_step": "Send the one-page overview; the prospect will forward it to the COO this week.",
    "objections": [
      "The COO usually prefers building things in-house."
    ],
    "confidence": "High: fields are supported by explicit prospect statements. No external competitor was named; the prospect is comparing against doing it internally."
  }
]
next-to-close0.50
Plain-English question: Which 3 deals are closest to signature right now?
Prompt — what the model saw
Which 3 deals are closest to signature right now? Use the CRM data and the Slack thread; Slack often has the real status before the CRM does. For each: deal, amount, why it is close, what is left. Under 250 words.
cost $0.0748 · 18s · in 32,544 / out 778 tok
✗ exactly_three✓ top3_named✓ slack_first✗ blocked_excluded

✓ Correct Answer (ground truth)

Plain-English answer: Deal-547B2B ($11,200), Deal-B7EBD1 ($9,000), and Deal-403845 ($9,000) — all DS5 COMMIT deals closing September 10-11, with what is left on each.
blocked_excludeddeal_alias=Deal-2465CE; reason=champion left + procurement freeze (Slack #sales-sparring 2026-09-04)
exactly_three1
ranking_ruleSlack signature-imminent deal ranks first; remaining ranked by stage (DS5>DS4), forecast (COMMIT>BEST_CASE>PIPELINE), close_date asc, amount desc
top3rank=1; deal_alias=Deal-547B2B; amount=11200; stage=DS5; forecast_category=COMMIT; close_date=2026-09-11 | rank=2; deal_alias=Deal-B7EBD1; amount=9000; stage=DS5; forecast_category=COMMIT; close_date=2026-09-10 | rank=3; deal_alias=Deal-403845; amount=9000; stage=DS5; forecast_category=COMMIT; close_date=2026-09-11
graded checksexactly_three · top3_named · slack_first · blocked_excluded

Model's actual answer

show the model's full answer
Based on the supplied September 4 Slack thread and CRM snapshot, not a verified current status:

1. Deal-547B2B — $11,200
   Why close: Strongest signature signal. Slack says redlines are clean, the signing page is out, and the VP People said they would sign September 5. CRM lists DS5/COMMIT.
   What is left: Customer signature. No subsequent confirmation of signing is provided.

2. Deal-403845 — $9,000
   Why close: Slack says the order form is with finance; CRM lists DS5/COMMIT.
   What is left: Finance processing/approval and signature; the exact remaining steps are not stated. Ownership needs clarification: CRM names Alex Franklin, while Dana Mercer gives the Slack update.

3. Deal-A2B47C — $6,360
   Why close: CRM lists DS5/COMMIT, and Slack confirms it remains warm with normal legal-review progress. Its September 11 CRM close date puts it ahead of Deal-D348E1, which has the same Slack status but an October 15 close date.
   What is left: Complete legal review and obtain signature. No signing commitment is provided.

Deal-2465CE is excluded despite DS5/COMMIT: Slack reports a departed champion and procurement freeze. Deal-B7EBD1 also has DS5/COMMIT status, but no Slack evidence establishes signature readiness.
product-gap-detectionPASS 1.00
Plain-English question: Which prospect complaints are true product gaps versus rollout or enablement issues?
Prompt — what the model saw
Surface true product gaps raised by prospects in active deals. For each candidate: quote the prospect line, classify as real gap, plan gate, or rollout/enablement issue using the product docs excerpt, and name the deal and amount. Only real gaps go in the summary. Rep language does not count as prospect voice.
cost $0.1070 · 27s · in 44,238 / out 1,224 tok
✓ real_gaps✓ plan_gate_classified✓ enablement_classified✓ gap_summary

✓ Correct Answer (ground truth)

Plain-English answer: Two real gaps (SCIM provisioning, ADP Workforce Now integration) separated from plan gates and enablement issues, each with the prospect quote and deal.
allowed_newTG-001, TG-002, TG-003, TG-004, TG-005, SCIM user provisioning, ADP Workforce Now integration, Custom report builder
enablement_issuestranscript_id=TG-004; deal_alias=Deal-84DBA6; amount=16000; feature=Slack integration recognition adoption; quote=We turned on the Slack integration in July, but half our managers still haven't given a single recognition — nobody ever trained them on it.; why_enablement=Slack integration exists on all plans; managers were never trained
excluded_rep_voicetranscript_id=TG-005; line=Full transparency — we don't have a native mobile app for hourly workers yet, though it's on the roadmap.; reason=rep voice, not prospect-raised
plan_gatestranscript_id=TG-003; deal_alias=Deal-CFE7F4; amount=32000; feature=Custom report builder; tier_required=Enterprise; quote=I tried to build a custom report combining recognition frequency with turnover and couldn't — is that possible on our current plan?
real_gapstranscript_id=TG-001; deal_alias=Deal-EC3025; amount=62000; feature=SCIM user provisioning; quote=We need SCIM user provisioning — our security team will not approve any tool that requires manual account management.; why_real=not listed in product docs at any tier | transcript_id=TG-002; deal_alias=Deal-D0D6B5; amount=23800; feature=ADP Workforce Now integration; quote=Our payroll runs on ADP Workforce Now and I don't see ADP anywhere in your integration list — that's a dealbreaker for us.; why_real=supported HRIS list is Workday/BambooHR/Gusto/Rippling only
graded checksreal_gaps · plan_gate_classified · enablement_classified · gap_summary

Model's actual answer

show the model's full answer
Candidate classifications

1. Deal-EC3025 — Real gap
   Prospect (IT Security Lead): “We need SCIM user provisioning — our security team will not approve any tool that requires manual account management.”
   Docs: SCIM user provisioning is explicitly not currently listed as supported.
   Amount: not provided.

2. Deal-D0D6B5 — Real gap
   Prospect (HRIS Manager): “Our payroll runs on ADP Workforce Now and I don't see ADP anywhere in your integration list — that's a dealbreaker for us.”
   Docs: ADP Workforce Now is explicitly not currently listed as supported.
   Amount: not provided.

3. Deal-CFE7F4 — Plan gate
   Prospect (VP People): “I tried to build a custom report combining recognition frequency with turnover and couldn't — is that possible on our current plan?”
   Docs: Custom report builder is Enterprise-only. The prospect’s current plan and support for this specific report combination are not provided, so the plan-gate classification is provisional.
   Amount: not provided.

4. Deal-84DBA6 — Rollout/enablement issue
   Prospect (People Ops Manager): “We turned on the Slack integration in July, but half our managers still haven't given a single recognition — nobody ever trained them on it.”
   Docs: Slack integration is available on all plans. The prospect identifies missing training, not missing functionality.
   Amount: not provided.

5. Deal-36C33F — Not a prospect-raised gap; excluded
   Prospect (HR Manager): “Good to know. The web version should be fine for our office staff for now.”
   The mobile-app claim comes only from Alex Franklin. The prospect does not request it, and the docs excerpt does not establish mobile-app availability. No supported gap/plan-gate/enablement classification.
   Amount: not provided.

Summary — real gaps only

SCIM user provisioning: Deal-EC3025.
ADP Workforce Now integration: Deal-D0D6B5.

Arithmetic: 1 SCIM gap + 1 ADP gap = 2 prospect-raised real gaps.

Combined deal amount cannot be calculated because neither amount is provided. Active-deal status is also not provided, so neither deal can be verified as active.
rep-performance 5 tests
stale-pipeline-by-rep0.67
Plain-English question: Which open deals have had no email, call, or meeting in the last 7 days?
Prompt — what the model saw
The snapshot date is 2026-09-05. Using the open deals and the engagements-by-deal table, list every open deal with no logged email, call, or meeting in the last 7 days. Compute recency from the engagements table's last_email/last_call/last_meeting fields, NOT from any date field on the deal (the deal's last_contacted_field is unreliable). For each stale deal give: deal alias, owner name, stage, amount, days since last contact. Group by owner, order by amount descending within each owner. Then state, per owner, how many stale deals they carry and the total stale amount. (Do not write outreach text.)
cost $0.3395 · 86s · in 67,586 / out 9,570 tok
✓ stale_recall_ge_80✓ grouped_by_owner✓ bryce_top_amount✓ cole_stale_amt✗ dana_stale_amt✗ stale_count

✓ Correct Answer (ground truth)

Plain-English answer: 65 open deals are stale; the largest is Deal-2D1F1B at $240,000, with Bryce Harmon carrying the most stale dollars at $626,243.
The answer in one line: jake_amt = 626,243; jenn_amt = 261,645; kevin_amt = 252,905.03; top_amount = 240,000; stale_count = 65; top_alias = Deal-2D1F1B
jake_amt626,243
jenn_amt261,645
kevin_amt252,905.03
stale_count65
top_amount240,000
top_aliasDeal-2D1F1B
graded checksstale_recall_ge_80 · grouped_by_owner · bryce_top_amount · cole_stale_amt · dana_stale_amt · stale_count

Model's actual answer

show the model's full answer
As of 2026-09-05, last contact = latest nonblank last_email, last_call, or last_meeting on or before the snapshot. Days since contact = 2026-09-05 − last contact; stale means more than 7 days. Future-dated engagements are excluded. The deal’s last_contacted_field is not used. Currency is not specified.

Each row shows the date subtraction used to calculate recency.

Alex Franklin
Deal alias   | Stage | Amount    | Days since last contact
Deal-CC08D1  | DS1   | 24,000.00 | 09-05 − 08-20 = 16
Deal-E73427  | DS3   | 18,000.00 | 09-05 − 08-26 = 10
Deal-885F45  | DS2   |  9,300.00 | 09-05 − 08-24 = 12
Deal-C2FF3C  | DS1   |  8,316.00 | 09-05 − 08-26 = 10
Deal-0D2F7A  | DS3   |  5,100.00 | 09-05 − 08-24 = 12
Deal-6C60D4  | DS3   |  4,800.00 | 09-05 − 08-24 = 12
Deal-13FEBD  | DS2   |  4,680.00 | 09-05 − 08-24 = 12
Deal-819506  | DS1   |  4,400.00 | 09-05 − 08-28 = 8
Deal-9D0060  | DS3   |  3,840.00 | 09-05 − 08-24 = 12
Deal-690476  | DS2   |  3,600.00 | 09-05 − 08-18 = 18
Deal-C6D97A  | DS4   |  3,240.00 | 09-05 − 08-28 = 8
Deal-EE195F  | DS3   |  3,120.00 | 09-05 − 08-28 = 8
Deal-278DEC  | DS3   |  2,700.00 | 09-05 − 08-28 = 8
Deal-635B8E  | DS3   |  2,600.00 | 09-05 − 08-18 = 18
Deal-6883F3  | DS1   |  2,400.00 | 09-05 − 08-20 = 16
Deal-4A13AD  | DS3   |  2,160.00 | 09-05 − 08-10 = 26
Deal-F67D31  | DS2   |  1,800.00 | 09-05 − 08-28 = 8
Deal-5FDCE4  | DS3   |  1,600.00 | 09-05 − 08-24 = 12
Deal-BA571A  | DS4   |  1,080.00 | 09-05 − 08-18 = 18

Bryce Harmon
Deal alias   | Stage | Amount     | Days since last contact
Deal-2D1F1B  | DS1   | 240,000.00 | 09-05 − 06-16 = 81
Deal-66D1FC  | DS1   |  99,000.00 | 09-05 − 08-20 = 16
Deal-950043  | DS1   |  70,000.00 | 09-05 − 08-17 = 19
Deal-B23205  | DS1   |  45,000.00 | 09-05 − 08-20 = 16
Deal-7BBDFA  | DS3   |  37,440.00 | 09-05 − 07-21 = 46
Deal-332637  | DS2   |  36,000.00 | 09-05 − 08-27 = 9
Deal-1BEEBF  | DS1   |  31,500.00 | 09-05 − 08-17 = 19
Deal-A414F6  | DS1   |  25,200.00 | 09-05 − 08-17 = 19
Deal-C5658B  | DS1   |  23,400.00 | 09-05 − 08-20 = 16
Deal-40522D  | DS3   |  21,000.00 | 09-05 − 08-17 = 19
Deal-C1FA6D  | DS1   |  18,000.00 | 09-05 − 08-20 = 16
Deal-01E193  | DS1   |  12,600.00 | 09-05 − 08-28 = 8
Deal-F0EBBB  | DS3   |  11,400.00 | 09-05 − 08-12 = 24
Deal-927338  | DS1   |  10,920.00 | 09-05 − 08-18 = 18
Deal-E25A09  | DS1   |   6,000.00 | 09-05 − 08-27 = 9
Deal-C9C286  | DS2   |   5,502.00 | 09-05 − 08-27 = 9
Deal-012CB1  | DS1   |       1.00 | 09-05 − 08-13 = 23
Deal-3795AD  | DS2   |       1.00 | 09-05 − 08-28 = 8

Cole Ingram
Deal alias   | Stage | Amount    | Days since last contact
Deal-D04904  | DS2   | 58,529.25 | 09-05 − 08-25 = 11
Deal-B25F40  | DS3   | 40,000.00 | 09-05 − 08-28 = 8
Deal-813836  | DS2   | 32,175.00 | 09-05 − 08-25 = 11
Deal-1BA595  | DS2   | 31,750.00 | 09-05 − 08-25 = 11
Deal-CFE1E8  | DS3   | 18,000.00 | 09-05 − 08-25 = 11
Deal-CD47A6  | DS2   | 12,168.00 | 09-05 − 08-25 = 11
Deal-627646  | DS3   | 11,193.00 | 09-05 − 08-25 = 11
Deal-FF809F  | DS2   |  7,781.20 | 09-05 − 08-25 = 11
Deal-AF932D  | DS2   |  7,225.40 | 09-05 − 08-25 = 11
Deal-A71728  | DS2   |  6,947.50 | 09-05 − 08-25 = 11
Deal-8BC9F5  | DS2   |  5,616.00 | 09-05 − 08-26 = 10
Deal-175395  | DS3   |  4,779.88 | 09-05 − 08-25 = 11
Deal-481E24  | DS3   |  4,140.00 | 09-05 − 08-26 = 10
Deal-C7F9BF  | DS2   |  3,360.00 | 09-05 − 08-25 = 11
Deal-2F3A66  | DS3   |  3,334.80 | 09-05 − 08-25 = 11
Deal-342E96  | DS2   |  2,700.00 | 09-05 − 08-12 = 24
Deal-E568D5  | DS3   |  1,875.00 | 09-05 − 08-25 = 11
Deal-FD9F4E  | DS5   |  1,330.00 | 09-05 − 08-26 = 10

Dana Mercer
Deal alias   | Stage | Amount    | Days since last contact
Deal-44EA29  | DS2   | 60,000.00 | 09-05 − 08-26 = 10
Deal-E51FB7  | DS2   | 43,875.00 | 09-05 − 08-24 = 12
Deal-B42F46  | DS1   | 27,000.00 | 09-05 − 08-17 = 19
Deal-BA3DDC  | DS3   | 23,400.00 | 09-05 − 08-21 = 15
Deal-9DDE86  | DS2   | 20,000.00 | 09-05 − 08-21 = 15
Deal-215CCA  | DS3   | 18,900.00 | 09-05 − 08-19 = 17
Deal-5EED42  | DS3   | 16,250.00 | 09-05 − 08-25 = 11
Deal-57887A  | DS2   | 15,000.00 | 09-05 − 08-28 = 8
Deal-944310  | DS4   | 10,500.00 | 09-05 − 08-03 = 33
Deal-B7EBD1  | DS5   |  9,000.00 | 09-05 − 08-20 = 16
Deal-3974EB  | DS4   |  9,000.00 | 09-05 − 08-28 = 8
Deal-F40F04  | DS2   |  8,100.00 | 09-05 − 08-21 = 15
Deal-7599B8  | DS3   |  7,350.00 | 09-05 − 08-18 = 18
Deal-87DDD1  | DS1   |  5,000.00 | 09-05 − 08-17 = 19
Deal-F336B6  | DS3   |  4,200.00 | 09-05 − 08-21 = 15
Deal-0660B4  | DS4   |  1,920.00 | 09-05 − 08-20 = 16

Farid Osman
Deal alias   | Stage | Amount   | Days since last contact
Deal-8BA24E  | DS1   | 2,880.00 | 09-05 − 08-28 = 8
Deal-8FDCD2  | DS1   |     1.00 | 09-05 − 08-21 = 15

Missing engagement records

These open deals have no row in the engagements table. No contact is logged in the supplied table, but days since last contact cannot be calculated; they are not included in the confirmed-stale totals.

Owner          | Deal alias  | Stage | Amount   | Days since last contact
Alex Franklin  | Deal-3EED2C | DS2   | 7,200.00 | Unknown
Elena Sinclair | Deal-57FF13 | DS1   | 2,100.00 | Unknown

Per-owner confirmed-stale totals

Alex Franklin: 19 deals.
24,000 + 18,000 + 9,300 + 8,316 + 5,100 + 4,800 + 4,680 + 4,400 + 3,840 + 3,600 + 3,240 + 3,120 + 2,700 + 2,600 + 2,400 + 2,160 + 1,800 + 1,600 + 1,080 = 106,736.00.

Bryce Harmon: 18 deals.
240,000 + 99,000 + 70,000 + 45,000 + 37,440 + 36,000 + 31,500 + 25,200 + 23,400 + 21,000 + 18,000 + 12,600 + 11,400 + 10,920 + 6,000 + 5,502 + 1 + 1 = 692,964.00.

Cole Ingram: 18 deals.
58,529.25 + 40,000 + 32,175 + 31,750 + 18,000 + 12,168 + 11,193 + 7,781.20 + 7,225.40 + 6,947.50 + 5,616 + 4,779.88 + 4,140 + 3,360 + 3,334.80 + 2,700 + 1,875 + 1,330 = 252,905.03.

Dana Mercer: 16 deals.
60,000 + 43,875 + 27,000 + 23,400 + 20,000 + 18,900 + 16,250 + 15,000 + 10,500 + 9,000 + 9,000 + 8,100 + 7,350 + 5,000 + 4,200 + 1,920 = 279,495.00.

Farid Osman: 2 deals.
2,880 + 1 = 2,881.00.

Elena Sinclair: 0 confirmed-stale deals; confirmed-stale amount 0.00. Deal-57FF13 remains unknown because its engagement record is missing.
activity-mix-vs-outcomePASS 1.00
Plain-English question: Per rep, what did the last 30 days of emails, calls, and meetings look like next to the deals that advanced?
Prompt — what the model saw
The snapshot date is 2026-09-05. For each sales rep (owner), compute the last-30-day totals of emails, calls, and meetings from the engagements table, and the number of deals that entered DS2 in the last 30 days (t_ds2 within 30 days of the snapshot). Report per rep: the activity mix as percentages (emails/calls/meetings share of that rep's total activities), total activities per DS2 entry, and rank the reps by that efficiency ratio (lowest activities per DS2 entry = most efficient). Name the most efficient rep and the highest-volume rep, and note if they differ. Do not attribute cause.
cost $0.4426 · 74s · in 131,057 / out 7,283 tok
✓ alex_ds2_30d✓ alex_volume_high✓ efficient_named✓ mix_present✓ no_causal_claim

✓ Correct Answer (ground truth)

Plain-English answer: Alex Franklin is the standout signal with 18 deals entering DS2 in 30 days (next closest: Bryce Harmon with 4), alongside per-rep activity totals.
The answer in one line: alex_ds2_30d = 18; bryce_ds2_30d = 4; note = legacy alias justin_ds2=4 actually held Bryce Harmon count; Alex Franklin entered 18 deals into DS2 in the 30d window
alex_ds2_30d18
bryce_ds2_30d4
notelegacy alias justin_ds2=4 actually held Bryce Harmon count; Alex Franklin entered 18 deals into DS2 in the 30d window
graded checksalex_ds2_30d · alex_volume_high · efficient_named · mix_present · no_causal_claim

Model's actual answer

show the model's full answer
Window: 2026-08-06 through 2026-09-05, including t_ds2 dates within 30 days of the snapshot. Activities use the supplied emails_30d, calls_30d, and meetings_30d fields, joined to deal owners.

Rank  Rep             Emails  Calls  Meetings  Total activities       DS2 entries  Activities/DS2
1     Alex Franklin   304     36     41        304 + 36 + 41 = 381    18           381 ÷ 18 = 21.17
2     Bryce Harmon    162     0      43        162 + 0 + 43 = 205     4            205 ÷ 4 = 51.25
3     Cole Ingram     96      14     1         96 + 14 + 1 = 111      2            111 ÷ 2 = 55.50
4     Farid Osman     38      0      34        38 + 0 + 34 = 72       1            72 ÷ 1 = 72.00
5     Dana Mercer     84      18     11        84 + 18 + 11 = 113     1            113 ÷ 1 = 113.00
—     Elena Sinclair  Missing engagement data                        0            Undefined

Activity mix: each activity count ÷ that rep’s total activities × 100.

Rep             Emails                     Calls                     Meetings
Alex Franklin   304 ÷ 381 × 100 = 79.79%   36 ÷ 381 × 100 = 9.45%    41 ÷ 381 × 100 = 10.76%
Bryce Harmon    162 ÷ 205 × 100 = 79.02%   0 ÷ 205 × 100 = 0.00%     43 ÷ 205 × 100 = 20.98%
Cole Ingram     96 ÷ 111 × 100 = 86.49%    14 ÷ 111 × 100 = 12.61%   1 ÷ 111 × 100 = 0.90%
Farid Osman     38 ÷ 72 × 100 = 52.78%     0 ÷ 72 × 100 = 0.00%      34 ÷ 72 × 100 = 47.22%
Dana Mercer     84 ÷ 113 × 100 = 74.34%    18 ÷ 113 × 100 = 15.93%   11 ÷ 113 × 100 = 9.73%
Elena Sinclair  Not computable             Not computable            Not computable

Qualifying DS2 entries, regardless of current stage:
  Alex Franklin (18): Deal-403845, Deal-1FC049, Deal-3EED2C, Deal-7FA0C3, Deal-E531A6, Deal-5296C9, Deal-36C33F, Deal-EE195F, Deal-F436DA, Deal-317E6F, Deal-D1E6C2, Deal-D9A72E, Deal-CA5E44, Deal-4F775F, Deal-898FC5, Deal-46988D, Deal-E73427, Deal-92D97D.
  Bryce Harmon (4): Deal-25F752, Deal-D73B89, Deal-CA7DC0, Deal-1CCE5C.
  Cole Ingram (2): Deal-42326B, Deal-1BA595.
  Farid Osman (1): Deal-499BF6.
  Dana Mercer (1): Deal-57887A.

Alex Franklin is both the most efficient rep by the requested ratio (21.17 activities per DS2 entry) and the highest-volume rep (381 activities). They do not differ.

Data limits: Elena Sinclair’s Deal-57FF13 has no engagement row and no t_ds2. Archived owners Gavin Porter and Hugo Lindqvist have no supplied deals or linked engagements, so their activity totals, mixes, and efficiency cannot be computed. Counts use the supplied 30-day fields as given; individual activity timestamps are unavailable to reconcile future-dated last_meeting values.
ae-qtd-scorecardPASS 1.00
Plain-English question: Build Alex Franklin's QTD scorecard: bookings versus quota, pipeline, and conversion.
Prompt — what the model saw
Build a QTD scorecard for the named AE as of 2026-09-05: bookings vs quota with attainment percent, new vs expansion split, active pipeline by stage with amount, rolling 90-day DS2-to-won rate, win and loss counts with top loss reason, activity volume by type in the last 30 days. Exclude any closed-won deal dated before the quarter. Close with three coaching observations grounded in the numbers.
cost $0.3813 · 126s · in 118,880 / out 10,906 tok
✓ bookings✓ quota✓ attainment✓ new_expansion_split✓ top_loss_reason✓ coaching_obs

✓ Correct Answer (ground truth)

Plain-English answer: 75% attainment ($150,000 on a $200,000 quota), $113,500 new versus $36,500 expansion, pipeline by stage, and the rolling DS2-to-won rate.
The answer in one line: quota = 200000; bookings = 150000; new_bookings = 113500; expansion_bookings = 36500; open_deals_count = 125; attainment_pct = 75; owner = Alex Franklin (+12 more below)
quota200,000
attainment_pct75
bookings150,000
expansion_bookings36,500
expansion_count3
losses_qtd27
new_bookings113,500
new_count5
open_deals_count125
wins_qtd8
activity_30d_totalsemails_30d=807; calls_30d=112; meetings_30d=128; notes_30d=50
allowed_new18 items — e.g. Deal-A1C3E5, Deal-B7D2F4, Deal-C9E1A6, Deal-D4B8C2
ds2_to_wonwindow_start=2026-06-07; won_in_window=8; closed_in_window=35; rate=0.2286
ownerAlex Franklin
pipeline_by_stage{"DS1": {"count": 20, "amount": 284621}, "DS2": {"count": 28, "amount": 353760}, "DS3": {"count": 67, "amount": 552705}, "DS4": {"count": 5, "amount": 23574}, "DS5": {"count": 5, "amount": 45730}}
prior_quarter_excludeddeal_alias=Deal-B3E6F1; close_date=2026-06-20; amount=24000
quarter2026-Q3
snapshot_date2026-09-05
top_loss_reasonreason=Lost- Timing (1 year or more); count=13
graded checksbookings · quota · attainment · new_expansion_split · top_loss_reason · coaching_obs

Model's actual answer

show the model's full answer
Alex Franklin — QTD scorecard as of 2026-09-05

Period: 2026-07-01 through 2026-09-05, assuming calendar quarters. Amounts use the supplied units; currency is not specified.

Bookings vs quota
  QTD bookings: 150,000 across 8 wins
  Q3 quota: 200,000
  Attainment: 150,000 ÷ 200,000 × 100 = 75.0%
  Remaining: 200,000 − 150,000 = 50,000

New vs expansion
  New: Deal-A1C3E5 40,000 + Deal-B7D2F4 35,000 + Deal-C9E1A6 21,000 + Deal-D4B8C2 11,000 + Deal-E6F3A9 6,500 = 113,500
  Expansion: Deal-F2C7D8 20,000 + Deal-A8B4D6 12,000 + Deal-C5D9E2 4,500 = 36,500
  Split: 113,500 ÷ 150,000 = 75.7% new; 36,500 ÷ 150,000 = 24.3% expansion.

Excluded: Deal-B3E6F1, 24,000, closed 2026-06-20 before the quarter.

Active pipeline
All supplied open deals, including those with expected close dates after Q3:

  Stage   Deals       Amount
  DS1        20      284,621
  DS2        28      353,760
  DS3        67      552,705
  DS4         5       23,574
  DS5         5       45,730
  Total     125    1,260,390

Amount arithmetic: 284,621 + 353,760 + 552,705 + 23,574 + 45,730 = 1,260,390.

Rolling 90-day DS2-to-won rate
Window: 2026-06-08 through 2026-09-05, inclusive. Defined as deals entering DS2 in that window that were won by the snapshot date, divided by all deals entering DS2 in that window.

  DS2-entry cohort: 8 won + 27 lost + 76 still open = 111
  Conversion: 8 ÷ 111 × 100 = 7.2%

This includes still-open entrants in the denominator; it is not a closed-deal win rate.

QTD wins and losses
  Wins: 8
  Losses: 27
  Top loss reason: Lost- Timing (1 year or more)
  Frequency: 13 ÷ 27 × 100 = 48.1% of losses
  Amount: 7,200 + 25,200 + 11,520 + 7,200 + 19,600 + 25,000 + 5,760 + 2,340 + 24,000 + 40,001 + 3,360 + 6,300 + 7,200 = 184,681

Activity — last 30 days
Using the supplied _30d counters as of the snapshot, excluding Deal-B3E6F1 as instructed. Individual activity dates are not provided.

  Type       Lost deals + Open deals + QTD wins = Total
  Emails            109 +        599 +       89 = 797
  Calls              25 +         54 +       31 = 110
  Meetings           13 +         90 +       23 = 126
  Notes              25 +          1 +       21 = 47

Total recorded activity: 797 + 110 + 126 + 47 = 1,080. These are deal-linked counters, not a verified complete AE activity log.

Three coaching observations

1. Build a specific quarter-end closing plan. The quota gap is 50,000. Open deals dated through September total 109,363, but only 34,204 is in DS4/DS5: 1,920 + 9,000 + 11,200 + 6,360 + 2,484 + 3,240. That covers 34,204 ÷ 50,000 = 68.4% of the gap, so closing those alone would not reach quota.

2. Qualify timing earlier. Lost- Timing (1 year or more) accounts for 13 of 27 losses and 184,681 in amount. Confirm the buying deadline and decision process before advancing opportunities.

3. Prioritize live engagement on open deals. Of 125 open deals, 62 show neither calls nor meetings in the last 30 days: 62 ÷ 125 = 49.6%. Review those opportunities for a confirmed next conversation; email volume alone does not establish buyer engagement.
multithreading-gap0.75
Plain-English question: Which open deals are single-threaded or under-threaded?
Prompt — what the model saw
Find every open deal that is single-threaded (fewer than 2 active contacts) or under-threaded (fewer than 3, or all contacts in one persona). Active means engaged in the last 60 days and not marked former. Personas: economic buyer, champion, HR admin, IT security, finance. For each flagged deal: amount, stage, active contact count, personas present, personas missing, the most valuable persona to add given the stage, and any on-file unengaged contact at that company who fits; otherwise say none on file.
cost $0.0953 · 30s · in 14,584 / out 1,417 tok
✗ flagged_recall✓ persona_language✓ single_thread_term✓ on_file_fit

✓ Correct Answer (ground truth)

Plain-English answer: Deals with fewer than 2 active contacts (single-threaded) or fewer than 3, or all in one persona (under-threaded), each flagged with the missing persona.
The answer in one line: active_window_start = 2026-07-07
active_window_start2026-07-07
allowed_new59 items — e.g. C-077A0E, C-0D15DF, C-1BDFE0, C-2AE3AA
best_persona_ruleDS2/DS3: economic buyer > IT security > finance > HR admin > champion; DS4/DS5: economic buyer > finance > IT security > HR admin > champion; first persona not present among active contacts
flag_ruleflag if active_contacts < 2 (single-threaded) or active_contacts < 3 or all active contacts share one persona (under-threaded); active = engaged within 60 days and not former
flaggeddeal_alias=Deal-36C33F; amount=15000; stage=DS2; active_count=1; reason=single-threaded; best_persona_to_add=economic buyer | deal_alias=Deal-50D386; amount=36000; stage=DS2; active_count=2; reason=under-threaded; best_persona_to_add=economic buyer | deal_alias=Deal-5408B0; amount=14850; stage=DS4; active_count=2; reason=under-threaded; best_persona_to_add=economic buyer | deal_alias=Deal-5BFE3B; amount=18000; stage=DS2; active_count=2; reason=under-threaded; best_persona_to_add=economic buyer | deal_alias=Deal-885F45; amount=9300; stage=DS2; active_count=2; reason=under-threaded; best_persona_to_add=IT security | deal_alias=Deal-92D97D; amount=60000; stage=DS2; active_count=1; reason=single-threaded; best_persona_to_add=economic buyer … (+5 more)
healthy_not_flaggedDeal-4B0BEB, Deal-84DBA6, Deal-D348E1
persona_universeeconomic buyer, champion, HR admin, IT security, finance
graded checksflagged_recall · persona_language · single_thread_term · on_file_fit

Model's actual answer

show the model's full answer
Three required inputs are missing:

- As-of date: needed to calculate the 60-day engagement window and active contact counts.
- Deal status: needed to establish which deals are open.
- Deal amount and stage: neither is supplied. Stage is also needed to recommend the most valuable persona to add.

Please provide the as-of date and a deal table with deal_id, open/closed status, amount, and stage—or confirm that all listed deals are open.

The calculation will be:
  Active contacts = contacts engaged within 60 days of the as-of date, excluding is_former=true.
  Single-threaded = active contact count < 2.
  Under-threaded = active contact count < 3, or all active contacts share one persona.

The supplied unengaged_contacts.csv can then identify on-file matches for the recommended missing persona.
talk-track-from-transcriptsPASS 1.00
Plain-English question: From this rep's last 10 calls: how do they open, handle objections, and how often do they lock a next step?
Prompt — what the model saw
From this rep's last 10 calls: what they lead with in the first five minutes, how they handle the three most common objections, how often a concrete next step is agreed (rate), and every competitor a prospect raised. Quote at most one line per finding and only lines that appear in the transcripts. End with two coaching notes.
cost $0.1329 · 34s · in 51,101 / out 2,319 tok
✓ objection_count✓ next_step_rate✓ competitors✓ coaching✓ quote_used

✓ Correct Answer (ground truth)

Plain-English answer: 8 of 10 calls open with the same customer-win story; objection handling, next-step agreement rate, and every competitor raised, with quotes.
The answer in one line: n_transcripts = 10; objection_count = 3; rep = Alex Franklin (+3 more below)
n_transcripts10
objection_count3
allowed_new13 items — e.g. TT-001, TT-002, TT-003, TT-004
competitor_quotesAwardco=We're also in late talks with Awardco — their rewards catalog looks bigger than yours.; Kudos=How are you different from Kudos? Our CEO used them at her last company.
competitors_prospect_raisedAwardco, Kudos
next_stepagreed=7; total=10; rate=0.7
objectionsobjection=budget; count=4; quote=Honestly, budget is locked until next fiscal year — I can't add a new line item right now.; transcript_id=TT-001 | objection=timing; count=3; quote=This is good, but can we revisit it next quarter? Open enrollment starts in October and we're underwater.; transcript_id=TT-002 | objection=in-house; count=3; quote=We already do recognition with a spreadsheet and quarterly gift cards — why would we change?; transcript_id=TT-004
opener_themetheme=opens with a customer recognition win story (400-person retailer, 18% turnover reduction); count=8; total=10; quote=Quick one to kick off: a 400-person retailer we work with cut regretted turnover 18% in two quarters after automating milestone awards, and I wanted to walk you through exactly how they did it.; transcript_id=TT-001
repAlex Franklin
rep_raised_competitors_excludedWorkhuman
graded checksobjection_count · next_step_rate · competitors · coaching · quote_used

Model's actual answer

show the model's full answer
First five minutes

Alex Franklin leads with the same retailer outcome story in 8/10 calls = 80%: automated milestone awards and a claimed 18% reduction in regretted turnover over two quarters. These are the rep’s claims, not independently verified results.

Calls: Deal-D348E1, Deal-5408B0, Deal-547B2B, Deal-C61CF7, Deal-60C2C2, Deal-EDC141, Deal-D9A12F, Deal-84DBA6.

Exceptions:
- Deal-403845: security review first, then pricing.
- Deal-1E2498: pricing first, responding to the prospect’s prior request.

Three most common objections and handling

1. Budget or approval constraints: 5/10 calls = 50%.
   Locked budgets occur in Deal-D348E1, Deal-547B2B, Deal-60C2C2, Deal-84DBA6; Deal-403845 adds a budget-committee constraint. Arithmetic: 4 + 1 = 5 distinct calls. Deal-84DBA6 also mentions committee approval but is counted once.
   Alex counters locked budgets with turnover savings and the retailer’s claimed avoided-backfill savings:
   “Totally fair. Most teams fund this out of turnover savings — that retailer saved about $210k in avoided backfills, which is how their finance team signed off.”
   — Deal-D348E1, minute 8
   When committee approval blocks Deal-403845 and Deal-84DBA6, he accepts the delay without securing an agreed follow-up.

2. Timing or workload: 3/10 calls = 30%.
   Deal-5408B0, Deal-C61CF7, Deal-D9A12F.
   Alex acknowledges the workload and proposes a smaller pilot:
   “Makes sense. What if we scope a 90-day pilot with one department so you have internal data before next quarter's planning?”
   — Deal-5408B0, minute 8
   All three agree to a working session; the transcripts do not establish agreement to the pilot itself.

3. Existing spreadsheet-and-gift-card process: 3/10 calls = 30%.
   Deal-403845, Deal-EDC141, Deal-1E2498.
   Alex emphasizes automation and recognition analytics:
   “Spreadsheets work until they scale — the difference is automation: milestones fire without HR lifting a finger, and you get analytics on who is being recognized.”
   — Deal-403845, minute 8
   Only Deal-1E2498 agrees to a concrete next step.

Concrete next-step rate

Counting an explicitly accepted activity with a specified time: 7/10 × 100 = 70%.

Agreed: Deal-D348E1, Deal-5408B0, Deal-547B2B, Deal-C61CF7, Deal-60C2C2, Deal-D9A12F, Deal-1E2498. Each accepts a Thursday, 2pm working session and says they will bring the HRIS manager.

No concrete next step agreed: Deal-403845, Deal-EDC141, Deal-84DBA6.

Every competitor raised by a prospect

- Awardco — Deal-547B2B, minute 4: “We're also in late talks with Awardco — their rewards catalog looks bigger than yours.”
- Kudos — Deal-EDC141, minute 4: “How are you different from Kudos? Our CEO used them at her last company.”

Workhuman is mentioned only by Alex in Deal-C61CF7, not raised by a prospect.

Coaching notes

1. After the retailer story, ask about the prospect’s own turnover costs or recognition workload before applying the savings argument. The provided excerpts show a repeated proof point, not a prospect-specific business case.
2. For committee delays and low urgency, identify the decision owner, approval timing, and a dated follow-up rather than ending with acceptance of the stall. Apply this particularly to Deal-403845, Deal-EDC141, and Deal-84DBA6.
reporting-analytics 5 tests
q3-weighted-forecastPASS 1.00
Plain-English question: Build the Q3 2026 weighted forecast from the open pipeline, and show the math deal by deal.
Prompt — what the model saw
Below are the open deals in the default sales pipeline with a close date between 2026-07-01 and 2026-10-15, pulled from the CRM warehouse on 2026-09-05. Stages: DS1 (discovery) through DS5 (contract). Build the Q3 2026 forecast (quarter = 2026-07-01 to 2026-09-30). Weighting rule: 100 percent of COMMIT plus 35 percent of BEST_CASE. PIPELINE deals count zero. Only deals with a close date inside the quarter count. Report: COMMIT total, BEST_CASE total, weighted forecast, count of deals in each category inside the quarter, the deals you excluded for being outside the quarter (count and total amount), and the top 5 BEST_CASE deals by amount inside the quarter. Show arithmetic. Then, in a final '## Data quality' section of no more than four sentences, name the data quality problems in this extract that would make the forecast unreliable if it were run unattended.
cost $0.2075 · 61s · in 69,894 / out 5,406 tok
✓ commit_total_44729✓ best_case_total_203565✓ weighted_115976.75✓ commit_count_7✓ best_case_count_24✓ excluded_count_32✓ excluded_amount_227575✓ top5_aliases✓ flags_blank_owner✓ flags_missing_why_buys

✓ Correct Answer (ground truth)

Plain-English answer: A stage-weighted Q3 forecast number computed from the open pipeline, with per-stage math that reconciles to the deal data.
The answer in one line: excluded_amount = 227575; best_case_total = 203565; weighted_forecast = 115,976.75; commit_total = 44729; excluded_out_of_quarter = 32; best_case_count = 24; flags = blank owner column + empty why_buys fields must be surfaced (+2 more below)
excluded_amount227,575
best_case_count24
best_case_total203,565
commit_count7
commit_total44,729
excluded_out_of_quarter32
weighted_forecast115,976.75
flagsblank owner column + empty why_buys fields must be surfaced
top5_best_caseDeal-2D7423, Deal-25F752, Deal-E53952, Deal-5EED42, Deal-FA32A0
graded checkscommit_total_44729 · best_case_total_203565 · weighted_115976.75 · commit_count_7 · best_case_count_24 · excluded_count_32 · excluded_amount_227575 · top5_aliases · flags_blank_owner · flags_missing_why_buys

Model's actual answer

show the model's full answer
Q3 2026 forecast: 115,976.75, using close dates from 2026-07-01 through 2026-09-30 inclusive. Currency is not specified in the extract.

In-quarter totals and counts
  COMMIT:       7 deals; 44,729
  BEST_CASE:   24 deals; 203,565
  PIPELINE:    23 deals; 201,637.40; zero forecast contribution
  Total:       54 deals

Arithmetic
  COMMIT = 11,200 + 9,000 + 9,000 + 6,360 + 5,400 + 2,520 + 1,249
         = 44,729

  BEST_CASE = 38,935 + 24,000 + 19,656 + 16,250 + 11,116 + 10,800
            + 10,500 + 9,890 + 9,720 + 9,000 + 7,200 + 3,840
            + 3,780 + 3,600 + 3,240 + 3,150 + 3,120 + 3,060
            + 2,916 + 2,760 + 2,484 + 2,100 + 1,920 + 528
            = 203,565

  Weighted forecast = (100% × 44,729) + (35% × 203,565) + (0% × 201,637.40)
                    = 44,729 + 71,247.75 + 0
                    = 115,976.75

Excluded for closing outside Q3: 32 deals, totaling 227,575. All close in October 2026.

Deal alias       Close date    Amount
Deal-E51FB7      2026-10-01     43,875
Deal-B936FE      2026-10-09     18,000
Deal-D9A12F      2026-10-15     17,000
Deal-D348E1      2026-10-15     13,770
Deal-4062CF      2026-10-15     10,800
Deal-293AF3      2026-10-09      9,000
Deal-034D49      2026-10-15      9,000
Deal-E0ADD8      2026-10-15      7,920
Deal-9F2E43      2026-10-08      7,690
Deal-FCBE5B      2026-10-07      7,500
Deal-712010      2026-10-15      7,200
Deal-6691E0      2026-10-15      5,700
Deal-C61CF7      2026-10-09      5,400
Deal-600CD9      2026-10-02      5,400
Deal-A92065      2026-10-15      5,400
Deal-1D532E      2026-10-15      5,400
Deal-48B656      2026-10-15      5,160
Deal-E531A6      2026-10-15      4,800
Deal-D1E6C2      2026-10-09      4,400
Deal-D9E112      2026-10-09      4,300
Deal-5AD94B      2026-10-15      4,000
Deal-901332      2026-10-15      3,600
Deal-47AE31      2026-10-09      3,600
Deal-15D24F      2026-10-09      3,600
Deal-766C74      2026-10-14      3,300
Deal-ED725A      2026-10-08      2,400
Deal-8AD4A5      2026-10-07      1,800
Deal-D7E999      2026-10-15      1,800
Deal-ED13B0      2026-10-09      1,680
Deal-5FDCE4      2026-10-01      1,600
Deal-7FA0C3      2026-10-01      1,400
Deal-F5A622      2026-10-08      1,080

Excluded amount arithmetic:
  43,875 + 18,000 + 17,000 + 13,770 + 10,800 + 9,000 + 9,000 + 7,920
  + 7,690 + 7,500 + 7,200 + 5,700 + 5,400 + 5,400 + 5,400 + 5,400
  + 5,160 + 4,800 + 4,400 + 4,300 + 4,000 + 3,600 + 3,600 + 3,600
  + 3,300 + 2,400 + 1,800 + 1,800 + 1,680 + 1,600 + 1,400 + 1,080
  = 227,575

Top 5 in-quarter BEST_CASE deals by amount
  1. Deal-2D7423 — 38,935 — 2026-09-30
  2. Deal-25F752 — 24,000 — 2026-09-25
  3. Deal-E53952 — 19,656 — 2026-09-30
  4. Deal-5EED42 — 16,250 — 2026-09-30
  5. Deal-FA32A0 — 11,116 — 2026-09-25

## Data quality
Owner is blank for every deal except Deal-C9C286, and most records have why_buys_chars = 0, leaving ownership and buying-rationale evidence missing.
Deal-A5E80A is DS1 but COMMIT, and Deal-499BF6 is DS2 but COMMIT; these early-stage commitments need validation before receiving 100% weight.
Deal-333EBB, Deal-57FF13, Deal-31AD2C, and Deal-7A2454 remain open with close dates before the 2026-09-05 extraction date, indicating potentially stale close dates.
Currency and amount basis are unspecified, and this open-deals-only extract omits Q3 closed-won actuals, so the calculated open-deal forecast is not a complete Q3 revenue forecast.
aha-moment-2x2PASS 1.00
Plain-English question: Do companies that hit 5+ givers and one redemption in month one actually retain better at 24 months?
Prompt — what the model saw
Test the activation hypothesis on the company cohort below: companies with 5+ unique givers (m1_users >= 5) AND 1+ successful redemption (m1_redemptions >= 1) in their first calendar month retain better at 24 months than those with one signal or neither. Every company in this extract is already 25+ months old, so current_status = 'active' means retained at 24 months. Report the full 2x2 (cohort size and 24-month retention rate for: both signals, givers-only, redemption-only, neither), state how many companies were excluded from the denominator and why (if none, say so), name the single signal with the largest retention lift, and state what this does and does not prove.
cost $0.1982 · 45s · in 73,452 / out 3,868 tok
✓ both_66✓ givers_47✓ redemp_31✓ neither_40✓ excluded_0✓ largest_lift_givers

✓ Correct Answer (ground truth)

Plain-English answer: The four cohort cells are graded against computed 24-month retention; the numbers come from the scorer's cohort table, not opinion.
The answer in one line: both_5plus_givers_and_redemption = 66.0% (31 of 47)
both_5plus_givers_and_redemption66.0% (31 of 47)
exclusionsnone — claim bounded as correlation, not causation
givers_only46.9% (23 of 49)
largest_single_signal_liftunique givers
neither40.0% (38 of 95)
redemption_only31.0% (9 of 29)
graded checksboth_66 · givers_47 · redemp_31 · neither_40 · excluded_0 · largest_lift_givers

Model's actual answer

show the model's full answer
The cohort supports the hypothesis descriptively: companies with both first-month signals had the highest 24-month retention.

First-month signals                         Companies   Retained   24-month retention
Both: m1_users >= 5, m1_redemptions >= 1     47          31         31 / 47 = 65.96%
Givers-only: m1_users >= 5, redemptions = 0  49          23         23 / 49 = 46.94%
Redemption-only: m1_users < 5, redemptions >= 1
                                           29           9          9 / 29 = 31.03%
Neither: m1_users < 5, redemptions = 0       95          38         38 / 95 = 40.00%

Denominator: 47 + 49 + 29 + 95 = 220 companies. Excluded: 0. All companies are mature enough, and all have the required signal and status fields. Only current_status = 'active' counts as retained; 'non_renewing' and 'cancelled' remain in the denominator as not retained. Missing industry or country does not affect eligibility.

Both-signal retention exceeds:
  Givers-only:     65.96% − 46.94% = 19.02 percentage points
  Redemption-only: 65.96% − 31.03% = 34.92 percentage points
  Neither:         65.96% − 40.00% = 25.96 percentage points

Largest single-signal lift: 5+ unique givers.
  Givers-only versus neither: 46.94% − 40.00% = +6.94 percentage points.
  Redemption-only versus neither: 31.03% − 40.00% = −8.97 percentage points.

The same ranking holds when measuring each signal across the whole cohort:
  Givers present versus absent: 54 / 96 − 47 / 124 = 56.25% − 37.90% = +18.35 percentage points.
  Redemption present versus absent: 40 / 76 − 61 / 144 = 52.63% − 42.36% = +10.27 percentage points.

This establishes an observed association in this extract, using the supplied retention definition. It does not prove causation, statistical significance, that these thresholds are optimal, or that driving these behaviors will improve retention in other cohorts.
arr-rollforward-reconciliationPASS 1.00
Plain-English question: Does CRM ARR match billing ARR — and if not, exactly where is the variance?
Prompt — what the model saw
Reconcile ARR between CRM company records and billing subscriptions as of 2026-09-05. Billing ARR is MRR x 12. Report both totals, the variance, and a decomposition into buckets (status mismatch, rounding, missing records, other) that sums exactly to the variance. List each mismatched account with a suggested owner. Business rule: any subscription with a term other than 12 months must have cf_agreement_end_date populated; list violations.
cost $0.1857 · 42s · in 48,637 / out 3,183 tok
✓ billing_total✓ crm_total✓ variance✓ buckets_sum✓ violations_listed

✓ Correct Answer (ground truth)

Plain-English answer: Billing $604,739.28 vs CRM $603,581.76: a $1,157.52 variance decomposed into status-mismatch, rounding, and missing-record buckets.
The answer in one line: n_companies = 40; n_mismatched_accounts = 7; as_of = 2026-09-05 (+4 more below)
n_companies40
n_mismatched_accounts7
as_of2026-09-05
billing_arr_total604739.28
bucket_sum_equals_variance1
buckets{"status_mismatch": {"amount": "-13158.48", "accounts": [{"company_alias": "C-0C8323BF", "billing_arr": "0.00", "hubspot_arr": "4905.24", "diff": "-4905.24", "suggested_owner": "Dana Mercer"}, {"compa
crm_arr_total603581.76
ruleterm_months != 12 requires cf_agreement_end_date
variance1157.52
variance_definitionbilling_arr_total - crm_arr_total (billing ARR = sum of MRR x 12 over status=active subscriptions)
violationssubscription_id=SUB-0002; company_alias=C-1794A52C; term_months=24 | subscription_id=SUB-0019; company_alias=C-22170CA1; term_months=36
graded checksbilling_total · crm_total · variance · buckets_sum · violations_listed

Model's actual answer

show the model's full answer
ARR reconciliation as of 2026-09-05

Basis: active subscriptions count toward billing ARR; cancelled subscriptions count as zero. The files provide no historical status dates, so this assumes the supplied statuses apply on 2026-09-05. Currency is not specified.

Totals and variance
  CRM ARR: 603,581.76
  Active billing MRR: 50,394.94
  Billing ARR: 50,394.94 × 12 = 604,739.28
  Variance (CRM − billing): 603,581.76 − 604,739.28 = −1,157.52

Variance decomposition
  Status mismatch: 4,905.24 + 8,253.24 = +13,158.48
  Rounding candidates: 16.00 + 20.00 = +36.00
  Missing records: 16,497.24 − 28,449.24 = −11,952.00
  Other: 24,396.00 − (2,233.00 × 12) = −2,400.00

  Check: 13,158.48 + 36.00 − 11,952.00 − 2,400.00 = −1,157.52

The rounding bucket is provisional: the data shows small differences but does not establish their cause.

Mismatched accounts
Suggested owners are roles; actual account owners were not provided.

Company alias | CRM ARR | Billing ARR | CRM − billing | Bucket | Suggested owner
C-0C8323BF | 4,905.24 | 0.00 | +4,905.24 | Status mismatch: cancelled | RevOps; confirm cancellation with Billing
C-0DC4FB8C | 8,253.24 | 0.00 | +8,253.24 | Status mismatch: cancelled | RevOps; confirm cancellation with Billing
C-0D66DF9E | 23,200.00 | 1,932.00 × 12 = 23,184.00 | +16.00 | Rounding candidate | RevOps
C-14D70CE0 | 18,200.00 | 1,515.00 × 12 = 18,180.00 | +20.00 | Rounding candidate | RevOps
C-21629AA4 | Missing | 2,370.77 × 12 = 28,449.24 | −28,449.24 | Missing CRM record | RevOps
C-0D5BBE3A | 16,497.24 | Missing | +16,497.24 | Missing billing record | Billing / Finance
C-0F7269D7 | 24,396.00 | 2,233.00 × 12 = 26,796.00 | −2,400.00 | Other: unexplained amount difference | RevOps + Billing / Finance

Missing records contribute zero on the missing side for reconciliation purposes; that does not establish their actual ARR.

Agreement-end-date violations
Subscription | Company alias | Term | cf_agreement_end_date | Suggested owner
SUB-0002 | C-1794A52C | 24 months | Missing | Billing / Deal Desk
SUB-0019 | C-22170CA1 | 36 months | Missing | Billing / Deal Desk

If cancelled MRR is included literally, total billing ARR is 51,491.48 × 12 = 617,897.76, and CRM − billing is −14,316.00. Under that basis, the status bucket is zero; 36.00 − 11,952.00 − 2,400.00 = −14,316.00.
kvm-month-over-monthPASS 1.00
Plain-English question: How did the four core KVMs move in August versus July?
Prompt — what the model saw
Report the four core KVMs for 2026-08 against 2026-07: giving rate, redemptions per user, 1:1 meetings engagement, pulse check engagement. For each: value, prior value, absolute and relative change, direction. Name the KVM with the largest relative move and the segment (plan_tier or size_band) driving it if the data supports that; otherwise say the data does not. Under 200 words plus one table.
cost $0.2067 · 61s · in 68,874 / out 5,423 tok
✓ four_kvms✓ biggest_mover_named✓ direction_correct✓ segment_named✓ word_cap

✓ Correct Answer (ground truth)

Plain-English answer: Giving rate 60.27% (up 0.07%), redemptions per user 1.7302 (up 0.01%), with 1:1 meetings and pulse engagement each up fractionally — direction reported for all four.
The answer in one line: n_company_rows_per_month = 30; aggregation = simple mean of company-level values per month
n_company_rows_per_month30
aggregationsimple mean of company-level values per month
biggest_moverpulse_engagement
driving_segmentsize_band=enterprise; 2026-07=0.55; 2026-08=0.2743; rel_change_pct=-50.13
kvms{"giving_rate": {"2026-07": 0.6023, "2026-08": 0.6027, "abs_change": 0.0004, "rel_change_pct": 0.07, "direction": "up"}, "redemptions_per_user": {"2026-07": 1.73, "2026-08": 1.7302, "abs_change": 0.00
segment_detail{"smb": {"2026-07": 0.6588, "2026-08": 0.6573, "rel_change_pct": -0.23}, "mid_market": {"2026-07": 0.593, "2026-08": 0.5942, "rel_change_pct": 0.2}, "enterprise": {"2026-07": 0.55, "2026-08": 0.2743,
graded checksfour_kvms · biggest_mover_named · direction_correct · segment_named · word_cap

Model's actual answer

show the model's full answer
Values are unweighted means across 30 companies; user counts and metric denominators are missing, so user-weighted KVMs cannot be calculated.

| KVM | 2026-08 | 2026-07 | Absolute change | Relative change | Direction |
|---|---:|---:|---:|---:|---|
| Giving rate | 60.2713% | 60.2297% | +0.0417 pp | +0.0692% | Up |
| Redemptions per user | 1.730163 | 1.729983 | +0.000180 | +0.0104% | Up |
| 1:1 meetings engagement | 44.7177% | 44.6887% | +0.0290 pp | +0.0649% | Up |
| Pulse check engagement | 50.8610% | 60.0587% | −9.1977 pp | −15.3145% | Down |

Arithmetic: mean = sum ÷ 30; absolute change = August − July; relative change = (August − July) ÷ July × 100. July → August sums:
Giving: 18.0689 → 18.0814.
Redemptions: 51.8995 → 51.9049.
1:1: 13.4066 → 13.4153.
Pulse: 18.0176 → 15.2583.
Rate differences are expressed in percentage points (pp).

Pulse check engagement had the largest relative move. size_band=enterprise drives the decline: 54.998% → 27.428%, or −27.570 pp ÷ 54.998% × 100 = −50.1291%. Its contribution is −27.570 × (10/30) = −9.190 pp, accounting for 99.9166% of the net decline. plan_tier cannot distinguish a driver: every company is tier_three.
rewards-ytd-sectionPASS 1.00
Plain-English question: Produce the redemption section of the weekly rewards report through August.
Prompt — what the model saw
Produce the redemption section of the weekly rewards report through the last completed month (2026-08): redemption count, spend, unique redeemers, redemptions per redeemer, provider mix as percent of spend, top 5 countries by redemptions. Say which month is the last completed month. Provider shares must sum to 100.
cost $0.3016 · 66s · in 61,946 / out 7,346 tok
✓ count✓ spend✓ unique_redeemers✓ month_named✓ provider_mix✓ top_country

✓ Correct Answer (ground truth)

Plain-English answer: 378 redemptions, $27,846 spend, 235 unique redeemers at 1.61 each, custom rewards at 39% of spend, US leading all countries.
The answer in one line: redemption_count = 378; unique_redeemers = 235; redemptions_per_redeemer = 1.61; total_spend_usd = 27846.00
redemption_count378
redemptions_per_redeemer1.61
unique_redeemers235
provider_mix_pct_of_spendTangoCard=11.6; Tremendous=30.54; custom=39.05; Snappy=18.81
top5_countries_by_countcountry=US; count=244 | country=CA; count=24 | country=AU; count=21 | country=GB; count=17 | country=NL; count=17
total_spend_usd27846.00
graded checkscount · spend · unique_redeemers · month_named · provider_mix · top_country

Model's actual answer

show the model's full answer
Redemptions — January–August 2026

Last completed month: August 2026 (2026-08).

Redemption count: 90 + 192 + 37 + 59 = 378.
Spend: $3,230 + $8,505 + $10,873 + $5,238 = $27,846.
Unique redeemers: 235 distinct user_key values across all providers.
Redemptions per redeemer: 378 ÷ 235 = 1.61.

Provider mix as percent of spend:
  TangoCard: $3,230 ÷ $27,846 × 100 = 11.60%
  Tremendous: $8,505 ÷ $27,846 × 100 = 30.54%
  custom: $10,873 ÷ $27,846 × 100 = 39.05%
  Snappy: $5,238 ÷ $27,846 × 100 = 18.81%
  Total: 11.60% + 30.54% + 39.05% + 18.81% = 100.00%

Top 5 countries by redemption count:
  1. US: 244
  2. CA: 24
  3. AU: 21
  4. GB: 17 (tied)
  4. NL: 17 (tied)
customer-success 5 tests
churn-save-eligibilityPASS 1.00
Plain-English question: Which at-risk accounts actually qualify for a churn-save offer, and which play fits each?
Prompt — what the model saw
Which at-risk accounts qualify for a churn-save offer under the documented eligibility rules, what amount is at stake per account and in total, and which play fits each (usage revival, executive touch, commercial concession)? Cite the signal that justifies each play. List accounts that look at risk but do not qualify and why.
cost $0.1405 · 43s · in 49,489 / out 3,554 tok
✓ eligible_set✓ total_at_stake✓ plays_cited✓ noneligible_named✓ rules_applied

✓ Correct Answer (ground truth)

Plain-English answer: Only accounts meeting all three rules qualify — e.g. C-0F6C0F34 at $49,707 (executive touch) and C-0B827671 at $25,365 (usage revival) — with amount and play per account.
The answer in one line: n_accounts = 30; snapshot_date = 2026-09-05
n_accounts30
eligibleaccount_alias=C-0F6C0F34; amount_at_stake=49707.00; play=executive touch; justifying_signal=champion_active is false - no executive sponsor engaged | account_alias=C-0B827671; amount_at_stake=25365.00; play=usage revival; justifying_signal=usage_trend_3m=declining over the last 3 months | account_alias=C-0B360C78; amount_at_stake=35748.00; play=commercial concession; justifying_signal=usage stable/growing with seat utilization 75% - risk is commercial, not adoption | account_alias=C-0B0F1BAB; amount_at_stake=5494.00; play=executive touch; justifying_signal=champion_active is false - no executive sponsor engaged | account_alias=C-0CA21961; amount_at_stake=16829.00; play=usage revival; justifying_signal=seat utilization 26% is below 50% | account_alias=C-0E9C27D1; amount_at_stake=41235.00; play=commercial concession; justifying_signal=usage stable/growing with seat utilization 85% - risk is commercial, not adoption … (+2 more)
non_eligible_at_riskaccount_alias=C-0BC71BDD; health_score=55 | account_alias=C-0BA71F12; health_score=52 | account_alias=C-0F6694C3; health_score=43 | account_alias=C-0BE96399; health_score=54 | account_alias=C-0F876796; health_score=47 | account_alias=C-0FCCD2DF; health_score=43 … (+1 more)
ruleshealth_score < 60, churn_save_eligible_amount > 0, renewal within 120 days of snapshot
snapshot_date2026-09-05
total_amount_at_stake224601.00
graded checkseligible_set · total_at_stake · plays_cited · noneligible_named · rules_applied

Model's actual answer

show the model's full answer
8 accounts qualify, with 224,601.00 in churn-save eligible amounts and 454,380.00 in total ARR. Currency is not specified in the data.

Eligibility requires all three rules: health_score < 60 (R1), churn_save_eligible_amount > 0 (R2), and renewal within 120 days of 2026-09-05 (R3), through 2027-01-03. Dates below use that snapshot, not today.

Qualifying accounts and recommended plays

Play-selection rules were not provided. These are signal-based recommendations, not documented eligibility requirements. No pricing objection or concession request is supplied, so a commercial concession cannot be justified from the data alone.

Account       Eligible amount  Total ARR   Renewal (days)   Play and supporting signal
C-0F6C0F34    49,707.00        86,741.00   2026-10-03 (28) Executive touch: champion_active=false despite growing usage; health=51.
C-0B827671    25,365.00        72,088.00   2026-11-14 (70) Usage revival: declining usage; seats used=113/202 × 100=55.9%; health=56.
C-0B360C78    35,748.00        60,427.00   2026-10-28 (53) Executive touch to diagnose risk: health=57 despite growing usage and an active champion. No specific executive-engagement gap is documented.
C-0B0F1BAB     5,494.00        15,391.00   2026-09-23 (18) Executive touch: champion_active=false, health=38, and renewal in 18 days.
C-0CA21961    16,829.00        31,501.00   2026-12-28 (114) Usage revival: flat usage; seats used=84/325 × 100=25.8%; health=58.
C-0E9C27D1    41,235.00        75,093.00   2026-09-24 (19) Executive touch to diagnose risk: health=39 and renewal in 19 days despite an active champion and seats used=134/157 × 100=85.4%. Cause of poor health is missing.
C-0CEF69FD    32,621.00        79,324.00   2026-11-21 (77) Executive touch: champion_active=false despite growing usage; health=53.
C-0D3278C7    17,602.00        33,815.00   2026-11-12 (68) Usage revival: declining usage; seats used=126/380 × 100=33.2%; health=54.

Arithmetic

Churn-save eligible amount:
49,707 + 25,365 + 35,748 + 5,494 + 16,829 + 41,235 + 32,621 + 17,602 = 224,601.00

Total ARR across qualifying accounts:
86,741 + 72,088 + 60,427 + 15,391 + 31,501 + 75,093 + 79,324 + 33,815 = 454,380.00

The eligible amount is not a documented discount budget or a forecast of ARR lost.

At-risk accounts that do not qualify

All accounts below meet R1 but fail R2, R3, or both.

Account       Health  Eligible amount  Days to renewal  Reason excluded
C-0BC71BDD    55           0.00         52               Fails R2: eligible amount is zero.
C-0BA71F12    52       6,824.00        218               Fails R3: renewal exceeds 120 days.
C-0F6694C3    43           0.00        197               Fails R2 and R3: zero eligible amount; renewal exceeds 120 days.
C-0BE96399    54           0.00         54               Fails R2: eligible amount is zero.
C-0F876796    47      19,958.00        154               Fails R3: renewal exceeds 120 days.
C-0FCCD2DF    43           0.00        230               Fails R2 and R3: zero eligible amount; renewal exceeds 120 days.
C-10A56B0F    54           0.00         98               Fails R2: eligible amount is zero.

Source: churnzero_accounts.csv; eligibility_rules.csv, R1–R3.
expansion-kitPASS 1.00
Plain-English question: Build the expansion kit for the customer who just replied to our outreach.
Prompt — what the model saw
A customer replied to outreach (reply included). Build the expansion kit: seat coverage (licensed over headcount), two lines on usage health, headroom in seats and ARR at the current per-seat rate, who replied and whether they can buy, the right buyer from our contacts, and a reply email under 150 words that moves toward the right person without being pushy and cites one real usage fact.
cost $0.1044 · 24s · in 43,145 / out 1,141 tok
✓ coverage_pct✓ headroom_arr✓ right_buyer✓ replier_not_buyer✓ usage_fact✓ email_under_150

✓ Correct Answer (ground truth)

Plain-English answer: 37.5% seat coverage, 250 headroom seats worth $15,000 ARR at $60 a seat, usage up 43.2% in six months — and the replier cannot buy; Dana R. (VP People) is the right buyer.
The answer in one line: headroom_seats = 250; coverage_pct = 37.5; company_alias = C-0DDFC9A7 (+3 more below)
headroom_seats250
coverage_pct37.5
company_aliasC-0DDFC9A7
headroom_arr15000.00
per_seat_annual_rate60.00
replierMaria S.
replier_can_buy0
right_buyername=Dana R.; title=VP People
usage_factsactive_users_2026_08=126; active_users_2026_03=88; growth_pct_6m=43.2; licensed_seat_utilization_pct=84
usage_health_linesMonthly active users grew from 88 (2026-03) to 126 (2026-08), +43.2% over 6 months., Last month 126 of 150 licensed seats were active (84.0% utilization), approaching the seat cap.
graded checkscoverage_pct · headroom_arr · right_buyer · replier_not_buyer · usage_fact · email_under_150

Model's actual answer

show the model's full answer
C-0DDFC9A7

Seat coverage: 150 licensed seats ÷ 400 headcount × 100 = 37.5%.

Usage health:
  Monthly active users increased every month, from 88 in March to 126 in August: +38 users; 38 ÷ 88 × 100 = 43.2% growth.
  Active users as a share of licensed seats rose from 88 ÷ 150 = 58.7% to 126 ÷ 150 = 84.0%, indicating strengthening adoption.

Expansion headroom:
  Seats: 400 − 150 = 250 additional seats to cover current headcount.
  Current annual per-seat rate: $9,000 ÷ 150 = $60.
  Potential incremental ARR: 250 × $60 = $15,000, assuming full-headcount coverage at the current rate; not confirmed demand.

Who replied: Maria S., People Operations Coordinator, on 2026-09-02. She explicitly says she does not make purchasing decisions and offers an introduction.

Right buyer: Dana R., VP People. Maria S. confirms Dana R. owns budget and seat expansion. Last engaged: 2026-05-18.

Reply email

Hi Maria,

Thanks for the update. Monthly active users at C-0DDFC9A7 rose from 88 in March to 126 in August, so there’s a useful usage story to share with Dana.

An introduction would be helpful, if you’re comfortable making one. I can send Dana a short usage summary and discuss whether broader seat coverage makes sense for the team, with no assumption that you need to expand now.

Thanks,
Cole
onboarding-statusPASS 1.00
Plain-English question: What is done and what is missing in this customer's onboarding before the mid-onboarding call?
Prompt — what the model saw
Prep the CSM for the mid-onboarding call: what is complete (integrations connected, allowance set, admins added, first recognition given, first redemption), what is not, early engagement signals, and the three things to cover on the call. Mark nothing complete without a data field that shows it.
cost $0.1364 · 29s · in 55,572 / out 1,394 tok
✓ complete_items✓ incomplete_items✓ three_items✓ grounded

✓ Correct Answer (ground truth)

Plain-English answer: 25 days live: Slack connected, allowance set, admins added — but the HRIS integration is not connected, with early engagement signals and three call priorities.
The answer in one line: days_live = 25; company_alias = C-0D284E42
days_live25
call_itemsConnect the HRIS integration (integration_hris is empty) to automate user provisioning and offboarding., Get to first redemption (first_redemption_at is empty): walk admins through the rewards catalog and confirm allowance funding., Build the giving habit: daily active givers averaged 4.3 in week 1 vs 13.0 last week - agree on manager nudges and a recognition cadence.
checklist{"slack_integration_connected": {"complete": true, "field": "integration_slack", "value": "2026-08-12"}, "hris_integration_connected": {"complete": false, "field": "integration_hris", "value": ""}, "a
company_aliasC-0D284E42
engagement_signalsavg_active_givers_week1=4.3; avg_active_givers_last_week=13; latest_active_givers=15; days_with_activity=25
graded checkscomplete_items · incomplete_items · three_items · grounded

Model's actual answer

show the model's full answer
C-0D284E42 — mid-onboarding call prep

Complete, supported by onboarding_account.csv:
- Slack connected: integration_slack = 2026-08-12.
- Allowance set: allowance_set = 2026-08-13.
- Admins added: admins_added = 2.
- First recognition given: first_recognition_at = 2026-08-15 14:22.

Not confirmed complete:
- HRIS connection: integration_hris is blank.
- First redemption: first_redemption_at is blank.
Blank fields indicate missing evidence, not proof these steps never happened.

Early engagement signals, from onboarding_usage.csv:
- Daily active givers rose from 3 on 2026-08-11 to 15 on 2026-09-04: 15 − 3 = 12 more; 12 ÷ 3 × 100 = 400% growth.
- First seven days averaged 30 ÷ 7 = 4.29 active givers/day; last seven days averaged 91 ÷ 7 = 13.00. Giving activity is trending upward.
- Employee count and unique-giver totals are missing, so adoption rate and breadth of participation cannot be assessed.
- Active givers appear before the recorded first recognition. The data does not explain that discrepancy.

Three things to cover on the call:
1. Confirm HRIS status, identify any blocker, and agree on an owner and next step if it remains unconnected.
2. Confirm whether anyone has redeemed; if not, walk through redemption and identify any blocker.
3. Review growing giving activity, ask whether participation is broad or concentrated, and reconcile the active-giver dates with the first-recognition timestamp.
renewal-risk-conflicting-datesPASS 1.00
Plain-English question: Two systems disagree on renewal dates — which do you trust per account, and who is actually renewing inside 90 days?
Prompt — what the model saw
Produce a 90-day renewal risk brief. Two systems hold renewal dates and disagree on some accounts; decide which to trust per account and say why (multi-year contracts are known to be wrong in ChurnZero). For every renewal: company, CSM, ARR, date used, seat utilization, 3-month usage trend, risk rating with one sentence of evidence. Flag every disagreement. Close with total ARR renewing and ARR at risk.
cost $0.2195 · 70s · in 41,812 / out 7,118 tok
✓ total_renewing✓ arr_at_risk✓ disagreements_flagged✓ trust_rule

✓ Correct Answer (ground truth)

Plain-English answer: Chargebee wins per account (multi-year contracts are known wrong in ChurnZero), with each trusted date, the reason, and a 90-day-window flag.
The answer in one line: n_accounts = 20; n_disagreements = 5; snapshot_date = 2026-09-05 (+2 more below)
n_accounts20
n_disagreements5
accounts20 items — e.g. account_alias=C-0B144C78; csm=Cole Ingram; arr=30899.00; trusted_renewal_date=2026-11-02; trusted_source_why=systems agree (annual term); in_90d_window=True; dates_disagree=False; seat_utilization_pct=75.4; usage_3m_ratio=1.03; risk=low; evidence=3-month usage ratio 1.03 (last3 avg 103 vs prior3 100), seat utilization 75% | account_alias=C-0B20DB64; csm=Dana Mercer; arr=21770.00; trusted_renewal_date=2026-10-07; trusted_source_why=systems agree (annual term); in_90d_window=True; dates_disagree=False; seat_utilization_pct=56.6; usage_3m_ratio=1; risk=medium; evidence=3-month usage ratio 1.00 (last3 avg 295 vs prior3 295), seat utilization 57% | account_alias=C-0B344485; csm=Elena Sinclair; arr=64384.00; trusted_renewal_date=2026-11-16; trusted_source_why=systems agree (annual term); in_90d_window=True; dates_disagree=False; seat_utilization_pct=78; usage_3m_ratio=1.04; risk=low; evidence=3-month usage ratio 1.04 (last3 avg 241 vs prior3 231), seat utilization 78% | account_alias=C-0B369871; csm=Dana Mercer; arr=85128.00; trusted_renewal_date=2026-10-29; trusted_source_why=systems agree (annual term); in_90d_window=True; dates_disagree=False; seat_utilization_pct=75.1; usage_3m_ratio=1.04; risk=low; evidence=3-month usage ratio 1.04 (last3 avg 330 vs prior3 316), seat utilization 75%
arr_at_risk_90d359409.00
disagreementsaccount_alias=C-0B7D2C30; cz_renewal_date=2026-09-10; cb_renewal_date=2026-09-15; is_multi_year=True; trusted=2026-09-15 | account_alias=C-0BBE3E60; cz_renewal_date=2027-09-26; cb_renewal_date=2026-09-26; is_multi_year=True; trusted=2026-09-26 | account_alias=C-0BCDB8C2; cz_renewal_date=2027-09-18; cb_renewal_date=2026-09-18; is_multi_year=True; trusted=2026-09-18 | account_alias=C-0D2AB865; cz_renewal_date=2026-09-10; cb_renewal_date=2026-09-22; is_multi_year=True; trusted=2026-09-22 | account_alias=C-0F5D2323; cz_renewal_date=2026-09-10; cb_renewal_date=2026-09-29; is_multi_year=True; trusted=2026-09-29
snapshot_date2026-09-05
total_arr_renewing_90d1048715.00
trust_rulemulti-year contracts: Chargebee is authoritative (ChurnZero known wrong); otherwise systems agree or Chargebee wins
window2026-09-05 to 2026-12-04
graded checkstotal_renewing · arr_at_risk · disagreements_flagged · trust_rule

Model's actual answer

show the model's full answer
90-day renewal risk brief

$1,048,715 ARR renews in the assumed window; $549,338 is flagged at risk, including $359,409 rated High.

Scope and method
  No reporting start date was provided. Assuming September 1–November 29, 2026 inclusive, the 90 days following the latest usage month supplied.
  Company names are missing; account aliases identify companies below.
  Seat utilization = seats_used ÷ seats × 100.
  Three-month trend shows June → July → August active users; change = (August − June) ÷ June × 100.
  Analyst-defined risk: High if utilization <40% or usage declines ≥10%; otherwise Medium if utilization <70%; otherwise Low.
  These are screening ratings, not churn predictions. Seats_used and monthly active_users are distinct measures; their definitions and snapshot alignment are missing.

Date decisions and disagreements
  C-0B7D2C30: DISAGREEMENT — use Chargebee 2026-09-15, not ChurnZero 2026-09-10; the 36-month contract falls under the known ChurnZero multi-year error.
  C-0BCDB8C2: DISAGREEMENT — use Chargebee 2026-09-18, not ChurnZero 2027-09-18; the 36-month contract falls under the known ChurnZero multi-year error.
  C-0D2AB865: DISAGREEMENT — use Chargebee 2026-09-22, not ChurnZero 2026-09-10; the 24-month contract falls under the known ChurnZero multi-year error.
  C-0BBE3E60: DISAGREEMENT — use Chargebee 2026-09-26, not ChurnZero 2027-09-26; the 24-month contract falls under the known ChurnZero multi-year error.
  C-0F5D2323: DISAGREEMENT — use Chargebee 2026-09-29, not ChurnZero 2026-09-10; the 24-month contract falls under the known ChurnZero multi-year error.
  Every remaining account has a 12-month term and matching dates; use the agreed date shown below.

Renewals

C-0B7D2C30
  CSM: Dana Mercer | ARR: $65,901 | Date: 2026-09-15
  Seats: 274/476 = 57.6% | Usage: 97 → 94 → 84; (84−97)/97 = −13.4%.
  High: Usage fell 13.4% with only 57.6% seat utilization.

C-0BCDB8C2
  CSM: Cole Ingram | ARR: $54,427 | Date: 2026-09-18
  Seats: 232/424 = 54.7% | Usage: 127 → 118 → 110; (110−127)/127 = −13.4%.
  High: Usage fell 13.4% with only 54.7% seat utilization.

C-0D2AB865
  CSM: Elena Sinclair | ARR: $38,022 | Date: 2026-09-22
  Seats: 250/407 = 61.4% | Usage: 125 → 117 → 109; (109−125)/125 = −12.8%.
  High: Usage declined each month, falling 12.8% overall.

C-0BBE3E60
  CSM: Dana Mercer | ARR: $30,993 | Date: 2026-09-26
  Seats: 74/114 = 64.9% | Usage: 39 → 35 → 33; (33−39)/39 = −15.4%.
  High: Usage fell 15.4% despite 64.9% seat utilization.

C-0F5D2323
  CSM: Cole Ingram | ARR: $90,647 | Date: 2026-09-29
  Seats: 111/390 = 28.5% | Usage: 20 → 21 → 18; (18−20)/20 = −10.0%.
  High: Seat utilization is only 28.5%, and usage fell 10.0%.

C-0EC6999D
  CSM: Elena Sinclair | ARR: $79,419 | Date: 2026-10-03
  Seats: 31/112 = 27.7% | Usage: 17 → 16 → 15; (15−17)/17 = −11.8%.
  High: Seat utilization is only 27.7%, and usage declined each month.

C-0B20DB64
  CSM: Dana Mercer | ARR: $21,770 | Date: 2026-10-07
  Seats: 214/378 = 56.6% | Usage: 294 → 298 → 294; (294−294)/294 = 0.0%.
  Medium: Usage was unchanged overall, but seat utilization is only 56.6%.

C-0BBC4E7A
  CSM: Cole Ingram | ARR: $56,374 | Date: 2026-10-10
  Seats: 228/337 = 67.7% | Usage: 142 → 141 → 139; (139−142)/142 = −2.1%.
  Medium: Seat utilization is below 70%, alongside a small usage decline.

C-0FD551AB
  CSM: Elena Sinclair | ARR: $48,815 | Date: 2026-10-14
  Seats: 210/376 = 55.9% | Usage: 123 → 122 → 126; (126−123)/123 = +2.4%.
  Medium: Usage improved slightly, but only 55.9% of seats are used.

C-0F9F8F13
  CSM: Dana Mercer | ARR: $46,230 | Date: 2026-10-18
  Seats: 199/352 = 56.5% | Usage: 185 → 185 → 182; (182−185)/185 = −1.6%.
  Medium: Only 56.5% of seats are used, with a slight usage decline.

C-0BC34584
  CSM: Cole Ingram | ARR: $16,740 | Date: 2026-10-22
  Seats: 327/494 = 66.2% | Usage: 104 → 104 → 106; (106−104)/104 = +1.9%.
  Medium: Slight usage growth does not remove the below-70% seat-utilization flag.

C-0B7A7546
  CSM: Elena Sinclair | ARR: $35,062 | Date: 2026-10-25
  Seats: 182/205 = 88.8% | Usage: 64 → 65 → 63; (63−64)/64 = −1.6%.
  Low: Seat utilization is 88.8%, and the three-month usage decline is small.

C-0B369871
  CSM: Dana Mercer | ARR: $85,128 | Date: 2026-10-29
  Seats: 317/422 = 75.1% | Usage: 326 → 330 → 333; (333−326)/326 = +2.1%.
  Low: Seat utilization exceeds 70%, and usage increased each month.

C-0B144C78
  CSM: Cole Ingram | ARR: $30,899 | Date: 2026-11-02
  Seats: 169/224 = 75.4% | Usage: 101 → 101 → 106; (106−101)/101 = +5.0%.
  Low: Seat utilization is 75.4%, with 5.0% usage growth.

C-0FC4DBB8
  CSM: Elena Sinclair | ARR: $94,732 | Date: 2026-11-05
  Seats: 356/464 = 76.7% | Usage: 189 → 191 → 193; (193−189)/189 = +2.1%.
  Low: Seat utilization exceeds 70%, and usage increased each month.

C-0D5BBE3A
  CSM: Dana Mercer | ARR: $39,740 | Date: 2026-11-09
  Seats: 85/102 = 83.3% | Usage: 88 → 90 → 91; (91−88)/88 = +3.4%.
  Low: Seat utilization is 83.3%, with steady usage growth.

C-0FB9D5AF
  CSM: Cole Ingram | ARR: $63,158 | Date: 2026-11-13
  Seats: 144/199 = 72.4% | Usage: 173 → 173 → 176; (176−173)/173 = +1.7%.
  Low: Seat utilization exceeds 70%, and usage increased overall.

C-0B344485
  CSM: Elena Sinclair | ARR: $64,384 | Date: 2026-11-16
  Seats: 224/287 = 78.0% | Usage: 238 → 240 → 244; (244−238)/238 = +2.5%.
  Low: Seat utilization is 78.0%, with usage increasing each month.

C-0CB2C1B4
  CSM: Dana Mercer | ARR: $40,628 | Date: 2026-11-20
  Seats: 386/473 = 81.6% | Usage: 47 → 48 → 49; (49−47)/47 = +4.3%.
  Low under the stated rubric: Seat utilization is 81.6% and usage is growing, but 49 monthly active users versus 386 seats_used warrants validating the metric definitions.

C-22170CA1
  CSM: Cole Ingram | ARR: $45,646 | Date: 2026-11-24
  Seats: 251/294 = 85.4% | Usage: 143 → 148 → 146; (146−143)/143 = +2.1%.
  Low: Seat utilization is 85.4%, and August usage remains above June despite July’s peak.

ARR totals

  High = $65,901 + $54,427 + $38,022 + $30,993 + $90,647 + $79,419 = $359,409.
  Medium = $21,770 + $56,374 + $48,815 + $46,230 + $16,740 = $189,929.
  Low = $35,062 + $85,128 + $30,899 + $94,732 + $39,740 + $63,158 + $64,384 + $40,628 + $45,646 = $499,377.

  Total ARR renewing = $359,409 + $189,929 + $499,377 = $1,048,715.
  ARR at risk (High + Medium, full account ARR, not expected loss) = $359,409 + $189,929 = $549,338, or $549,338 ÷ $1,048,715 × 100 = 52.4%.
ticket-theme-synthesisPASS 1.00
Plain-English question: What are the real themes in this quarter's support tickets, ignoring the existing tags?
Prompt — what the model saw
Synthesize the quarter's support tickets into 4 to 7 themes. Do not trust existing tags; read the text. Per theme: name, count, share, distinct accounts, ARR affected, two ticket ids, one-line recommendation. Separate broad patterns from single-account noise and rank by ARR exposure, not volume.
cost $0.1420 · 35s · in 55,163 / out 2,508 tok
✓ theme_count_range✓ counts_verifiable✓ single_account_flagged✓ arr_ranked✓ ticket_ids

✓ Correct Answer (ground truth)

Plain-English answer: 80 tickets in 4-7 themes — points-not-posting leads (20 tickets, 25%, $31,100 ARR affected) — each with counts, distinct accounts, ARR, and a recommendation.
The answer in one line: total_tickets = 80; quarter_window = 2026-06-01 through 2026-08-31
total_tickets80
noteexisting tags are noisy/unreliable; classify by text
quarter_window2026-06-01 through 2026-08-31
ranking_by_arr_exposurehris_provisioning, redemption_failures, billing_dispute, points_not_posting, slack_sync
sample_ticket_ids80 items — e.g. IC-460004, IC-460016, IC-460001, IC-460041
single_account_noise_themebilling_dispute
themes{"points_not_posting": {"count": 20, "share_pct": 25.0, "distinct_accounts": 9, "arr_affected": "31100.00", "example_ticket_ids": ["IC-460001", "IC-460002"], "single_account": false}, "slack_sync": {"
graded checkstheme_count_range · counts_verifiable · single_account_flagged · arr_ranked · ticket_ids

Model's actual answer

show the model's full answer
Five themes, ranked by affected-account ARR, based on ticket text rather than tags. Shares use all 80 tickets supplied for June–August 2026. ARR counts each account once per theme; it is exposure, not confirmed revenue loss.

1. HRIS provisioning failures — cross-account pattern
   Count/share: 12 tickets; 12 ÷ 80 = 15.0%.
   Distinct accounts: 3.
   ARR affected: C-0B2213A9 ($36,000) + C-0F6C0F34 ($30,000) + C-0DDFC9A7 ($48,000) = $114,000.
   Evidence: IC-460059, IC-460060. New hires are not created, including skipped hires without logged errors.
   Recommendation: Reconcile HRIS rosters against provisioned users and investigate silent sync failures across these accounts.

2. Redemption and gift-card fulfillment failures — broad pattern
   Count/share: 18 tickets; 18 ÷ 80 = 22.5%.
   Distinct accounts: 7.
   ARR affected: C-0CEF69FD ($8,900) + C-0B827671 ($10,700) + C-0FCCD2DF ($9,600) + C-0F876796 ($8,700) + C-14264ABD ($11,000) + C-0D9CA315 ($9,600) + C-0B0F1BAB ($10,300) = $68,800.
   Evidence: IC-460025, IC-460024. Checkout fails, codes or emails do not arrive, and some failed orders still deduct points.
   Recommendation: Trace checkout through fulfillment and reconcile deducted points against failed or unfulfilled orders.

3. Incorrect seat counts and renewal pricing — single-account concentration
   Count/share: 16 tickets; 16 ÷ 80 = 20.0%.
   Distinct accounts: 1.
   ARR affected: C-0E9C27D1 ($52,000) = $52,000, not 16 times that amount.
   Evidence: IC-460069, IC-460078. Reports include 200 seats billed versus 150 licensed and incorrect renewal-tier pricing.
   Recommendation: Audit this account’s contracted seats, pricing, and invoices together; do not treat its repeated tickets as evidence of a broad billing problem.

4. Recognition points missing or delayed — broadest account reach
   Count/share: 20 tickets; 20 ÷ 80 = 25.0%.
   Distinct accounts: 9.
   ARR affected: C-0D3278C7 ($3,500) + C-0BF20542 ($4,500) + C-0D0B047C ($4,500) + C-0BE96399 ($2,700) + C-0D284E42 ($3,400) + C-0D6CC8E3 ($4,200) + C-21FEBCBB ($2,900) + C-0DD0626C ($2,500) + C-0B2895EF ($2,900) = $31,100.
   Evidence: IC-460004, IC-460016. Recognitions appear delivered but balances do not update, sometimes across whole teams.
   Recommendation: Reconcile delivered recognitions with points-ledger entries and investigate posting delays across accounts.

5. Slack sync, authentication, and command failures — cross-account pattern
   Count/share: 14 tickets; 14 ÷ 80 = 17.5%.
   Distinct accounts: 4.
   ARR affected: C-0B843542 ($4,400) + C-10A56B0F ($5,400) + C-0BA71F12 ($3,900) + C-8C2E8F00 ($5,200) = $18,900.
   Evidence: IC-460047, IC-460051. Recognition syncing stops, settings or reauthorization do not persist, and slash commands fail.
   Recommendation: Test sync-state persistence, reauthorization, and slash commands across the affected accounts.

Reconciliation: 12 + 18 + 16 + 20 + 14 = 80 tickets; shares total 100%. The themes cover 24 distinct accounts and $114,000 + $68,800 + $52,000 + $31,100 + $18,900 = $284,800 in account ARR. Resolution status, root causes, and actual churn or revenue loss are not provided.
marketing 5 tests
case-study-matchPASS 1.00
Plain-English question: Which three existing customers are the best case-study matches for this prospect?
Prompt — what the model saw
Given the prospect, rank the three most similar existing customers with public case studies for use as social proof. Similarity on industry, size band, use case, region. Cite the fields that drove each match. Do not name any customer without a case study and never include billing data or contact names.
cost $0.1060 · 21s · in 44,522 / out 1,122 tok
✓ top3✓ decoys_excluded✓ cites_fields✓ no_billing

✓ Correct Answer (ground truth)

Plain-English answer: Three ranked matches on industry, size, use case, and region (e.g. C-11C31562), with the higher-scoring decoy explicitly excluded for a mismatched field.
The answer in one line: prospect_alias = C-82AF3719
constraintsno_billing_data, no_contact_names, only_customers_with_has_case_study_true_named
decoys_excludedcompany_alias=C-D6217CAA; similarity_score=4; reason=no_case_study | company_alias=C-C153868F; similarity_score=4; reason=no_case_study
prospect_aliasC-82AF3719
prospect_fields_usedindustry, size_band, use_case, region
top3company_alias=C-11C31562; similarity_score=3 | company_alias=C-64171065; similarity_score=3 | company_alias=C-A13C193D; similarity_score=3
graded checkstop3 · decoys_excluded · cites_fields · no_billing

Model's actual answer

show the model's full answer
Using equal weight for industry, size band, use case, and region, the top three tie for first. Each has has_case_study=true.

1 (tie). C-64171065
   Matches: industry=Technology; size_band=Mid-Market; use_case=employee_recognition.
   Mismatch: region=NA-East versus prospect NA-West.
   Score: 1 + 1 + 1 + 0 = 3/4 (75%).

1 (tie). C-11C31562
   Matches: size_band=Mid-Market; use_case=employee_recognition; region=NA-West.
   Mismatch: industry=Manufacturing versus prospect Technology.
   Score: 0 + 1 + 1 + 1 = 3/4 (75%).

1 (tie). C-A13C193D
   Matches: industry=Technology; size_band=Mid-Market; region=NA-West.
   Mismatch: use_case=retention versus prospect employee_recognition.
   Score: 1 + 1 + 0 + 1 = 3/4 (75%).

Case-study URLs and content were not provided; eligibility is based solely on has_case_study=true.
channel-attribution-mathPASS 1.00
Plain-English question: Compute cost per SQM, cost per SQO, and pipeline per dollar for every channel over six months.
Prompt — what the model saw
Compute channel performance for the trailing 6 months. Paid: spend, SQMs, SQOs, cost per SQM, cost per SQO, SQM-to-SQO rate, pipeline amount, pipeline per dollar. Organic: volume, SQO rate, pipeline. Report a channel with spend and zero SQMs as undefined, not zero. Flag rows where SQO date precedes SQM date. Finish with a reallocation recommendation and your confidence given sample sizes.
cost $0.1398 · 37s · in 53,086 / out 2,720 tok
✓ undefined_channel✓ undefined_not_zero✓ flagged_rows✓ best_channel✓ reallocation

✓ Correct Answer (ground truth)

Plain-English answer: LinkedIn Ads at $960/SQM and $3,000/SQO, organic search at $90,000 pipeline on zero spend, paid social flagged at $18,000 with zero SQMs, and a channel winner named.
The answer in one line: best_pipeline_per_dollar_channel = paid_search
best_pipeline_per_dollar_channelpaid_search
confidence_notewebinars and referral have small SQM samples (<20); treat rates as directional.
flagged_rows_sqo_before_sqmCT-000041, CT-000044
months2026-03, 2026-04, 2026-05, 2026-06, 2026-07, 2026-08
per_channel{"linkedin_ads": {"spend_usd": 24000, "sqms": 25, "sqos": 8, "pipeline_amount": 96000, "cost_per_sqm": 960.0, "cost_per_sqo": 3000.0, "sqm_to_sqo_rate": 0.32, "pipeline_per_dollar": 4.0}, "organic_sea
reallocation_recommendationfrom=paid_social; to=paid_search; rationale=paid_social has spend but zero SQMs (cost per SQM undefined, not zero); paid_search has the highest pipeline per dollar.
undefined_channelspaid_social
graded checksundefined_channel · undefined_not_zero · flagged_rows · best_channel · reallocation

Model's actual answer

show the model's full answer
Period: March–August 2026, the six months supplied. Volume means contacts with an SQM date, not all leads. Attribution uses the supplied first-touch channel. SQOs below include all populated SQO dates; invalid date sequences are flagged separately.

Paid performance

Channel       Spend calculation   SQMs  SQOs  Cost/SQM          Cost/SQO
paid_search   $6,000 × 6 = $36,000   40    18  $36,000/40=$900   $36,000/18=$2,000
linkedin_ads  $4,000 × 6 = $24,000   25     8  $24,000/25=$960   $24,000/8=$3,000
paid_social   $3,000 × 6 = $18,000    0     0  Undefined          Undefined
webinars      $1,500 × 6 =  $9,000   12     5  $9,000/12=$750    $9,000/5=$1,800

Channel       SQM-to-SQO rate       Pipeline calculation     Pipeline per dollar
paid_search   18/40 = 45.00%        18 × $40,000 = $720,000   $720,000/$36,000 = 20.00
linkedin_ads   8/25 = 32.00%         8 × $12,000 =  $96,000   $96,000/$24,000 = 4.00
paid_social    0/0 = Undefined      $0                       $0/$18,000 = 0.00
webinars       5/12 = 41.67%         5 × $12,000 =  $60,000   $60,000/$9,000 = 6.67

paid_social has spend but no attributed contacts in the supplied contact file. Its unit costs and conversion rate are undefined, not zero. Its recorded pipeline per dollar is zero.

Organic performance

Channel         Volume (SQMs)  SQOs  SQO rate           Pipeline calculation
organic_search             30    10  10/30 = 33.33%     10 × $9,000 = $90,000
referral                   15     6   6/15 = 40.00%      6 × $8,000 = $48,000
Combined                   45    16  16/45 = 35.56%     $90,000 + $48,000 = $138,000

Organic acquisition costs and pre-SQM lead volumes are not provided.

Date-quality flags

Contact     Channel       SQM date    SQO date    Issue
CT-000041   linkedin_ads  2026-06-14  2026-06-09  SQO precedes SQM
CT-000044   linkedin_ads  2026-07-23  2026-07-18  SQO precedes SQM

If those two SQOs and their associated pipeline are excluded pending correction, linkedin_ads retains 25 SQMs but has:
  SQOs: 8 − 2 = 6
  SQM-to-SQO rate: 6/25 = 24.00%
  Cost per SQO: $24,000/6 = $4,000
  Pipeline: $96,000 − (2 × $12,000) = $72,000
  Pipeline per dollar: $72,000/$24,000 = 3.00

Reallocation recommendation and confidence

Pause paid_social pending an attribution check; if its zero volume is confirmed, shift that budget primarily to paid_search. Reduce linkedin_ads until the date errors are resolved, and test a smaller incremental allocation to webinars. paid_search produces the most pipeline per dollar; webinars has the lowest observed cost per SQO.

Confidence is moderate in this direction, low in the optimal allocation or returns from additional spend. paid_search has 40 SQMs/18 SQOs, linkedin_ads 25/8 reported (6 excluding flagged rows), and webinars only 12/5. Organic samples are also small: organic_search 30/10 and referral 15/6. These data do not establish statistical separation, scaling capacity, or realized revenue; pipeline is not revenue.
competitor-battlecard-updatePASS 1.00
Plain-English question: Update the Rivally battlecard with sourced pricing and where each side wins.
Prompt — what the model saw
Update the battlecard for the competitor. Sections: one-line positioning, pricing with source and date (newer source wins, note the conflict), where they win, where we win, objections and responses, recent changes, our 12-month win/loss record against them. Cite a snippet id for every factual claim. Rep opinion on a call is not a fact about the competitor. Anything from the old card you cannot re-source gets marked unverified.
cost $0.1453 · 45s · in 28,511 / out 3,257 tok
✓ competitor_named✓ pricing✓ win_loss✓ unverified_marked✓ citations

✓ Correct Answer (ground truth)

Plain-English answer: Rivally at $7 per user per month (the August source wins over the stale $5, conflict noted), a 13-7 win/loss record, and two unverified old claims flagged.
The answer in one line: competitor = Rivally
citation_ids_available25 items — e.g. S01, S02, S03, S04
competitorRivally
pricingvalue_per_user_per_month=7; source_snippet=S17; source_date=2026-08-12
rep_opinion_snippet_ids_not_factsS09, S21
unverified_old_claimsRivally lacks a Slack integration., Rivally was acquired by WorkHuman in 2025.
win_loss_12mowins=13; losses=7
graded checkscompetitor_named · pricing · win_loss · unverified_marked · citations

Model's actual answer

show the model's full answer
Battlecard: Rivally

One-line positioning
Points-based recognition with a survey add-on and EU data residency; the old card’s mid-market targeting remains unverified. [S02, S06, S15; existing_battlecard.md]

Pricing
- Latest supplied official source: Recognition Starter is $7/user/month, annual billing required, dated 2026-08-12. [S17]
- Conflict: January listed Recognition at $5; April listed Recognition Starter at $5. The newer August source supersedes both. Starter’s increase: ($7 − $5) ÷ $5 × 100 = 40%. [S03, S08, S17]
- Deal-specific reports, not published pricing: $6.50/user/month for a 500-seat prospect on an annual term, dated 2026-06-02; $7/user/month with a reported 15% discount for a three-year term, dated 2026-08-14. Discount arithmetic: $7 × (1 − 0.15) = $5.95/user/month. Do not present either as generally available pricing. [S13, S18]
- Pulse is a separately priced add-on, not bundled; its price is missing. [S23]

Where they win
These are reported strengths, not verified reasons for our recorded losses.
- Recognition experience: reviewers praise the points-based feed and engagement. [S02, S16]
- Setup and Slack: one mid-market reviewer reports setup under a week and Slack working out of the box. [S04]
- EU fit: an EU enterprise reviewer praises distributed-team suitability and multi-language support; EU data residency was announced generally available. [S12, S15]
- Support: one reviewer praises response times under four hours; this is not an SLA. [S22]

Where we win
- Analytics: an 800-seat prospect reportedly chose Bonusly over Rivally for analytics depth. Rivally reviewers also describe limited analytics and basic reporting. This supports an analytics angle, not a universal superiority claim. [S25, S02, S07]
- Other evaluation openings: reviewers report missing SCIM, lagging admin tooling, CSV-only analytics exports, and no bulk recognition editing. Bonusly’s corresponding capabilities are not supplied, so comparative wins remain unverified. [S10, S16, S20, S24]

Objections and responses
- “Rivally is only $5.” Response: the latest supplied official Starter price is $7/user/month with annual billing; $5 is superseded. [S08, S17]
- “Rivally is better for our EU team.” Response: acknowledge the reviewer-reported language strengths and announced EU residency. Validate country-specific rewards requirements: one reviewer reports a thinner EMEA catalog than the US catalog. No Bonusly EU comparison is supplied. [S12, S15, S14]
- “Rivally’s analytics are enough.” Response: test required dashboards and export workflows. Reviews flag basic reporting and CSV-only exports; one prospect reportedly chose Bonusly for analytics depth. [S07, S20, S25]
- “Rivally cannot integrate with Slack.” Response: remove that claim. A reviewer reports a working Slack integration; the old card’s blanket claim is contradicted. [S04]
- “Rivally discounts aggressively” / “its UI is clunky.” Response: neither is established competitor fact. Both are explicitly rep opinions; the reported three-year discount is only a deal-specific anecdote. [S09, S21, S18]

Recent changes
- 2026-03-05: announced Pulse, a lightweight engagement survey add-on; 2026-09-01: announced its exit from beta, separately priced rather than bundled. [S06, S23]
- 2026-05-09: hired an ex-Workday VP EMEA to lead European expansion; 2026-07-01: opened a Dublin office and announced EU data residency generally available. [S11, S15]
- 2026-08-12: published Starter pricing of $7/user/month, replacing the earlier $5 listing. [S08, S17]
- 2026-08-20: announced Microsoft Teams app v2 in public preview, not general availability. [S19]
- Old-card acquisition claim, “acquired by WorkHuman in 2025”: unverified; no supplied snippet supports it. The supplied funding announcement reports a $40M Series C, but does not independently disprove an acquisition. [existing_battlecard.md; S01]
- Old-card EU enterprise/multi-language claim: supported only as attributed reviewer feedback, not established segment-wide leadership. [S12]

Our 12-month win/loss record
Window: September 2025–August 2026, the 12 months covered by the supplied deal file. No snippet IDs accompany these rows; exact deal aliases are cited instead. A later rolling 12-month record cannot be established from the supplied data.

Wins:
Deal-A9FD43, Deal-7AA785, Deal-44C524, Deal-0D0CD6, Deal-D5B790, Deal-5C636E, Deal-67BE14, Deal-1B6969, Deal-F03E7B, Deal-072E31, Deal-F65C8F, Deal-E46EAB, Deal-1D2392.

Losses:
Deal-7767F5, Deal-5645A5, Deal-C6FFAA, Deal-D263E0, Deal-935746, Deal-9066A6, Deal-72A02F.

Arithmetic: 13 wins + 7 losses = 20 decisions. Win rate = 13 ÷ 20 × 100 = 65%; loss rate = 7 ÷ 20 × 100 = 35%. [deals_with_competitor.csv; aliases above]

Loss reasons are missing. The September prospect anecdote has no deal alias and is outside this window; it is not added to the record. [deals_with_competitor.csv; S25]
nurture-sequence-diagnosisPASS 1.00
Plain-English question: What is broken in each email sequence — rates, overlaps, tracking errors?
Prompt — what the model saw
Review the sequences. Per sequence: sent, open rate, reply rate, meeting rate, weakest step. Identify tracking errors (opened above sent), audience overlap between sequences, and the failure mode for anything under 2 percent reply. One change per weak sequence and which to fix first. Under 350 words.
cost $0.2083 · 44s · in 49,190 / out 2,517 tok
✓ tracking_error✓ overlap_pair✓ weakest_step✓ fix_priority

✓ Correct Answer (ground truth)

Plain-English answer: Per-sequence open, reply, and meeting rates with the weakest step (all three weakest at step 3), an opened-above-sent tracking error, and audience overlap flagged.
The answer in one line: fix_priority_first = Cold Outbound - HR Leaders
all_pairwise_overlaps['Cold Outbound - HR Leaders', 'Cold Outbound - People Ops'], ['Expansion Nurture', 'New Logo Nurture']
audience_overlap_pairCold Outbound - HR Leaders, Cold Outbound - People Ops
failure_modeCold Outbound - HR Leaders=opens healthy (~37%) but replies near zero from step 1 -> body/CTA not landing, not a deliverability or subject-line problem
fix_priority_firstCold Outbound - HR Leaders
per_sequence{"New Logo Nurture": {"sent": 1386, "opened": 490, "replied": 90, "meetings": 27, "open_rate": 0.3535, "reply_rate": 0.0649, "meeting_rate": 0.0195, "weakest_step": 3}, "Expansion Nurture": {"sent": 8
tracking_errorsequence=Expansion Nurture; step=2; sent=300; opened=340; issue=opened exceeds sent
under_2pct_replyCold Outbound - HR Leaders
graded checkstracking_error · overlap_pair · weakest_step · fix_priority

Model's actual answer

show the model's full answer
Rates = summed events ÷ summed sends × 100, not unique-contact conversion. Weakest step = lowest reply rate.

New Logo Nurture
Sent: 500+458+428 = 1,386.
Open: 490/1,386 = 35.35%; reply: 90/1,386 = 6.49%; meeting: 27/1,386 = 1.95%.
Weakest: step 3, 18/428 = 4.21% reply.

Expansion Nurture
Sent: 300+300+275 = 875.
Open: 565/875 = 64.57%, unreliable; reply: 59/875 = 6.74%; meeting: 12/875 = 1.37%.
Weakest: step 3, 12/275 = 4.36% reply.
Tracking error: step 2 opened 340 > 300 sent; 340/300 = 113.33%, exceeding sends by 40. Change: reconcile open-event counting before using opens.

Cold Outbound - HR Leaders
Sent: 600+595+590 = 1,785.
Open: 545/1,785 = 30.53%; reply: 8/1,785 = 0.45%; meeting: 0/1,785 = 0%.
Weakest: step 3, 1/590 = 0.17% reply; all steps are below 2%.
Failure mode: opens without replies or meetings. Targeting, message relevance, and CTA are possible causes, not proven; copy and delivery data are missing.
Change: test a revised HR-specific value proposition.

Cold Outbound - People Ops
Sent: 400+386+377 = 1,163.
Open: 340/1,163 = 29.23%; reply: 29/1,163 = 2.49%; meeting: 6/1,163 = 0.52%.
Weakest: step 3, 6/377 = 1.59% reply.
Failure mode: late-step response deterioration; fatigue versus weak content cannot be determined.
Change: replace step 3 with a new value proposition.

Audience overlap
New Logo Nurture ↔ Expansion Nurture: CT-000301, CT-000624.

Cold Outbound - HR Leaders ↔ Cold Outbound - People Ops: 21 contacts:
CT-000849, CT-000884, CT-000890, CT-000908, CT-001033, CT-001097, CT-001101, CT-001103, CT-001105, CT-001130, CT-001153, CT-001159, CT-001217, CT-001227, CT-001236, CT-001255, CT-001258, CT-001277, CT-001285, CT-001311, CT-001345.

Overlap establishes shared enrollment, not simultaneous sends; dates are missing.

Fix first: Cold Outbound - HR Leaders, with the most sends, lowest reply rate, and zero meetings.
weekly-marketing-goals-rollupPASS 1.00
Plain-English question: Are we on pace on the quarter's marketing goals?
Prompt — what the model saw
Produce the weekly marketing goals update for the active quarter: SQMs, SQOs, DS2s, closed-lost MIA rate, same-quarter close count, active pipeline coverage against target. For each: QTD actual, target, delta, pace (on/behind/ahead given days elapsed). One paragraph on what moved this week.
cost $0.1158 · 25s · in 24,524 / out 1,381 tok
✓ metrics_present✓ deltas✓ coverage_ratio✓ pace✓ paragraph

✓ Correct Answer (ground truth)

Plain-English answer: SQMs ahead of pace, SQOs on pace, DS2s behind (35 short of expected at day 66 of 92), with deltas and same-quarter closes.
The answer in one line: days_in_quarter = 92; days_elapsed = 66; coverage_ratio_active_pipeline = 0.75; elapsed_fraction = 0.7174 (+1 more below)
days_in_quarter92
coverage_ratio_active_pipeline0.75
days_elapsed66
elapsed_fraction0.7174
metrics{"SQMs": {"qtd_actual": 230, "target": 300, "delta": -70, "expected_at_day_elapsed": 215.22, "pace": "ahead"}, "SQOs": {"qtd_actual": 84, "target": 120, "delta": -36, "expected_at_day_elapsed": 86.09,
pace_ruleexpected = target * days_elapsed/days_in_quarter; within +/-5% of expected = on, above = ahead, below = behind; lower_better metrics invert.
this_week_paragraph_required1
graded checksmetrics_present · deltas · coverage_ratio · pace · paragraph

Model's actual answer

show the model's full answer
Q3-2026 weekly marketing goals update
Provided quarter: July 1–September 30. Elapsed: 66 ÷ 92 = 71.74%.

Delta = actual − target. Volume pace compares actual with target × 66 ÷ 92. Rate and pipeline coverage are assessed against their full targets, not prorated.

Metric                     QTD actual   Target       Delta        Pace
SQMs                       230          300          −70          Ahead
SQOs                       84           120          −36          Behind
DS2s                       40           75           −35          Behind
Closed-lost MIA rate        20%          10% max      +10 pp       Behind
Same-quarter close count   10           20           −10          Behind
Active pipeline coverage   75%          100%         −25 pp       Behind

Arithmetic:
  SQMs: 230 − 300 = −70; expected by day 66 = 300 × 66 ÷ 92 = 215.22; actual is 14.78 ahead.
  SQOs: 84 − 120 = −36; expected = 120 × 66 ÷ 92 = 86.09; actual is 2.09 behind.
  DS2s: 40 − 75 = −35; expected = 75 × 66 ÷ 92 = 53.80; actual is 13.80 behind.
  Closed-lost MIA rate: 5 ÷ 25 = 20%; 20% − 10% = +10 percentage points, worse than target.
  Same-quarter closes: 10 − 20 = −10; expected = 20 × 66 ÷ 92 = 14.35; actual is 4.35 behind.
  Active pipeline: 3,000,000 ÷ 4,000,000 = 75% coverage; 3,000,000 − 4,000,000 = −1,000,000. As a point-in-time balance, coverage is below target; its weekly trajectory is unavailable.

What moved this week: Cannot determine from the supplied QTD snapshot. No prior-week actuals or weekly activity data were provided.
communication 5 tests
ceo-slack-compressionPASS 1.00
Plain-English question: Compress the forecast analysis into one Slack message the CEO can read in ten seconds.
Prompt — what the model saw
Compress the forecast analysis below into a single Slack message for the CEO. Rules: at most three sentences; under 80 words; lead with the decision or the so-what; keep every distinct number that matters and the one caveat that changes the conclusion; cut any sentence that does not carry a claim, a number, or a directive. No em dashes, no throat-clearing phrases.
cost $0.1001 · 18s · in 43,575 / out 729 tok
✓ max_3_sentences✓ under_80_words✓ weighted_present✓ caveat_present✓ no_em_dash

✓ Correct Answer (ground truth)

Plain-English answer: Three sentences, under 80 words, decision first, every distinct number kept plus the one caveat — all checked by deterministic text rules.
The answer in one line: weighted_number = 115976; must_keep = data-quality caveat — deleting the bad news fails
weighted_number115,976
limitmax 3 sentences, max 80 words, no em-dashes
must_keepdata-quality caveat — deleting the bad news fails
graded checksmax_3_sentences · under_80_words · weighted_present · caveat_present · no_em_dash

Model's actual answer

show the model's full answer
Use 115,976.75 as the Q3 forecast: 100% × 44,729 COMMIT + 35% × 203,565 BEST_CASE = 44,729 + 71,247.75; PIPELINE counts zero. Only 54 of 86 deals qualify; exclude 32 totaling 227,575 after September 30, including Deal-D348E1 (13,770 COMMIT, October 15) and 9 BEST_CASE totaling 28,240. Treat this as unvalidated: owners are blank on 85/86, preventing rep/quota checks, and why-buys on 71/86, including all 7 in-quarter COMMIT deals.
stalled-deal-followup-emailPASS 1.00
Plain-English question: Write a follow-up email for this stalled deal — under 120 words, one prior touch, one ask.
Prompt — what the model saw
Write a follow-up email for this stalled deal. Under 120 words. Reference one specific prior touch from the engagement history, make one clear ask, and match the tone of the prior thread. No invented details.
cost $0.0740 · 9s · in 11,873 / out 195 tok
✓ word_cap✓ references_touch✓ single_ask

✓ Correct Answer (ground truth)

Plain-English answer: An email that references the July 28 milestone-awards demo, makes one clear ask, and invents nothing — Deal-0D2F7A, 31 days since last engagement.
The answer in one line: amount = 5100; word_cap = 120; days_since_last_engagement = 31; deal_alias = Deal-0D2F7A (+4 more below)
amount5,100
days_since_last_engagement31
word_cap120
allowed_facts14 items — e.g. Deal-0D2F7A, $5,100, DS3, Alex Franklin
allowed_newENG-00001, ENG-00002, ENG-00003, ENG-00004, ENG-00005
deal_aliasDeal-0D2F7A
must_referenceengagement_id=ENG-00002; date=2026-07-28; type=meeting; why=most substantive prior touch; concrete prospect reactions to milestone awards and points catalog
ownerAlex Franklin
single_cta1
stageDS3
tonewarm, concise, professional (matches prior thread)
graded checksword_cap · references_touch · single_ask

Model's actual answer

show the model's full answer
Subject: Following up on the demo recap

Hi team,

I’m following up on my August 5 email with the July 28 demo recap and pricing for 150 seats. Automated milestone awards and the points catalog stood out in our conversation, especially with birthdays and anniversaries tracked manually today.

Is recognition still a priority for your team?

Best,
Alex Franklin
funnel-digest-paragraphPASS 1.00
Plain-English question: Write the weekly GTM digest — one paragraph per team, every number traceable.
Prompt — what the model saw
Write the weekly GTM digest: one paragraph per team (Marketing, Sales, CS, PLG) in Bonusly brand voice. Every number must trace to the extracts. No customer contact names, no employee performance commentary, no compensation. Four paragraphs, under 400 words total.
cost $0.1057 · 19s · in 23,672 / out 619 tok
✓ four_paragraphs✓ word_cap✓ teams

✓ Correct Answer (ground truth)

Plain-English answer: Four paragraphs (Marketing, Sales, CS, PLG) using only the allowed numbers, no contact names, no performance commentary, under 400 words.
The answer in one line: word_cap = 400
word_cap400
allowed_numbers16 items — e.g. 2, 3, 9, 08
bannedcustomer contact names, employee performance commentary, compensation
required_paragraphsMarketing, Sales, CS, PLG
graded checksfour_paragraphs · word_cap · teams

Model's actual answer

show the model's full answer
Marketing: SQMs finished at 46 against a weekly target of 52, a shortfall of 6 (46 − 52 = −6). Webinars delivered 18 of those SQMs. Paid search cost per SQM held at $150.

Sales: Sales converted 14 SQOs and set 9 DS2 meetings, creating $310,000 in new pipeline. Same-quarter close count was 3 for the week.

CS: The team saved 2 renewals, and NPS moved to 61. The prior NPS was not provided, so the size and direction of the change are unknown. There are 3 open red-flag accounts heading into next week.

PLG: PLG added 412 new signups, with activation at 31%. Another milestone: 38 companies reached the aha moment of 10 recognition gives.
partner-digest-sectionPASS 1.00
Plain-English question: Write the partner digest so every partner in the roster appears exactly once.
Prompt — what the model saw
Write the partner digest for the period. Every partner in the roster appears once: active plays get status with numbers, quiet partners get one line. Partner-sourced pipeline count and amount must match the deal data. Nothing omitted, nothing added.
cost $0.1098 · 20s · in 24,551 / out 850 tok
✓ all_partners✓ pipeline_amount✓ quiet_one_line

✓ Correct Answer (ground truth)

Plain-English answer: Four active partners with status and numbers, four quiet partners with one line each, and partner-sourced pipeline that reconciles to the deal data.
The answer in one line: rule = every roster partner named exactly once; quiet partners get one line; no partners outside roster.
active_no_dealsWorkWell Group
partner_sourced_pipelinecount=5; amount=550000
per_partner_pipeline{"Apex Rewards Co": {"count": 2, "amount": 275000}, "HRCloud Partners": {"count": 1, "amount": 140000}, "CultureBridge": {"count": 2, "amount": 135000}}
quiet_partnersRecogniq, KudosWave, PeopleFirst Advisors, TotalPerk
rosterpartner_name=Apex Rewards Co; status=active | partner_name=HRCloud Partners; status=active | partner_name=CultureBridge; status=active | partner_name=WorkWell Group; status=active | partner_name=Recogniq; status=quiet | partner_name=KudosWave; status=quiet … (+2 more)
ruleevery roster partner named exactly once; quiet partners get one line; no partners outside roster.
graded checksall_partners · pipeline_amount · quiet_one_line

Model's actual answer

show the model's full answer
Partner digest | 2026-08-24 to 2026-09-04

Partner-sourced pipeline: 2 + 1 + 2 + 0 = 5 deals; 275,000 + 140,000 + 135,000 + 0 = 550,000. All five deals have utm_source = Partner. Currency was not provided.

Apex Rewards Co: 2 sourced opportunities, both in DS1. Deal-DDAAF2 (180,000) + Deal-2CF33E (95,000) = 275,000. Co-webinar confirmed for 09-15.

HRCloud Partners: 1 sourced opportunity. Deal-F1CDA5 = 140,000. Security review completed; opportunity moved to DS2.

CultureBridge: Lunch-and-learn produced 2 sourced opportunities, both early stage. Deal-096E1D (60,000) + Deal-067213 (75,000) = 135,000.

WorkWell Group: 0 sourced deals; amount = 0. Q4 joint-playbook restart under discussion; planning call booked for 09-09.

Recogniq: Quiet; no activity or sourced deals provided.
KudosWave: Quiet; no activity or sourced deals provided.
PeopleFirst Advisors: Quiet; no activity or sourced deals provided.
TotalPerk: Quiet; no activity or sourced deals provided.
rfp-answer-from-policyPASS 1.00
Plain-English question: Answer the 10 RFP items from the policy excerpts only — no guessing.
Prompt — what the model saw
Answer the 10 questionnaire items from the policy excerpts only. Cite the excerpt id for every answer. Where the excerpts do not answer the question, mark it unanswerable and say what document would; do not guess. Tone: vendor response, direct.
cost $0.0804 · 13s · in 12,467 / out 697 tok
✓ answered_count✓ unanswerable_flagged✓ citations✓ no_guess

✓ Correct Answer (ground truth)

Plain-English answer: Seven answered with excerpt citations (Q1-Q7), three flagged unanswerable with the document that would answer them (Q8-Q10).
The answer in one line: answered_count = 7; rule = unanswerable items must be flagged and the document that would answer them named; do not guess.
answered_count7
answer_map{"Q1": ["P1"], "Q2": ["P1"], "Q3": ["P2"], "Q4": ["P3"], "Q5": ["P4"], "Q6": ["P5"], "Q7": ["P6"]}
ruleunanswerable items must be flagged and the document that would answer them named; do not guess.
unanswerableQ8, Q9, Q10
valid_citation_idsP1, P2, P3, P4, P5, P6, P7, P8
graded checksanswered_count · unanswerable_flagged · citations · no_guess

Model's actual answer

show the model's full answer
Q1: Yes. All customer data is encrypted at rest using AES-256. [P1]

Q2: Yes. Data in transit is encrypted with TLS 1.2 or higher. [P1]

Q3: Yes. Bonusly maintains a SOC 2 Type II report, renewed annually. Prospects can obtain the report under NDA. [P2]

Q4: Customer data is retained for 90 days after contract termination, then deleted from production systems. [P3]

Q5: The current subprocessor list is published at trust.bonusly.com. Customers receive 30 days’ notice before a new subprocessor is added. The notification method is not specified; the subprocessor notification policy would be needed to confirm it. [P4]

Q6: Yes. SAML 2.0 single sign-on is available on the Enterprise plan. [P5]

Q7: Production databases are backed up daily. Backups are encrypted and retained for 35 days. [P6]

Q8: Unanswerable. The excerpts do not state whether Bonusly holds ISO 27001 certification. A current ISO 27001 certificate would be needed to confirm certification. [P1–P8]

Q9: Unanswerable. The excerpts do not specify a contractual uptime SLA percentage. The contractual Service Level Agreement would be needed. [P1–P8]

Q10: Unanswerable. The excerpts do not state whether Bonusly will sign a HIPAA Business Associate Agreement. A BAA policy or approved BAA template would be needed. [P1–P8]
ops-maintenance 5 tests
skill-file-reconciliationPASS 1.00
Plain-English question: Audit a skill set against its manifest and flag every conflict with a severity and an action.
Prompt — what the model saw
Reconcile the skill set below against its manifest. Find and report (with severity CRITICAL/WARNING/INFO and an action type TRIM_DESC, UPDATE_BODY, DELETE_SKILL, MERGE, or REVIEW): (1) any two skills whose ALWAYS-trigger phrases overlap/duplicate, (2) any circular delegation chain between skills (name it), (3) any delegation target that does not exist (dangling, name it), (4) any version conflict and which skill should survive, (5) how many manifest descriptions exceed 1,024 characters, (6) any hardcoded page ids, dates, or person names in skill bodies, (7) manifest drift in BOTH directions (files with no manifest row, and manifest rows with no file). Make one proposal per finding; rewrite nothing.
cost $0.9642 · 74s · in 331,441 / out 5,601 tok
✓ duplicate_trigger_found✓ circular_chain✓ dangling_target✓ oversized_desc_count✓ hardcoded_values✓ manifest_drift✓ severity_and_action

✓ Correct Answer (ground truth)

Plain-English answer: A severity-ranked list of TRIM_DESC / UPDATE_BODY / DELETE_SKILL / MERGE / REVIEW actions covering every rule violation found.
The answer in one line: files_audited = 14; rank = CRITICAL / WARNING / INFO with one fix action per issue
files_audited14
findingsduplicate trigger, circular delegation chain, dangling file reference, oversized description, hardcoded values, manifest drift
rankCRITICAL / WARNING / INFO with one fix action per issue
graded checksduplicate_trigger_found · circular_chain · dangling_target · oversized_desc_count · hardcoded_values · manifest_drift · severity_and_action

Model's actual answer

show the model's full answer
Scope: only the supplied manifest and skill bodies. Missing targets below are absent from this supplied set; existence elsewhere is unknown.

1. Overlapping triggers

WARNING | MERGE | comms-drafter + email-drafter
Duplicate triggers include “write me an email,” “draft a follow-up,” “what should I say,” “bump email,” and “contract nudge.” Both also handle reviewing and rewriting existing emails.
Proposal: merge email-drafter into comms-drafter, preserving signature retrieval and updating callers.

WARNING | UPDATE_BODY | pipeline-intelligence-report + weekly-pipeline-report
ALWAYS-trigger overlap includes “pipeline update” versus “run the pipeline update,” and “what's the pipeline look like” versus “what does pipeline look like.”
Proposal: establish exclusive routing: scored active-deal views to pipeline-intelligence-report; weekly funnel, generation, and bookings performance to weekly-pipeline-report.

WARNING | UPDATE_BODY | pipeline-intelligence-report + sales-forecast
pipeline-intelligence-report claims VP requests for “pipeline health or forecast context”; sales-forecast claims current-quarter outlook, deal confidence, and quarter-close risk.
Proposal: reserve quarter revenue forecasts for sales-forecast and active-pipeline scoring for pipeline-intelligence-report.

WARNING | UPDATE_BODY | pipeline-intelligence-report + next-to-close
“Never answer pipeline questions inline” conflicts with next-to-close’s explicit conversational shortlist and instruction not to use pipeline-intelligence-report for that purpose.
Proposal: add a shortlist exception to pipeline-intelligence-report’s broad trigger rule.

INFO | REVIEW | analysis-validator + signalforge-claim-compressor + signalforge-feedback
Their mandatory scopes overlap on reports and analytical outputs. This is a staged workflow, not duplicate functionality, but analysis-validator calls itself the “last thing” before publication while the other skills run afterward.
Proposal: review and approve one explicit sequence with distinct finality rules.

2. Circular delegation

CRITICAL | UPDATE_BODY | deal-strategy-coach → email-drafter → deal-strategy-coach
deal-strategy-coach requires email-drafter for manager-to-prospect emails. email-drafter routes strategic diagnosis back to deal-strategy-coach. This is a conditional handoff cycle, not proof of an unavoidable infinite loop.
Proposal: make email-drafter terminal when invoked with an already-completed coaching diagnosis.

3. Dangling delegation and required skill targets

Each target below is absent from both the supplied files and manifest. One proposal per target:

CRITICAL | REVIEW | bonusly-data-questions
Caller: analysis-validator §12.4.
Proposal: verify and register the intended specialist.

CRITICAL | REVIEW | bonusly-product-questions
Caller: analysis-validator §12.4.
Proposal: verify and register the intended specialist.

CRITICAL | REVIEW | bonusly-business-reporting-questions
Caller: analysis-validator §12.4.
Proposal: verify and register the intended specialist.

CRITICAL | REVIEW | bonusly-rewards-questions
Caller: analysis-validator §12.4.
Proposal: verify and register the intended specialist.

CRITICAL | REVIEW | bonusly-ppp-questions
Caller: analysis-validator §12.4.
Proposal: verify and register the intended specialist.

CRITICAL | REVIEW | bonusly-feature-flag-questions
Caller: analysis-validator §12.4.
Proposal: verify and register the intended specialist.

CRITICAL | REVIEW | bonusly-deal-desk-questions
Caller: analysis-validator §12.4.
Proposal: verify and register the intended specialist.

CRITICAL | REVIEW | bonusly-datadog-questions
Caller: analysis-validator §12.4.
Proposal: verify and register the intended specialist.

CRITICAL | REVIEW | bonusly-brand
Required by comms-drafter, email-drafter, and sales-forecast; also referenced by signalforge-claim-compressor.
Proposal: verify and register the required brand skill.

CRITICAL | REVIEW | prospect-research-multithreading
Invoked by comms-drafter, email-drafter, and deal-strategy-coach.
Proposal: verify and register the intended research skill.

CRITICAL | REVIEW | signalforge-reports
Required through skill-file reads by pipeline-intelligence-report and weekly-pipeline-report.
Proposal: verify and register the required report-design skill.

WARNING | REVIEW | skill-orchestrator
Referenced for cascading updates by analysis-validator and terminal-step registration by signalforge-feedback; absent from the supplied set. This is a configuration dependency, not an explicit runtime delegation.
Proposal: verify the orchestrator and its registration contract.

4. Version conflicts

WARNING | UPDATE_BODY | analysis-validator
The current header, changelog, and footer identify v3.6, but the validation-trail template identifies “analysis-validator v3.2.” Execution instructions also retain G1-A–G1-H and G2-A–G2-E ranges despite checks extending through G1-L and G2-F.
Proposal: retain analysis-validator v3.6 and align its stale template and execution ranges.

INFO | REVIEW | No competing versioned skill files supplied
The manifest has no version column, so a manifest-versus-file version comparison is unavailable. pipeline-intelligence-report v6 referencing a v4 design system does not establish a skill-version conflict.
Proposal: review version metadata coverage before considering any version-based deletion.

5. Description length

INFO | REVIEW | No manifest descriptions exceed 1,024 characters
Arithmetic: count(description_chars > 1,024) = 0 of 14.
Largest declared length = 1,006; remaining allowance = 1,024 − 1,006 = 18 characters.
Proposal: retain current description lengths; no TRIM_DESC is warranted by the supplied manifest.

6. Hardcoded IDs, dates, and person names in bodies

WARNING | UPDATE_BODY | analysis-validator
Dates include April 26, 2026; May 9, 2026; March 28, 2023; May 4, 2026; May 2026; and the example period “Q1 2026 · Jan 1 – Mar 31.”
Person names include Manish, Amani, Bryce Harmon, Hugo Lindqvist, Dana Mercer, Alex Franklin, Cole Ingram, Gavin Porter, Alaina Loori, Shealagh Coughlin, Colleen Perry, Ellie Barton, Ashley Reyer, Ashley Le, Megan Franz, Elena Sinclair, Tracy, Youssef Elkhateeb, Amanda Czenkus, Ben Castelli, Amani Phipps, John Thomas, and Yasmin Wahid.
Proposal: replace operational roster and escalation assumptions with runtime resolution, keeping historical dates explicitly historical.

WARNING | UPDATE_BODY | closed-lost-analysis
Hardcoded dates include “demo on 4/13,” May 2026, “May 4–12,” and “June 3.”
Proposal: separate historical examples from reusable operational instructions and use runtime dates for execution.

WARNING | UPDATE_BODY | deal-strategy-coach
Hardcoded page ID: 2257879045. Dated references include April 2026 and “Pricing — 2026.” Named routing dependencies include Perseus and Farid.
Proposal: move the playbook destination, pricing effective period, and named routing assignments into verified configuration.

WARNING | UPDATE_BODY | model-selection
Hardcoded dates include last_checked: 2026-05-19, changelog date 2026-05-19, April 14, 2026, and model knowledge-cutoff dates.
Proposal: preserve these as dated registry metadata and enforce the stated freshness check before use.

WARNING | UPDATE_BODY | partner-digest
Page IDs: 2286321666, 2265382925, 2236940297, 2237825028, 2239365136, 2238283777. Destination folder ID: 2286616609.
Dates include May 19, 2026; June 2, 2026; May 16, 2026; 2026-05-17; and Q2/Q3 2026.
Names include Amani, Amani Phipps, Kelli, Jen Lee, Hani, Bryce, and Sara; the body also hardcodes Amani’s Slack user ID U03QLMBL7AR.
Proposal: configure destinations and resolve current owners/contacts dynamically, retaining dated issues only as historical references.

WARNING | UPDATE_BODY | pipeline-intelligence-report
The body hardcodes May 2026, March 2023, and the AE roster Bryce Harmon, Dana Mercer, Cole Ingram, Alex Franklin, and Gavin Porter.
Proposal: resolve the current roster at runtime and separate dated schema/design notes from current operating assumptions.

WARNING | UPDATE_BODY | sales-forecast
Hardcoded parent page ID: 2232582148. Names include Alaina and changelog references to Elena. Dates include April 27, 2026 and the example July 9, 2026. Operational headings still say “Open Q2 Deals” and “Q2 Narrative” despite quarter-agnostic instructions.
Proposal: configure the destination and leadership role, and derive all operational quarter labels dynamically.

WARNING | UPDATE_BODY | signalforge-feedback
Hardcoded page IDs: 2295136266, parent 2234417154, and activation-checklist target 2247295002. The checklist targets the Build Log rather than the stated Feedback Log. “Gavin Porter Rep Diagnostic” is a named example.
Proposal: use one configured feedback destination consistently across logging and activation checks.

WARNING | UPDATE_BODY | stale-pipeline-report
Hardcoded example dates: 5/15, 5/19, and 5/7; changelog date: 2026-06-10.
Proposal: mark these strictly as examples/history and derive actual outreach dates from each deal’s evidence.

WARNING | UPDATE_BODY | weekly-pipeline-report
Hardcoded person: Ben Lavin, also referenced as Ben throughout.
Operational dates include “April 1 – June 30, 2026,” Q2/Q3 buckets, Q2 targets, and Q1 2026 context.
Proposal: resolve the intended reviewer at runtime and parameterize the reporting quarter while preserving explicitly labeled historical comparisons.

INFO | REVIEW | signalforge-claim-compressor
Hardcoded changelog date: 2026-05-09; attribution identifier: JuliusBrussee/caveman.
Proposal: retain these only as historical provenance, not operational inputs.

7. Manifest drift in both directions

INFO | REVIEW | Files without manifest rows
None. All 14 supplied skill files have matching rows.
Arithmetic: 14 supplied files − 14 matched files = 0.
Proposal: retain the current file-to-manifest membership.

INFO | REVIEW | Manifest rows without files
None. All 14 manifest rows have matching supplied files, and declared_name matches each file alias.
Arithmetic: 14 manifest rows − 14 matched rows = 0.
Proposal: retain the current manifest-to-file membership.
runbook-from-slack-threadPASS 1.00
Plain-English question: Turn the resolved incident thread into a runbook.
Prompt — what the model saw
Turn the resolved incident thread into a runbook: ordered steps with the exact command or action taken, who ran it, how success was verified, and rollback for each step that changed state. Every step must trace to a message in the thread. Add nothing the thread does not contain; where the thread is unclear, mark the step as needing confirmation.
cost $0.0822 · 25s · in 32,265 / out 1,574 tok
✓ ordered_steps✓ rollback✓ needs_confirmation✓ sourced

✓ Correct Answer (ground truth)

Plain-English answer: Ordered steps from the thread, each with the exact command, who ran it, how success was verified, and rollback where state changed.
The answer in one line: rule = no steps beyond what the thread contains; every step traces to a message id.
ambiguous_steporder=4; msg=M05; mark=needs-confirmation
ruleno steps beyond what the thread contains; every step traces to a message id.
stepsorder=1; msg=M01; actor=Bryce Harmon; action=acknowledge PagerDuty alert, take IC; state_change=False; verified_by=M01 | order=2; msg=M02; actor=Farid Osman; action=check queue depth; command=bundle exec rake sidekiq:queue_depth; state_change=False; verified_by=M02 (48,213 pending) | order=3; msg=M04; actor=Farid Osman; action=pause auto-enqueue via feature flag; command=bin/rails runner 'FeatureFlag.disable(:auto_recognition_enqueue)'; state_change=True; rollback=bin/rails runner 'FeatureFlag.enable(:auto_recognition_enqueue)'; verified_by=M07 (queue depth falling) | order=4; msg=M05; actor=Elena Sinclair; action=cleared the dead set; state_change=True; needs_confirmation=thread gives no command or method for clearing the dead set | order=5; msg=M06; actor=Bryce Harmon; action=scale reward-worker to 6 replicas; command=kubectl scale deployment/reward-worker --replicas=6; state_change=True; rollback=kubectl scale deployment/reward-worker --replicas=3; verified_by=M07/M08 | order=6; msg=M08; actor=Cole Ingram; action=verify queue drained and error rate baseline; command=bundle exec rake sidekiq:queue_depth; state_change=False; verified_by=M08 (returns 0) … (+2 more)
graded checksordered_steps · rollback · needs_confirmation · sourced

Model's actual answer

show the model's full answer
1. Acknowledge the alert and take incident command.
   Who: Bryce Harmon.
   Action: Acknowledged the PagerDuty alert for reward-worker queue depth > 10k and took IC.
   Verification: Acknowledgment and IC assignment stated in the thread; no separate verification recorded.
   Rollback: Not provided.
   Source: [M01]

2. Check reward queue depth.
   Who: Farid Osman.
   Command: `bundle exec rake sidekiq:queue_depth`
   Verification: Returned 48,213 pending jobs; normal is under 500.
   Source: [M02]

3. Inspect the dead set.
   Who: Farid Osman.
   Action: Reported 112 dead jobs, all Redis::TimeoutError from around 13:58.
   Exact inspection command: Not provided; needs confirmation.
   Verification: Findings reported in the thread; no separate verification recorded.
   Source: [M03]

4. Pause enqueue.
   Who: Farid Osman.
   Command: `bin/rails runner 'FeatureFlag.disable(:auto_recognition_enqueue)'`
   Verification: No direct confirmation that enqueue stopped; needs confirmation. Subsequent queue decline was reported after worker scaling, so it does not isolate this change.
   Rollback: `bin/rails runner 'FeatureFlag.enable(:auto_recognition_enqueue)'`
   Source: [M04]; subsequent observation [M07]

5. Clear the dead set.
   Who: Elena Sinclair.
   Action: Cleared the dead set while in the console.
   Exact command: Not provided; needs confirmation.
   Verification: Completion stated, but no post-clear dead-set count recorded; needs confirmation.
   Rollback: Not provided; needs confirmation.
   Source: [M05]

6. Scale reward workers from 3 to 6 replicas.
   Who: Bryce Harmon.
   Command: `kubectl scale deployment/reward-worker --replicas=6`
   Verification: No replica-readiness check recorded; needs confirmation. Farid Osman subsequently reported queue depth at 9,400 and falling approximately 1,200/min.
   Rollback: `kubectl scale deployment/reward-worker --replicas=3`
   Source: [M06], [M07]

7. Verify the queue drained and errors returned to baseline.
   Who: Cole Ingram.
   Command: `bundle exec rake sidekiq:queue_depth`
   Action: Checked error rate in Datadog.
   Verification: Queue-depth command returned 0; Datadog error rate was back to baseline. Exact Datadog query or view was not provided; needs confirmation.
   Source: [M08]

8. Re-enable enqueue.
   Who: Bryce Harmon.
   Command: `bin/rails runner 'FeatureFlag.enable(:auto_recognition_enqueue)'`
   Verification: 40 new jobs processed cleanly in the next 3 minutes.
   Rollback: The inverse command appears in [M04]: `bin/rails runner 'FeatureFlag.disable(:auto_recognition_enqueue)'`. Its use as rollback for this step was not explicitly stated; needs confirmation.
   Source: [M09], [M04]

9. Scale reward workers back to 3 replicas and resolve the incident.
   Who: Bryce Harmon.
   Command: `kubectl scale deployment/reward-worker --replicas=3`
   Action: Declared the incident resolved.
   Verification: Queue remained stable at 0. No replica-readiness check recorded; needs confirmation.
   Rollback: The prior scale-up command appears in [M06]: `kubectl scale deployment/reward-worker --replicas=6`. Its use as rollback for this step was not explicitly stated; needs confirmation. No rollback for incident resolution was provided.
   Source: [M10], [M06]
datadog-incident-timelinePASS 1.00
Plain-English question: Sequence the failure from the logs — what broke first, and what cascaded after it?
Prompt — what the model saw
From the log slice, sequence the failure: first error with timestamp, the cascade in order, the service and job involved, and the Datadog query that would confirm the first error. State what the logs do not show.
cost $0.0818 · 24s · in 34,017 / out 1,176 tok
✓ first_error_ts✓ service_named✓ cascade_ordered✓ query✓ gaps_stated

✓ Correct Answer (ground truth)

Plain-English answer: A reward-service Redis timeout at 14:01:12Z cascades to sidekiq at 14:01:40 and api-gateway at 14:03:05, with the Datadog query that confirms the first error.
The answer in one line: job_class_involved = RewardGiveJob
cascade_orderservice=reward-service; first_error_at=2026-09-03T14:01:12Z | service=sidekiq; first_error_at=2026-09-03T14:01:40Z | service=api-gateway; first_error_at=2026-09-03T14:03:05Z | service=web-app; first_error_at=2026-09-03T14:03:30Z
clean_servicespostgres
confirming_queryservice:reward-service status:error "@message:Redis::TimeoutError"
first_errortimestamp=2026-09-03T14:01:12Z; service=reward-service; message=Redis::TimeoutError: Connection to Redis at redis-primary:6379 timed out after 5s
job_class_involvedRewardGiveJob
logs_do_not_showroot cause of the Redis timeout, redis-server's own logs, infrastructure metrics (CPU/memory/network), any deploy or config change preceding the incident, who applied mitigation or when (no actor fields in logs)
graded checksfirst_error_ts · service_named · cascade_ordered · query · gaps_stated

Model's actual answer

show the model's full answer
First error: 2026-09-03T14:01:12Z, reward-service:
Redis::TimeoutError: Connection to Redis at redis-primary:6379 timed out after 5s

Cascade in observed order (all timestamps UTC, 2026-09-03):
1. 14:01:20–14:01:40: reward-service reports retries exhausted for RewardGiveJob.
2. 14:01:40: sidekiq reports RewardGiveJob failed with Redis::TimeoutError; retrying in 60s. Further failures/retries follow. In sidekiq_jobs.csv, the earliest recorded failure is J-00005 at 14:01:46.
3. 14:02:30: sidekiq warns that Queue reward depth is above 10,000.
4. 14:02:36: RecognitionDigestJob also fails with Redis::TimeoutError, recorded as J-00013 in sidekiq_jobs.csv.
5. 14:03:05: api-gateway reports a 502 upstream timeout calling reward-service /gives.
6. 14:03:30: web-app reports Give form submission failed because of an upstream 502 from api-gateway. Job failures, gateway timeouts, and form failures continue through 14:06:52.
7. 14:22:10: reward-service reports Redis connection restored and job processing resuming.
8. 14:24:45: sidekiq reports Queue reward depth below 500.

Service and job: The first error is in reward-service, connecting to redis-primary:6379. The primary job named in the cascade is RewardGiveJob, processed by sidekiq. RecognitionDigestJob is also affected.

Datadog query:
service:reward-service status:error "Redis::TimeoutError" "redis-primary:6379" "timed out after 5s"

Set the UTC time window to 2026-09-03T14:01:00Z–2026-09-03T14:02:00Z and sort oldest first to confirm the 14:01:12Z entry.

What the logs do not show:
- Why the Redis connection timed out; the underlying cause is not established.
- Correlation IDs proving that individual job, gateway, and form failures belong to the same request.
- The full outage duration, exact queue peak, or total affected jobs/users.
- Whether failed jobs ultimately succeeded, the queue fully drained, or /gives and form submissions recovered. Connection restoration and a queue below 500 do not establish those outcomes.
feature-flag-state-summaryPASS 1.00
Plain-English question: Summarize what each feature flag controls and who has it on.
Prompt — what the model saw
Summarize the state of the listed feature flags: what each controls per the code excerpt, which companies or segments have it on, targeting rules, and any flag with no code reference or no targets. Company counts must match the export. Do not describe a flag that is not in the export.
cost $0.0832 · 20s · in 12,697 / out 918 tok
✓ orphans✓ no_targets✓ counts_match✓ no_invented

✓ Correct Answer (ground truth)

Plain-English answer: Recognition streaks on for 42 beta companies, redeem-flow redesign off with a 12-company target list, plus any flag with no code reference or no targets.
The answer in one line: rule = company counts must equal the export; do not describe flags absent from the export.
code_referencedrecognition_streaks_v2, points_budget_guardrails, slack_dm_nudges, redeem_flow_redesign, analytics_dashboard_v3, ms_teams_app_v2
flags{"recognition_streaks_v2": {"state": "on", "targeting_rules": "segment:beta_companies", "company_count": 42}, "points_budget_guardrails": {"state": "on", "targeting_rules": "all_companies", "company_c
flags_with_no_targetspaused_offboard_cleanup
orphan_flags_no_code_referencelegacy_give_modal, survey_boosters_q3, paused_offboard_cleanup
rulecompany counts must equal the export; do not describe flags absent from the export.
graded checksorphans · no_targets · counts_match · no_invented

Model's actual answer

show the model's full answer
Company counts below are copied directly from the export, not calculated. For flags marked off, listed targets do not mean those companies have the feature on. Individual company names were not provided.

- recognition_streaks_v2 — On. Controls recording recognition streaks via StreakTracker.record(give). Targets segment:beta_companies; company_count: 42.
- points_budget_guardrails — On. Controls enforcing giver points budgets via BudgetService.new(company).enforce!(giver, points). Targets all_companies; company_count: 220.
- slack_dm_nudges — On. Gates SlackDm.send_nudge(user); the job returns without sending when disabled. Targets segment:region_na; company_count: 87.
- redeem_flow_redesign — Off. Selects RedeemV2Component when enabled and RedeemV1Component otherwise. Targets targeted_list; company_count: 12. List members were not provided.
- analytics_dashboard_v3 — On. Controls initializing @dashboard with AnalyticsV3.new(company). Targets segment:tier_three; company_count: 65.
- ms_teams_app_v2 — Off. Controls TeamsAppV2.install(company). Targets targeted_list; company_count: 9. List members were not provided.
- legacy_give_modal — Off. No code reference in the provided excerpt; what it controls is unknown. Targets segment:legacy_plan; company_count: 14.
- survey_boosters_q3 — On. No code reference in the provided excerpt; what it controls is unknown. Targets segment:legacy_plan; company_count: 7.
- paused_offboard_cleanup — Off. No code reference in the provided excerpt; what it controls is unknown. Targeting rules are blank; company_count: 0, so there are no exported targets.

Segment definitions were not provided. Although legacy_give_modal and survey_boosters_q3 both specify segment:legacy_plan, their exported company counts differ; the supplied data does not explain why.
nda-triagePASS 1.00
Plain-English question: Triage each NDA green, yellow, or red — and quote the clause that decides it.
Prompt — what the model saw
Triage each NDA as GREEN (standard approval), YELLOW (counsel review), or RED (full legal review). Cite the clause that drives each classification by quoting it. Catch embedded non-solicits, non-competes, missing carve-outs, and unusual term or governing law. One line of reasoning per NDA.
cost $0.0794 · 15s · in 12,395 / out 590 tok
✓ green✓ yellow✓ red✓ non_solicit_caught✓ clause_quoted✓ correct_labels

✓ Correct Answer (ground truth)

Plain-English answer: NDA-1 green (standard mutual terms), NDA-2 yellow for an embedded 18-month non-solicit, with the driving clause quoted for every classification.
The answer in one line: non_solicit_in = NDA-2
checksclassification_per_nda, clause_quotes_exist_in_document, non_solicit_caught, carveouts_checked
classifications{"NDA-1": {"classification": "GREEN", "driving_reason": "standard mutual NDA: 2-year term, standard carve-outs, Delaware law", "driving_clause_quote": "Carve-outs: information that (a) is or becomes p
non_solicit_inNDA-2
graded checksgreen · yellow · red · non_solicit_caught · clause_quoted · correct_labels

Model's actual answer

show the model's full answer
NDA-1 — [PARTY A] and Bonusly: GREEN (standard approval). “Term: two (2) years from the Effective Date; obligations survive for three (3) years after termination,” with all five confidentiality carve-outs, Delaware law, and “no exclusivity”; no restrictive covenant is stated.

NDA-2 — [PARTY B] and Bonusly: YELLOW (counsel review). Embedded non-solicit/no-hire: “neither party shall solicit for employment or hire any employee of the other party with whom it came into contact under this Agreement,” applying during the three-year term plus eighteen months afterward (3 years + 1.5 years = 4.5 years).

NDA-3 — [PARTY C] and Bonusly: RED (full legal review). Broad three-year non-compete: “Recipient shall not, directly or indirectly, engage in or provide services to any business that competes with Discloser's business”; compounded by “No confidentiality carve-outs are stated,” a five-year one-way term, and Irish law with “exclusive jurisdiction of its courts.”