Tasks / Growth

Diagnose a retention problem

Can the model work out from raw data what's really happening to retention, separating churn from measurement and seasonality, and find the cause?

Task type v1.1 · 2 tasksLast changed 6 Oct 2026 · ChangelogDifficulty

What AI gets right here, and what you’ll still have to catch

From 18 graded outputs by 9 models. 78% were usable with at most a quick edit.

Reliably right

Respects explicit constraints100% pass
The output is a diagnosis under 800 words, addressed to Leo, and proposes actions that would enforce the constraint of holding acquisition spend.
GPT-6.1 Sol · API · Flat, while sign-ups grow
Produces the required deliverable100% pass
Provides a complete diagnosis with evidence and recommendations, in memo form for Leo, within 800 words, and is actionable as-is.
GPT-6.1 Sol · API · Flat, while sign-ups grow
Understands before it justifies100% pass
Tests the summer explanation against the data, finds it insufficient, and reaches its own diagnosis from the cohort and accounting evidence.
GPT-6.1 Sol · API · Flat, while sign-ups grow

Where it slips

Uses the supplied evidence correctly67% pass
The SMB adoption churn calculation is incorrect; 39% of ~$7.4k/month churn is ~$2.9k/month, not $34k MRR, and the annualized figure is ~$34.8k ARR, not $0.4M.
Sonnet 5.5 · API · The collapse that wasn't
Avoids unsupported claims72% pass
The output states as fact that removing goal-setting 'stripped away the user’s commitment and personal investment' and caused the drop, without labelling this as a hypothesis.
Gemini 3.8 Flash · API · Flat, while sign-ups grow
Gets the base of every number right75% pass
The SMB adoption churn calculation is incorrect; 39% of ~$7.4k/month churn is ~$2.9k/month, not $34k MRR, and the annualized figure is ~$34.8k ARR, not $0.4M.
Sonnet 5.5 · API · The collapse that wasn't

How it’s graded

Each check is built on how world-class product and growth leaders do this work, as they’ve described it on Lenny’s Podcast, and names the leader it comes from.

Understands before it justifies
Does the diagnosis start from what the data shows (which users, when, at what point in their life with the product) before accepting or proposing an explanation, including the one it was handed?
Passes when Tests the explanation it was given against the data, and reaches its own diagnosis from the evidence.
Bangaly KabaLenny’s Podcast
Splits growth into new, lost and returning
Does the diagnosis separate new users or revenue from what was lost and what came back, rather than reading a single total?
Passes when Uses growth accounting (new, retained, churned, returning, or new, expansion, contraction, churn) to show which flow changed.
Naomi GleitLenny’s Podcast
Reads the shape of the curves
Does the diagnosis read how retention curves behave over time (flattening, still falling, or dipping and recovering) and what each shape means?
Passes when Describes the curve shapes in the data and draws the right conclusion from each, such as a dip that recovers being seasonal.
Product and growth leadersOn Lenny’s Podcast

Plus our standard checks

Uses the supplied evidence correctly · Addresses the actual decision · Respects explicit constraints · Identifies material uncertainty · Avoids unsupported claims · Produces the required deliverable · Gets the base of every number right · and 3 written for each task, which you’ll see in the tasks below.

The tasks

Read the brief, then compare up to three outputs, each with the LLM judge’s verdict on every check. Highlights mark what a PM had to fix.

The brief

You're a Staff PM at Tandem. Our CEO, Maya Lind, believes retention collapsed in March and wants two squads on an AI assistant next quarter (her note is below). She's asked you to confirm the diagnosis and find where else we're losing customers. Write her a memo of no more than 1,200 words: what is and isn't happening to retention, the evidence for each, the size of the real problem in revenue, and what to do about it. The files are attached, as exported.

What the model was given6 items: About Tandem, ceo_note.md, active_accounts_by_month.csv (from the main dashboard), mrr_movements_by_month.csv, exit_survey_2026.csv (accounts that cancelled), product_and_pricing_changes.md

About TandemA shared workspace for teams: docs, wikis and project pages. Sold per seat per month. Three segments: education (schools and universities, mostly annual contracts renewing in August), SMB, and mid-market.

ceo_note.md1 lines · Download

From Maya (CEO): "Retention collapsed in March and hasn't recovered: active accounts are down by a fifth and education is in freefall. Notion AI is eating us. I want two squads on an AI assistant next quarter. Can you confirm the diagnosis and tell me where else we're bleeding?"

active_accounts_by_month.csv (from the main dashboard)month,segment,paying_accounts,active_accounts 2026-01,Education,610,525 2026-01,SMB,2140,1758 2026-01,Mid-market,420,378 2026-02,Education,617,531 2026-02,SMB,2166,1785 2026-02,Mid-market,425,381 2026-03,Education,625,428 2026-03,SMB,2191,1434 2026-03,Mid-market,430,314 2026-04,Education,632,439 2026-04,SMB,2217,1447 2026-04,Mid-market,435,317 2026-05,Education,639,443 2026-05,SMB,2243,1448 2026-05,Mid-market,440,322 2026-06,Education,647,293 2026-06,SMB,2268,1476 2026-06,Mid-market,445,325 2026-07,Education,654,201 2026-07,SMB,2294,1497 2026-07,Mid-market,450,327 2026-08,Education,661,296 2026-08,SMB,2320,1514 2026-08,Mid-market,455,330 2026-09,Education,669,461 2026-09,SMB,2345,1524 2026-09,Mid-market,460,335

mrr_movements_by_month.csv27 rows · Download

month,segment,starting_mrr_usd,new_mrr_usd,expansion_mrr_usd,contraction_mrr_usd,churned_mrr_usd,ending_mrr_usd
2026-01,Education,61000,2237,603,311,874,62655
2026-01,SMB,214000,9047,3152,1616,7154,217429
2026-01,Mid-market,268000,7253,8246,2374,2613,278512
2026-02,Education,62655,2092,584,300,934,64097
2026-02,SMB,217429,9117,3227,1592,7555,220626
2026-02,Mid-market,278512,6797,8384,2324,2769,288600
…

exit_survey_2026.csv (accounts that cancelled)response_id,segment,main_reason X501,Mid-market,Too expensive per seat after the price change X502,Mid-market,Too expensive per seat after the price change X503,Mid-market,Too expensive per seat after the price change X504,Mid-market,Too expensive per seat after the price change X505,Mid-market,Too expensive per seat after the price change X506,Mid-market,Too expensive per seat after the price change X507,Mid-market,Too expensive per seat after the price change X508,Mid-market,Too expensive per seat after the price change X509,Mid-market,Too expensive per seat after the price change X510,Mid-market,Too expensive per seat after the price change X511,Mid-market,Too expensive per seat after the price change X512,Mid-market,Consolidating tools X513,Mid-market,Consolidating tools X514,Mid-market,Consolidating tools X515,Mid-market,Switched to Notion X516,SMB,Company closed or downsized X517,SMB,Company closed or downsized X518,SMB,Company closed or downsized X519,SMB,Company closed or downsized X520,SMB,Company closed or downsized X521,SMB,Company closed or downsized X522,SMB,Company closed or downsized X523,SMB,Company closed or downsized X524,SMB,Company closed or downsized X525,SMB,Company closed or downsized X526,SMB,Company closed or downsized X527,SMB,Company closed or downsized X528,SMB,Company closed or downsized X529,SMB,Company closed or downsized X530,SMB,Too expensive X531,SMB,Too expensive X532,SMB,Too expensive X533,SMB,Too expensive X534,SMB,Too expensive X535,SMB,Too expensive X536,SMB,Too expensive X537,SMB,Too expensive X538,SMB,Too expensive X539,SMB,Switched to Notion X540,SMB,Switched to Notion X541,SMB,Switched to Notion X542,SMB,Switched to Notion X543,SMB,Didn't get the team to adopt it X544,SMB,Didn't get the team to adopt it X545,SMB,Didn't get the team to adopt it X546,SMB,Didn't get the team to adopt it X547,SMB,Didn't get the team to adopt it X548,SMB,Didn't get the team to adopt it X549,SMB,Didn't get the team to adopt it X550,SMB,Didn't get the team to adopt it X551,SMB,Didn't get the team to adopt it X552,SMB,Didn't get the team to adopt it X553,SMB,Didn't get the team to adopt it X554,SMB,Didn't get the team to adopt it X555,SMB,Didn't get the team to adopt it X556,SMB,Didn't get the team to adopt it X557,SMB,Didn't get the team to adopt it X558,SMB,Didn't get the team to adopt it X559,SMB,Didn't get the team to adopt it X560,Education,"End of academic year, will review" X561,Education,"End of academic year, will review" X562,Education,"End of academic year, will review" X563,Education,"End of academic year, will review" X564,Education,"End of academic year, will review" X565,Education,"End of academic year, will review" X566,Education,"End of academic year, will review" X567,Education,"End of academic year, will review" X568,Education,"End of academic year, will review" X569,Education,"End of academic year, will review" X570,Education,"End of academic year, will review" X571,Education,"End of academic year, will review" X572,Education,Budget cut X573,Education,Budget cut X574,Education,Budget cut X575,Education,Budget cut X576,Education,Budget cut X577,Education,Budget cut X578,Education,Switched to Google Workspace X579,Education,Switched to Google Workspace

product_and_pricing_changes.md6 lines · Download

# Product and pricing changes, 2026

- 2 March: Dashboard metric "active account" changed from "anyone logged in within 30 days" to "anyone created or edited a page within 30 days", to measure real use.
- 15 May: Mid-market plan from $12 to $15 per seat per month, at each account's next monthly bill.
- 15 May: Seat audit launched for admins: a page listing seats with no activity in 60 days, with a one-click "remove seat".
- 3 August: Notion launched its AI assistant on every paid plan.

What a strong answer doesThe answer key the graders mark against

Separates what isn't churn from what is. (1) March: active accounts fall about 17 points in every segment at once while paying accounts keep growing, because on 2 March 'active' changed from logging in to editing: a definition change, not a collapse. (2) Education: active accounts dip June to August and recover in September, the summer break; education MRR grows, with a modest churn bump at August renewals. (3) Logo churn is roughly flat throughout, and Notion is the main reason in only 5 of 79 exit responses, and Notion's AI launched in August, months after the 'collapse'. The real loss is mid-market contraction: after the 15 May rise from $12 to $15 per seat, shipped the same day as the seat audit, contraction jumps from about $2,400 a month to $12,000 to $22,000 a month (about $75,000 in May to September against about $9,600 in January to April), and mid-market MRR stops growing (flat at about $322,000 from June after rising about $10,000 a month), while 11 of 15 mid-market exits cite price per seat. A secondary issue is SMB adoption ('didn't get the team to adopt it' is the top SMB reason). Recommends fixing the mid-market seat economics (such as pricing active seats or a seat-band discount, and rethinking how the seat audit is presented), restating the dashboard on a consistent definition, and not putting two squads on an AI assistant as a retention fix without evidence it would help.

Critical failures (cap the score at 40)

Confirms that retention collapsed in March

Task

Staff-level · v1.1

The outputs

9 setups answered this brief.

GPT-6 AstraChatGPT100Usable with a quick edit Haiku 5.5API100Usable with a quick edit Opus 5.5Claude96Usable with a quick edit GPT-6 LunaAPI96Usable with a quick edit GPT-6.1 SolAPI93Usable with a quick edit Sonnet 5.5API79Needs more work Gemini 3.5 Flash-LiteGemini70Needs more work Gemini 4 ArgonAPI61Needs more work Gemini 3.8 FlashAPI59Needs more work

Gemini 4 ArgonAPI

Needs more than a quick edit8 of 13 checks passed

The memo contains a couple of minor arithmetic mistakes (SMB active ratio 65.4% vs. 65.5%, education churn baseline ~$906) that would require a double-check; otherwise it is well-structured and actionable.

Output

MEMORANDUM

TO: Maya Lind, CEO FROM: Staff Product Manager DATE: October 2026 SUBJECT: Retention Diagnosis, Revenue Impact, and Q4 Squad Allocation

Executive Summary

Retention did not collapse in March, Education is not in freefall, and Notion AI is not driving customer losses.

The apparent ~20% drop in active accounts in March was a reporting artifact caused by a stricter analytics definition launched on March 2. Furthermore, the summer dip in Education usage is normal academic seasonality, which fully rebounded in September. We should not allocate two squads to an AI assistant next quarter.

However, we are bleeding significant revenue in two other areas: 1. Mid-Market Contraction ($62,800/mo in excess contraction since May): On May 15, we simultaneously raised Mid-market prices by 25% ($12 to $15/seat) and launched a "one-click remove seat" audit tool for admins. This prompted admins to aggressively purge read-only seats, wiping out the gains from the price hike and stalling Mid-market growth. 2. SMB Gross Churn (~$7,400/mo in lost MRR): SMB is our largest source of full account cancellations, driven primarily by poor team onboarding and macro business closures.

Instead of building an AI assistant, we should redirect those two squads toward Mid-market pricing/packaging (introducing free "Viewer" seats) and SMB onboarding and activation.

---

1. What Isn’t Happening (Evaluating the Initial Diagnosis)

The March "Collapse" Was a Metric Definition Change * The Evidence: On March 2, we changed the dashboard definition of an "active account" from anyone logging in within 30 days to anyone creating or editing a page within 30 days. Immediately, reported active-to-paying ratios dropped by roughly 20% across all segments simultaneously: * Education: 86.1% (Feb) $\rightarrow$ 68.5% (Mar) * SMB: 82.4% (Feb) $\rightarrow$ 65.4% (Mar) * Mid-market: 89.6% (Feb) $\rightarrow$ 73.0% (Mar) * The Reality: Underlying retention remained completely stable. Paying accounts grew steadily in March (from 3,208 to 3,246 total), and total Churned MRR was flat ($11,258 in Feb vs. $11,403 in Mar). Under the new definition, engagement rates have remained completely stable from March through September (~65% in SMB, ~73% in Mid-market).

Education is Experiencing Normal Seasonality, Not Freefall * The Evidence: Education active accounts dropped to 293 in June and 201 in July, before rebounding to 461 in September (68.9% of paying accounts, matching March's 68.5% rate). Meanwhile, paying Education accounts grew every single month from January (610) to September (669). * The Reality: Schools and universities go on break over the summer. Because most Education contracts are annual and renew in August, we saw a predictable annual churn spike in August ($3,925 vs. a ~$906/mo baseline), with exit surveys citing "End of academic year, will review" (12/20 responses) and "Budget cut" (6/20). By September, Education MRR hit an all-time high of $72,211.

Notion AI is Not Eating Our Customer Base * The Evidence: Notion launched its AI assistant on August 3. However, our total Churned MRR across SMB and Mid-market actually decreased slightly after the launch ($10,386 in July $\rightarrow$ $10,189 in Aug $\rightarrow$ $10,173 in Sep). In our 2026 exit surveys (79 total responses), 0 customers mentioned AI, and only 5 (6.3%) switched to Notion (4 SMB, 1 Mid-market).

---

2. Where We Are Actually Bleeding (and Size of the Problem)

While logo churn is stable, revenue growth has stalled due to two distinct problems in Mid-market and SMB.

#### Problem A: Mid-Market Seat Contraction (The May 15 Self-Inflicted Wound) On May 15, we increased Mid-market pricing from $12 to $15/seat (+25%) and simultaneously launched the Admin Seat Audit page, which flags seats with no activity in 60 days and offers a one-click "remove seat" button.

The Dynamics: The price hike generated a one-time $31,585 expansion spike in May (~$23,200 above our ~$8,360/mo baseline). However, because we don't differentiate between "Editors" and "Viewers," the price increase gave admins sticker shock, and the audit tool gave them an immediate mechanism to cut costs by purging passive readers.
The Evidence: Mid-market Contraction MRR averaged $2,397/mo from Jan–Apr. Following the May 15 changes, it jumped to $10,239 in May, peaked at $22,351 in June, and remains elevated at $13,212 in September (over 5x the pre-May baseline). In exit surveys, 11 of 15 (73%) Mid-market cancellations cited "Too expensive per seat after the price change."
Revenue Impact:
From May to September (5 months), total Mid-market contraction was $74,802—representing ~$62,816 in excess contraction above our pre-May baseline.
This completely wiped out the ~$23.2k May price-hike expansion, resulting in a net revenue loss of ~$39,600 from the May 15 initiatives.
Mid-market MRR is now actively shrinking (falling from $334,910 in May to $321,932 in September), and we are still bleeding ~$10,800/mo in excess contraction ($129,600/yr annualized run-rate).

#### Problem B: SMB Churn & The Activation Gap SMB is our highest-churning segment in both percentage and absolute dollar terms.

The Evidence: SMB Churned MRR averages $7,444/mo (~$89,300/yr in lost ARR, or ~3.3% monthly gross MRR churn), accounting for 65% of all company Churned MRR.
The Drivers: Out of 44 SMB exit survey responses:
39% (17/44) left because they "Didn't get the team to adopt it."
32% (14/44) left because the "Company closed or downsized" (uncontrollable macro).
20% (9/44) cited "Too expensive."
Further, under our new active metric, 35% of paying SMB accounts (~821 accounts in September) did not create or edit a single page in the last 30 days. These dormant accounts are prime candidates for future churn.

---

3. Recommendations: What to Do Next Quarter

Instead of putting two squads on an AI assistant, we should deploy them against the two problems costing us nearly $220,000/yr in combined annualized MRR losses:

#### Squad 1: Mid-Market Packaging, Pricing & Admin Flows Goal: Stop the $10.8k/mo excess contraction bleed by aligning our pricing model with how teams actually use docs and wikis. 1. Introduce Free "Viewer / Commenter" Roles: Wikis and project pages require broad read access across a company, even if only a core group edits them. Charging $15/mo for passive readers forces admins to remove them, which shrinks our footprint inside the account. Make read/comment-only seats free, and charge $15/mo strictly for Editors/Creators. 2. Redesign the Admin Seat Audit Flow: Replace the destructive "one-click remove seat" button with a "Downgrade to Free Viewer" default action. Additionally, ensure the 60-day inactivity check tracks page views rather than just edits, so admins don't accidentally purge executives and cross-functional stakeholders who rely on Tandem to read documentation. 3. Win-Back & Annual Lock-In: Have Customer Success reach out to Mid-market accounts that contracted heavily since May, offering to restore removed employees as free Viewers and offering a $12/Editor rate in exchange for an annual contract commitment.

#### Squad 2: SMB Onboarding, Activation & Health Intervention Goal: Reduce the $7.4k/mo SMB churn rate by solving the #1 controllable churn driver ("Didn't get the team to adopt it") and activating the 35% of paying SMBs that are currently read-only/dormant. 1. Template-Driven Multiplayer Onboarding: New SMB workspaces fail when the admin faces a blank page. Build guided setup flows with pre-built templates (e.g., Company Wiki, Weekly Sprint Tracker) that prompt the admin to invite teammates and assign them specific sections to edit during the first 7 days. 2. Automated Zero-Edit Playbooks: Trigger automated lifecycle nudges and in-app guidance for paying SMB accounts that hit 14 days without creating or editing a page, helping admins migrate existing docs from Google Docs/Notion before renewal. 3. Pause/Downsize Flows: For the 32% of SMBs churning due to downsizing or closures, introduce a self-serve "Pause Subscription" or "Startup Relief" discount tier in the cancellation flow to retain logos through tough macro periods.

Claims the judge couldn’t find in the brief

Active-to-paying ratios dropped: Education 86.1% → 68.5%, SMB 82.4% → 65.4%, Mid-market 89.6% → 73.0%.
Education contracts are annual and renew in August; saw an annual churn spike in August of $3,925 vs. a ~$906/mo baseline.

Grades and run details

Decision model 62 · LLM judge 64

Decision model checks

failUses the supplied evidence correctly31%
failAddresses the actual decision89%
passRespects explicit constraints43%
failIdentifies material uncertainty95%
failAvoids unsupported claims22%
passProduces the required deliverable86%
passSeparates what isn't churn100%
passFinds and sizes the real loss98%
passAnswers the AI plan with evidence100%
passUnderstands before it justifies100%
failGets the base of every number right22%
passSplits growth into new, lost and returning88%
passReads the shape of the curves100%

Run

Run: #1
API response time: 22 s
Submitted: 13 Oct 2026

Check by check

Open a check to read the judge’s reasons. “Mixed” means the two graders disagreed.

Got wrong 4

Addresses the actual decisionWrong

The output commits to a clear call (no AI squads, redirect) but does not state what result or condition would change that recommendation.

Identifies material uncertaintyWrong

The memo does not name any unknowns that could change the decision or say how they would be resolved.

Avoids unsupported claimsWrong

The causal claim that the seat audit 'prompted admins to aggressively purge read-only seats' is presented as established fact without direct evidence beyond timing.

Gets the base of every number rightWrong

The SMB active-to-paying ratio for March is given as 65.4% instead of the correct 65.5%; the Education churn baseline is ~$906 not the computed ~$912, making figures slightly off.

Mixed 1

Uses the supplied evidence correctlyMixed

All key facts are taken correctly from the supplied context; minor arithmetic rounding issues do not invent new facts.

Got right 8

Respects explicit constraintsRight

The memo is well under 1,200 words, addressed to CEO Maya Lind, and covers the required topics.

Produces the required deliverableRight

The output is a complete, actionable memo that the CEO could use with light edits.

Separates what isn't churnRight

Clearly links the March drop to the 2 March definition change and shows education's summer dip recovering in September.

Finds and sizes the real lossRight

Identifies the mid-market contraction spike after the 15 May price rise and seat audit, and sizes it in MRR with specific numbers.

Answers the AI plan with evidenceRight

Uses exit surveys and the August Notion launch timeline to show AI is not behind the losses and advises against the AI squad plan.

Understands before it justifiesRight

Tests the CEO's diagnosis against the data and reaches its own conclusions from the evidence.

Splits growth into new, lost and returningRight

Uses new, expansion, contraction and churn MRR to show which flows changed (e.g., expansion spike, contraction jump).

Reads the shape of the curvesRight

Describes the dip-and-recover shape of education and the stable ratios after the definition change.

Results

Every setup we’ve tested on this task type, across all its tasks and repeats, graded on the current checklist. Calibrated.

#	Model · Harness	Decision model	LLM judge	Runs	Critical failures
1	GPT-6 AstrawithChatGPT	100.0	100.0	2	None
2	Haiku 5.5withAPI	100.0	100.0	2	None
3	GPT-6 LunawithAPI	96.2	100.0	2	None
4	GPT-6.1 SolwithAPI	92.3	100.0	2	None
5	Opus 5.5withClaude	85.7	100.0	2	None
6	Sonnet 5.5withAPI	92.3	89.3	2	None
7	Gemini 3.5 Flash-LitewithGemini	80.3	74.2	2	None
8	Gemini 4 ArgonwithAPI	70.4	82.1	2	None
9	Gemini 3.8 FlashwithAPI	76.4	70.9	2	None

What you get for the cost

GPT-6 Luna · API scores within 2 points of GPT-6 Astra · ChatGPT at about a fiftieth of the cost.

Best valueNobody beats these on both score and cost

Task score and relative cost by setup
Setup	Task score	Relative cost (cheapest = 1)
GPT-6 Luna · API	98	1.0×
Gemini 3.5 Flash-Lite · Gemini	76	3.0×
Haiku 5.5 · API	100	3.6×
Gemini 3.8 Flash · API	73	13×
GPT-6.1 Sol · API	96	15×
Gemini 4 Argon · API	76	17×
Sonnet 5.5 · API	89	50×
Opus 5.5 · Claude	93	77×
GPT-6 Astra · ChatGPT	100	79×

Cost is each output’s tokens at the model’s list price, shown relative to the cheapest setup. Runs done in an app have no usage meter, so theirs are estimated from the same setup’s measured runs. How we compare cost

The PM job

Working out why customers leave, and what to fix first.

Why it matters

Teams pick a story (the season, a competitor) and then pull data to justify it. Half the time the 'drop' is a definition change, and the real leak is somewhere else.

What good looks like

Starts from what the data shows, not a story
Splits growth into new, lost and returning users
Separates measurement and seasonality from churn
Sizes the real problem and fixes new-user retention first

Deliberately not measured

Dashboard design

Capability tested

Retention diagnosis

The failure we’re looking for

Confirming the story everyone already believes

Grading

Decision model and LLM judge, calibrated against a blind PM review