Tasks / Metrics & Experimentation

Pick a North Star metric

Can the model pick the one metric that measures the value a product creates, and break it into inputs teams can move?

Task type v1.1 · 2 tasksLast changed 5 Oct 2026 · ChangelogDifficulty

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

From 16 graded outputs by 8 models. 75% were usable with at most a quick edit.

Reliably right

Says how it could be gamed100% pass
Identifies gaming vectors (tapping through, notification overuse, short-term spikes, revenue manipulation, MAU drift) and sets specific guardrails (quality completion, push caps, blocking metrics, cohort retention checks, subscriber churn tracking).
Sonnet 5.5 · API · Revenue, actives, or something else?
Sees the growth is narrow100% pass
It shows the split (chains up, independents' filled shifts and fill rate down, more leaving) and tracks independent venues via segment reporting and guardrails.
GPT-6 Luna · API · Growing, on paper
Checks the North Star can be measured100% pass
It flags the 64% confirmation gap and proposes instrumenting check-in/out, reconciliation, and auditing to fix it before the board readout.
GPT-6 Luna · API · Growing, on paper

Where it slips

Identifies material uncertainty55% pass
It identifies the confirmation gap as an unknown but does not say how resolving it could change the North Star decision.
GPT-6.1 Sol · API · Growing, on paper
Addresses the actual decision58% pass
The output commits to a North Star but does not state what result or condition would change that recommendation.
GPT-6 Luna · API · Growing, on paper
Avoids unsupported claims67% pass
It presents 'easy to tap through' and 'we don't track today' as facts without support, and does not label them as assumptions.
Opus 5.5 · Claude · Revenue, actives, or something else?

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.

Measures value created, not captured
Does the chosen North Star measure the value customers get from the product, rather than revenue or a count of people who merely showed up?
Passes when Picks a metric of value delivered to customers (a core action that predicts retention), and keeps revenue as a separate business measure.
Itamar GiladLenny’s Podcast Sean EllisLenny’s Podcast
Breaks it into inputs teams can move
Does the output break the North Star into a few input metrics, each one a team could own and move week to week, and pair it with a separate business measure of value captured?
Passes when Gives a small tree of input metrics, each with an owner and a link to the North Star, plus the revenue or business measure it's paired with.
Itamar GiladLenny’s Podcast
Says how it could be gamed
Does the output pressure-test the North Star for how teams could raise it without delivering more value, and set guardrail metrics against that?
Passes when Names a realistic way the metric could be gamed in this product and the guardrails that would catch it.
Lauryn IsfordLenny’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 4 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 Crewmate. Every team has been steering by its own number, and the exec team wants to pick one North Star at next week's offsite, so the CEO can report it to the board three weeks later. Write the memo they'll read beforehand, in no more than 1,200 words: the North Star you recommend and why, its metric tree with an owner for each input, the business measure it's paired with, how it could be gamed and the guardrails against that, and what each team's current metric becomes. The numbers below are as the data team sent them. Not all of it matters equally.

What the model was given9 items: About Crewmate, What each exec wants, Shifts by quarter (monthly averages), GMV (monthly average), Venues, Workers, Same-day shifts, Completion data, From the COO

About CrewmateA marketplace for hospitality shifts: venues post shifts, and vetted workers pick them up. Crewmate takes 18% of each shift's value. Two kinds of venue: four enterprise hotel and restaurant chains (310 venues between them, on annual contracts) and 2,790 independent venues (bars, restaurants, cafes) who pay as they go.

What each exec wantsCEO: GMV, 'because it's what the board looks at'. Head of Demand: shifts posted. Head of Supply: active workers. COO: fill rate.

Shifts by quarter (monthly averages)Enterprise chains: Q2 posted 21,000, filled 19,500. Q3 posted 34,000, filled 31,600. Independent venues: Q2 posted 56,000, filled 44,200. Q3 posted 58,000, filled 39,400.

GMV (monthly average)Q2 $8.5M, Q3 $9.8M. Enterprise chains were 34% of GMV in Q2 and 46% in Q3. The chains raised hourly rates by 11% for the holiday season from August.

VenuesVenues whose shifts fill 90% of the time or more: 84% still posting six months later. Under 70%: 41%. In Q3, 9% of independent venues stopped posting altogether (5% in Q2). No chain venue stopped.

WorkersRegistrations: 6,100 in Q2, 7,900 in Q3, after Supply introduced a $25 sign-up bonus in July. 2,370 of the Q3 registrants have worked a shift. Workers who work three or more shifts in their first month: 66% still working after three months; everyone else: 12%.

Same-day shiftsShifts posted less than 24 hours before they start fill 58% of the time. They are 31% of independents' posts and 8% of the chains'.

Completion dataVenues confirm in the app that a shift was worked for 64% of filled shifts; the rest are assumed worked unless the venue raises a dispute. Of confirmed shifts, 6% were no-shows or disputed.

From the COO"Since September, Ops pays workers a $15 rescue bonus to pick up same-day shifts that would otherwise go unfilled. 6,200 rescue fills last month, $93,000. Rescue-filled shifts have a 13% no-show rate, against 4% for the rest. It's kept our fill rate from falling further."

What a strong answer doesThe answer key the graders mark against

Works out from the raw numbers what the pack doesn't say: all of the growth is the four chains (filled shifts up from 19,500 to 31,600 a month, plus an 11% rate rise), while independent venues filled fewer shifts (44,200 to 39,400, so their fill rate fell from 79% to 68%) and are leaving faster (9% stopped posting, against 5%); GMV rose mostly on the chains' volume and prices. Picks a North Star of value delivered to both sides, such as shifts worked, but designs it so the decline can't hide: measured for independents separately, or as reliably staffed venues, or with an explicit independent-venue guardrail. Notices that 'worked' is only confirmed for 64% of shifts, so the North Star can't be trusted yet, and makes fixing confirmation an early input or uses a measurable proxy until then. Treats the rescue bonus as a way the fill rate (and any fills-based North Star) is being bought: 13% no-shows against 4%, so it needs guardrails such as cost per worked shift and no-show rate. Shows how shifts posted, registrations (2,370 of 7,900 worked) and GMV each rose without more value. Builds a metric tree with owners, pairs it with revenue or GMV as the business measure, and redefines each team's metric as an input.

Critical failures (cap the score at 40)

Recommends GMV or shifts posted as the North Star

Task

Staff-level · v1.0

The outputs

8 setups answered this brief.

GPT-6 LunaAPI97Usable with a quick edit Sonnet 5.5API95Usable with a quick edit GPT-6.1 SolAPI93Usable with a quick edit GPT-6 AstraChatGPT93Usable with a quick edit Opus 5.5Claude88Usable with a quick edit Haiku 5.5API82Usable with a quick edit Gemini 3.8 FlashAPI73Usable with a quick edit Gemini 3.5 Flash-LiteGemini62Needs more work

GPT-6 AstraChatGPT

Usable with a quick edit12 of 14 checks passed

The memo is complete and actionable; a PM could add a sentence about what would change the North Star choice as a quick edit, without rechecking evidence or redoing reasoning.

Output

Decision memo: Crewmate’s North Star

To: Executive team From: Staff PM Decision for next week’s offsite: Adopt successful shifts worked per month as our North Star, paired with marketplace contribution dollars. Publish both with mandatory enterprise/independent breakdowns.

Why this, not today’s competing metrics?

Crewmate creates value when a venue gets the staffing it requested and a worker completes paid work—not when someone posts, registers, or accepts a shift.

Our headline growth currently masks deteriorating service:

Monthly average	Q2	Q3
---	---:	---:
Total filled shifts	63,700	71,000
Overall fill rate	82.7%	77.2%
Enterprise fill rate	92.9%	92.9%
Independent fill rate	78.9%	67.9%

GMV rose 15.3%, but independent GMV fell approximately 5.7%, from $5.61M to $5.29M. Enterprise growth and its 11% seasonal hourly-rate increase obscure that deterioration. Four contracted chains now produce 46% of GMV; none stopped posting, while independent cessation rose from 5% to 9%.

This matters beyond one quarter. Venues with fill rates of at least 90% have 84% six-month posting retention, versus 41% below 70%. That is an association, not proof of causation, but it makes independent reliability our clearest warning.

Neither fill rate nor filled-shift volume is sufficient. Fill rate can improve by discouraging difficult demand. Filled volume can grow through unreliable, subsidized matches: rescue fills have a 13% no-show rate versus 4% otherwise.

Successful shifts worked measures delivered marketplace value. It rewards demand growth, useful supply, and reliable execution together. Segment guardrails prevent enterprise growth from concealing independent failure.

Definition and metric tree

A successful shift is one genuine venue staffing request that a worker actually works and for which payment is owed, with no upheld no-show or material service dispute. Acceptance alone does not count. Rescue reassignment does not create another shift.

Count by shift date, using a seven-day reporting lag; restate later adjudications transparently.

For each venue segment × local market × posting-lead-time cohort:

> Successful shifts = genuine shifts posted × fill rate × successful-work rate among filled shifts

Sum cohort results for the North Star. Keep same-day requests in the denominator.

Input	Accountable owner	Operating levers and diagnostics
Genuine shifts posted	Head of Demand	Posting venues × shifts per venue; acquisition, repeat posting, advance notice, sustainable shift pricing
Fill rate	Head of Supply	Available, vetted workers in the right place/time; acceptance; matching; worker activation and retention
Successful-work rate	COO	Attendance, cancellations, rescue reliability, dispute resolution
Completion measurement coverage and accuracy	Head of Data, with Product delivering instrumentation	Evidence capture, representative audits, reconciliations, uncertainty reporting

Supply’s key activation diagnostic becomes the share of new workers completing three shifts within their first 30 days, measured only for mature cohorts. These workers have much stronger observed retention: 66% versus 12%. Treat that as a hypothesis to test, not permission to manufacture three token shifts.

Measurement: we do not yet know the North Star

Only 64% of filled shifts have venue-reported outcomes; 6% of those were no-shows or disputed. The other 36% are assumed worked. We cannot assume confirmed and unconfirmed shifts have identical outcomes—or that every dispute ultimately represents failure.

Therefore, 71,000 Q3 filled shifts is not a completed-work baseline.

Before the board meeting, Data and Ops will:

Reconcile outcome evidence, payouts, cancellations, and dispute status.
Audit a representative sample of unconfirmed shifts, stratified by segment, lead time, and rescue status.
Publish successful-shift estimates with uncertainty, known outcomes, and unresolved coverage. If the sample cannot support a defensible estimate, publish bounds rather than false precision.
Apply the same method to any historical comparison. Better confirmation coverage must not masquerade as growth.

Product should make outcome capture routine through venue confirmation and worker attendance evidence, with independent checks where those conflict.

Paired business measure

Use monthly marketplace contribution dollars:

> Net transaction revenue, after refunds and credits, minus worker/venue incentives and variable payment, verification, and servicing costs.

Finance owns this measure. Keep fixed overhead separate and define cost allocation consistently.

At an 18% take rate, Q3 GMV implies $1.764M in monthly gross fees before adjustments and costs. The latest month’s rescue bonuses alone cost $93,000. That is not necessarily waste—but rescue acceptance is not proof of incremental, profitable completion.

Continue reporting GMV to the board, with a bridge separating completed volume, shift hours/rates, and enterprise mix. GMV remains financially useful; it stops being our definition of customer value.

Gaming risks and guardrails

Risk	Guardrail
Enterprise growth hides independent deterioration	Always show North Star and contribution by segment. Independent fill below 70% is a red escalation threshold—already breached—not an acceptable equilibrium. Track venue-level reliability distribution and posting cessation, not just averages.
Teams reject difficult shifts to improve rates	Freeze demand definitions; retain legitimate failed requests in denominators. Report cancellations, rejected demand, and same-day performance.
Ops buys nominal fills	Report rescue cost per successful shift and contribution by cohort. Test rescue incrementality against a credible control; scale only with demonstrated completion benefit and explicit subsidy limits.
Teams split shifts or generate token work	Count original staffing requests; deduplicate replacements and artificial splits. Monitor worked hours and value per shift.
Teams pressure confirmations or hide disputes	Audit both confirmed and unconfirmed work; prohibit acceptance or silence as completion evidence. Track no-shows, disputes, unpaid work, and safety incidents separately.

Demand and Supply should jointly test earlier posting and better same-day coverage: same-day shifts fill only 58% of the time and constitute 31% of independent posts versus 8% of enterprise posts. Do not simply stop serving them.

What each current metric becomes

CEO—GMV: Board financial context and revenue bridge, not North Star.
Demand—shifts posted: Demand input, evaluated alongside repeat posting and independent reliability.
Supply—active workers: Capacity diagnostic, segmented by availability and market; prioritize productive activation and retained workers over registrations. Only 2,370 of 7,900 Q3 registrants have worked a shift.
COO—fill rate: Intermediate conversion metric, subordinate to successful work, quality, and contribution.

Approve the definition, owners, guardrails, and measurement plan at the offsite. Set the numerical North Star growth target only after establishing a trustworthy baseline.

Grades and run details

Decision model 93 · LLM judge 93

Decision model checks

passUses the supplied evidence correctly12%
passAddresses the actual decision86%
passRespects explicit constraints43%
passIdentifies material uncertainty99%
passAvoids unsupported claims35%
passProduces the required deliverable74%
passSees the growth is narrow100%
passChecks the North Star can be measured100%
passSees the fill rate being bought100%
passOne North Star for both sides100%
passMeasures value created, not captured99%
failGets the base of every number right16%
passBreaks it into inputs teams can move100%
passSays how it could be gamed100%

Run

Run: #1
Time to output: 61 s
Submitted: 5 Oct 2026

Check by check

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

Mixed 2

Addresses the actual decisionMixed

The memo commits to a North Star but does not state what result or condition would change that choice, as required by the criterion.

Gets the base of every number rightMixed

All derived percentages and differences (fill rates, GMV growth, independent GMV drop) are computed correctly from the supplied data, with bases clear.

Got right 12

Uses the supplied evidence correctlyRight

All factual claims are taken directly from the supplied context or derived by correct arithmetic, with no invented facts.

Respects explicit constraintsRight

The output is a memo under 1,200 words, includes the recommended North Star, metric tree with owners, paired business measure, gaming risks and guardrails, and redefines each team's current metric.

Identifies material uncertaintyRight

It identifies the confirmation gap as a major unknown, proposes a measurement plan with uncertainty bounds, and notes the fill-rate/retention link is an association, not proven causation.

Avoids unsupported claimsRight

Interpretations like 'headline growth masks deterioration' are supported by the numbers; causal claims are labelled as hypotheses or associations.

Produces the required deliverableRight

The memo is complete, in the right form for the exec team, within the word limit, and actionable with all requested sections.

Sees the growth is narrowRight

It explicitly splits enterprise vs. independent performance, shows independent filled shifts and fill rate falling, and mandates segment breakdowns with an independent-fill guardrail.

Checks the North Star can be measuredRight

It flags that only 64% of filled shifts are confirmed, proposes a reconciliation and audit plan, and requires outcome capture improvements before the North Star can be trusted.

Sees the fill rate being boughtRight

It uses the 13% no-show rate vs. 4% and the $93k cost to show rescue fills inflate the number, and sets guardrails like cost per successful shift and incrementality testing.

One North Star for both sidesRight

The North Star 'successful shifts worked' only rises when a venue gets staffing and a worker completes paid work, explicitly valuing both sides of the marketplace.

Measures value created, not capturedRight

It measures value delivered to customers (completed work) rather than revenue or mere participation, and keeps revenue as a separate business measure.

Breaks it into inputs teams can moveRight

It breaks the North Star into genuine shifts posted, fill rate, successful-work rate, and completion measurement, each with an accountable owner, and pairs it with marketplace contribution dollars.

Says how it could be gamedRight

It lists realistic gaming risks (segment hiding, rejecting difficult demand, buying fills, splitting shifts, pressuring confirmations) and specific guardrails for each.

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	Sonnet 5.5withAPI	95.9	96.7	2	None
2	GPT-6.1 SolwithAPI	97.7	93.3	2	None
3	GPT-6 LunawithAPI	95.5	88.3	2	None
4	GPT-6 AstrawithChatGPT	94.2	88.3	2	None
5	Opus 5.5withClaude	94.6	80.8	2	None
6	Haiku 5.5withAPI	88.8	81.7	2	None
7	Gemini 3.8 FlashwithAPI	78.9	65.8	2	None
8	Gemini 3.5 Flash-LitewithGemini	54.4	38.3	2	1 capped

What you get for the cost

GPT-6 Luna · API scores within 4 points of Sonnet 5.5 · API at about a twentieth of the cost.

Best valueNobody beats these on both score and costAt least one critical failure

Task score and relative cost by setup
Setup	Task score	Relative cost (cheapest = 1)
GPT-6 Luna · API	92	1.0×
Haiku 5.5 · API	86	1.1×
Gemini 3.5 Flash-Lite · Gemini	44	1.8×
Gemini 3.8 Flash · API	71	7.5×
GPT-6.1 Sol · API	96	9.6×
Sonnet 5.5 · API	96	22×
Opus 5.5 · Claude	86	36×
GPT-6 Astra · ChatGPT	91	52×

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

Choosing the metric the whole company steers by.

Why it matters

Pick revenue and you steer by the rear-view mirror. Pick a vanity metric and teams will happily grow it while the product gets worse.

What good looks like

Measures value delivered to customers, not value captured
Grounded in what predicts retention in the data
Broken into input metrics a team can own
Says how it could be gamed, with guardrails

Deliberately not measured

Dashboard design
Naming

Capability tested

Metric design

The failure we’re looking for

Revenue, or a vanity metric, as the North Star

Grading

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