Tasks / Challenge

Challenge an idea

Can the model find the strongest reason an idea may fail, backed by evidence?

Measures the systemTask v1.2 · 2 casesDifficulty

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

From 14 graded outputs by 7 models. 64% were usable with at most a quick edit.

Reliably right

  1. Addresses the actual decision100% pass
    It commits early to not committing three squads and authorizing a six-week test, and specifies what result would change that.
    GPT-6.1 Sol · API · The CEO's embedded-payments bet
  2. Identifies material uncertainty100% pass
    It names the unknowns (representativeness of payment data, large-account volume access, actual mix) and resolves them with a bounded test.
    GPT-6.1 Sol · API · The CEO's embedded-payments bet
  3. Uses the interviews faithfully100% pass
    All quotes are accurate and correctly attributed to A05, A08, A04, A07, and A12.
    GPT-6 Luna · API · An AI SDR for small agencies

Where it slips

  1. A cheap test that can actually read out64% pass
    The pre-registered gate requires signed deals and renewal rather than an early signal such as qualified meetings or proposals, so it risks being too late for a cheap read-out.
    GPT-6.1 Sol · API · An AI SDR for small agencies
  2. Re-estimates the revenue correctly71% pass
    It shows working but lands on a central $2.3M and $1M–$6M range, not roughly the $5–10M range required by the grading rubric, and its $9.1M pilot upper bound is not used as the main re-estimate.
    Sonnet 5.5 · API · The CEO's embedded-payments bet
  3. Avoids unsupported claims71% pass
    The memo claims the pilot achieved ~0.38% blended take as a fact, which is not in the evidence and is not derived from it arithmetically; it also treats the pro-rata $1.43B locked value as a hard constraint without flagging the assumption.
    Opus 5.5 · Claude · The CEO's embedded-payments bet

Case viewer

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

The brief

We plan to sell an AI sales-development rep, an agent that finds prospects and sends personalised outbound email, to agencies with 5–20 staff. Before we commit a year to it, find the strongest reason this fails. Use the interview material below, including the three transcripts. Write it up for the founding team: the one reason, the evidence for it (quote the interviews), where the evidence cuts the other way, what we would need to see to be proved wrong, and the cheapest test that would show it. Keep it under 600 words.

What the model was given7 items: Scenario, Founders' hypothesis (from our planning doc), Interview summary (14 agencies, 5–20 staff, last six weeks), Interview log, interview-A05-content-agency.md, interview-A04-shopify-agency.md, interview-A08-packaging-studio.md
ScenarioWe are a three-person founding team with £800k of pre-seed funding and about 20 months of runway. Planned price: $600 a month per agency. Two competitors have each raised over $10m in the last year, selling AI SDRs mostly to software companies with 50+ staff.
Founders' hypothesis (from our planning doc)“Small agencies live feast or famine. When a big client leaves they have no pipeline, because the founder is too busy delivering to sell. An AI SDR keeps the pipeline full in the background, so the famine never comes. At $600 a month, one extra client a year pays for it many times over.”
Interview summary (14 agencies, 5–20 staff, last six weeks)11 of 14 get most of their new revenue from referrals and repeat clients (86% on average across those 11). 9 had tried outbound in the last two years; 7 of those stopped within six months, and none of the 7 closed a deal they could attribute to it. The 2 who kept going (A04, A07) have 15–20 staff and a dedicated person for business development. 3 (A08, A12, A14) said they regularly turn work away. 6 described feast-or-famine swings in new business. Average deal size across the 14: $18k; typical sales cycle 6–10 weeks.
Interview logAgency (staff) · services · share of new revenue from referrals and repeat clients · outbound tried? · outcome · representative quote A01 (8) · brand and web studio · 90% · yes, cold-email agency · stopped after 3 months · "We got meetings with people who'd never buy from us." A02 (14) · performance marketing · 85% · yes, LinkedIn automation · stopped after 4 months, account restricted · "LinkedIn shut our founder's account down. That's our best referral network." A03 (6) · PR boutique · 95% · no · — · "Every client we have came from someone vouching for us." A04 (18) · Shopify development · 40% (plus 30% partners, 30% outbound) · yes, in-house BD lead · kept, 30% of new revenue · "It works because of the follow-up, not the first email." A05 (11) · content · 85% · yes, lead-gen agency · stopped after 2 months, no deals · "Content is a trust purchase." A06 (7) · video production · 85% · no · — · "Our clients find us through the videos. Someone shares one and we get a call." A07 (16) · SEO · 45% · yes, outsourced lead generation · kept · "It pays for itself, just. Most of the meetings are a waste of time, but one in ten turns into a retainer." A08 (5) · packaging design · 95% · no · — · "If you sent me ten more leads in October I'd have to say no to nine of them." A09 (12) · paid social · 75% · yes, cold email · stopped after 5 months · "The replies were mostly people asking us to take them off the list." A10 (20) · B2B marketing · 70% · yes, contract SDR · stopped after 6 months · "It cost us about £400 for every meeting, and the meetings didn't close." A11 (9) · WordPress web agency · 90% · yes, cold email · stopped after 3 months · "Our domain ended up on a spam list. Took weeks to fix." A12 (10) · branding · 85% · no · — · "We're booked out till March. I don't need more leads, I need another designer." A13 (15) · HubSpot partner · 48% (plus 40% from HubSpot's partner directory) · yes, cold email · stopped after 4 months · "HubSpot sends us more than we can handle in a good quarter." A14 (6) · UX research · 90% · no · — · "We're small on purpose. We say no to about a third of enquiries."
interview-A05-content-agency.md31 lines · Download
# Interview A05: content agency, 11 staff
Participant: founder and managing director. Interviewer: our co-founder. 34 minutes, lightly edited.

**Interviewer:** Where did your last five clients come from?

**Participant:** Let me think. Two were old clients coming back: one had changed jobs and brought us into her new company. Two were introductions, one from a web agency we partner with and one from a client's CFO who'd seen our work. The fifth found us through a talk I gave at a SaaS marketing meetup. So none of them from anything you'd call outbound.

…
interview-A04-shopify-agency.md23 lines · Download
# Interview A04: Shopify development agency, 18 staff
Participant: head of business development. Interviewer: our co-founder. 29 minutes, lightly edited.

**Interviewer:** Tell me how new business works for you.

**Participant:** We're a bit unusual for an agency our size. I'm a full-time business-development person, and outbound is about 30% of our new revenue. Referrals and repeat work are still the majority, maybe 40% referrals and the rest partners: Shopify's partner directory and a couple of app companies who send us projects.

…
interview-A08-packaging-studio.md23 lines · Download
# Interview A08: packaging design studio, 5 staff
Participant: founder and creative director. Interviewer: our co-founder. 22 minutes, lightly edited.

**Interviewer:** How do new clients find you?

**Participant:** Almost entirely word of mouth. Food and drink brands talk to each other. Someone launches a range, their packaging does well on the shelf, and the next founder asks who did it. I'd say 95% referral. We've never done any outbound.

…
What a strong answer doesThe answer key the graders mark against

Names the load-bearing assumption as 'small agencies' growth is limited by a shortage of new leads': the interviews say it is limited by trust and capacity. 11 of 14 win most work through referrals and repeat clients; 7 of 9 who tried outbound quit within six months with no attributable deals, for reasons tied to trust (A05's 'content is a trust purchase', A01's meetings with people who'd never buy) and reputation risk (A02's restricted LinkedIn account, A11's spam listing); 3 turn work away (A08: 'I'd have to say no to nine of them'). Quotes the interviews accurately and attributes them correctly. Engages the counter-evidence honestly: outbound works for A04 and A07, but both are 15–20 staff with a dedicated BD owner, already use tools like Apollo and Clay, and credit the follow-up rather than the first email, so the segment is small, tooled and not what the product automates. Treats feast or famine as a real pain but notes agencies fix it through past clients and referrals (A05's reactivation emails, and her willingness to pay for help keeping in touch with past clients), pointing at a different product. Says what would prove the challenge wrong (for example, a meaningful share of 10–20-staff agencies with a BD owner and budget, or a concierge outbound test producing qualified meetings that turn into proposals), and designs a cheap test that measures an early signal, since a 6–10 week sales cycle means a six-week test cannot observe won deals.

Critical failures (cap the score at 40)
  • Invents interview quotes or attributes a quote to the wrong agency
  • Treats the founders' feast-or-famine hypothesis as confirmed by the interviews
Case

Staff-level · v1.10 · synthetic · B2B, go-to-market, customer research

The outputs

7 setups answered this brief. Pick up to three to read side by side.

Needs more than a quick edit3 of 11 checks passed

A PM would need to re-check evidence and repair the test before acting. The core argument is useful but not reliable as written.

Output

The One Reason This Fails

Agency services are high-trust purchases that cold automated outreach cannot convert without intensive, bespoke human follow-up—which time-poor founders cannot provide. Automating email generation solves top-of-funnel activity, but the actual point of failure is trust and follow-up capacity.

---

The Evidence

1. Zero conversion for founder-led shops: 7 of 9 agencies that tried outbound stopped within six months having closed zero deals. 2. Cold outreach delivers the wrong buyers: Services require credibility upfront. * A05 (11 staff): “Content is a trust purchase... The people on those calls had no idea who we were and, honestly, most of them didn’t have a budget.” * A01 (8 staff): “We got meetings with people who’d never buy from us.” * A10 (20 staff): “Cost us about £400 for every meeting, and the meetings didn’t close.” 3. The constraint is manual follow-up, not initial emails: * A04 (18 staff, dedicated BD): “It works because of the follow-up, not the first email... What wins deals is that I then send them a teardown of their site, a short video... The owner does sales on the side. They don't have the time to follow up, and the follow-up is the job.” 4. Asymmetric downside risk: Small agencies live on reputation. * A11: “Our domain ended up on a spam list.” * A02: “LinkedIn shut our founder’s account down. That’s our best referral network.” * A05: “I’d worry about our name going out on emails I haven’t read.”

---

Where the Evidence Cuts the Other Way

  • Feast-and-famine is real: 6 of 14 experience painful dry spells. A05 panicked when three clients left simultaneously.
  • Outbound works at scale with dedicated staff: A04 generates 30% of revenue via outbound; A07 makes it pay (“one in ten turns into a retainer”).
  • Willingness to pay exists if positioned right: A04 would pay $600/month if it replaced Apollo/Clay and saved research hours. A05 would pay $200/month for an automated tool to reactivate past clients and referral networks.

---

What We Would Need to See to Be Proved Wrong

  1. A solo founder (with no dedicated BD rep) closing an outbound deal with typical agency pricing ($18k) using standard outbound messaging, without spending >5 hours/week on bespoke pitch work.
  2. Willingness of agency owners to allow an autonomous agent to send cold outbound without manual pre-approval of every single touchpoint.

---

The Cheapest Test to Prove It

A 3-week "Wizard of Oz" Concierge Test with 3 Agencies (Cost: <$500): * Partner with 3 agencies suffering feast-or-famine (e.g., A05, A09, A10) who will commit to running a free, 30-day outbound campaign. * Manually emulate the AI SDR: scrape prospect lists, generate personalized emails, and handle automated sequencing from a secondary domain. * Kill criteria: If after booking 10–15 meetings across these agencies, zero advance to contract negotiations—or founders drop out because they cannot keep up with follow-ups and custom scoping—kill the product.

(Pivot signal: A05 offered to pay immediately for an AI assistant that drafts warm re-engagement notes to past clients and referral partners. That solves the same famine problem without the friction of cold outbound.)

Claims the judge couldn’t find in the brief

  • 7 of 9 agencies that tried outbound stopped within six months having closed zero deals.
  • 6 of 14 agencies experience painful dry spells.
  • A04 generates 30% of revenue via outbound.
  • A05, A09, and A10 are agencies suffering feast-or-famine.

Grades and run details

Decision model 95 · LLM judge 3 of 12 checks
Decision model checks
  • passUses the supplied evidence correctly44%
  • passAddresses the actual decision96%
  • passRespects explicit constraints36%
  • passIdentifies material uncertainty97%
  • partialAvoids unsupported claims23%
  • passProduces the required deliverable71%
  • passFinds the load-bearing assumption26%
  • passUses the interviews faithfully74%
  • passEngages the counter-evidence79%
  • passA cheap test that can actually read out94%
  • passTigers, not paper tigers94%
Run
Run
#1
API response time
22 s
Submitted
1 Oct 2026

Usable with a quick edit10 of 11 checks passed

The memo is structurally complete and well‑reasoned; a quick edit to shift the test gate to an early signal (e.g., qualified proposals) would make it fully usable without rechecking evidence or redoing the reasoning.

Output

To: Founding team Subject: We risk automating a channel our customers cannot turn into revenue

Do not commit a year to this product yet. The strongest reason it fails is that finding prospects and sending personalised emails does not remove the constraint on small agencies’ sales: the human work of establishing trust and converting interest. Our promise assumes an agent can keep revenue flowing while the founder stays busy delivering. The interviews suggest we would generate more work for that same founder.

Eleven of 14 agencies get most new revenue from referrals and repeat clients, averaging 86% across those eleven. Of nine that tried outbound, seven stopped within six months without an attributable deal. This is more than unfamiliarity with the channel: most have bought or attempted it and abandoned it.

A05 shows the mechanism. Its lead-generation agency booked six or seven meetings a month, but prospects lacked budgets or wanted evidence of relevant experience. With a referred prospect, “the first call was about scope, not about whether we were any good”. Personalisation may earn attention; these interviews do not establish that it supplies the credibility needed to buy.

The successful case makes the constraint clearer. A04 generates 30% of new revenue through outbound, but has a dedicated BD lead who follows emails with site teardowns, videos and calls: “They don’t have the time to follow up, and the follow-up is the job.” Automating the first step leaves that job with the founder.

Some agencies do not need additional demand at all. A08 says, “Getting them isn’t. Handling them is,” and needs a senior designer. Three agencies regularly turn work away. Feast or famine is not a universal unmet need.

What cuts against this conclusion

Six agencies do report feast-or-famine swings. A04 and A07 prove outbound can work, although both have dedicated business-development capacity. A04 would pay $600 if we replaced research tools, saved time, integrated with HubSpot and required email approval. That supports a narrower research assistant proposition.

Past failures also do not prove our product would fail: targeting and execution may improve, and A05’s two-month trial barely covered the typical 6–10-week sales cycle. Fourteen interviews cannot establish market prevalence.

But “one extra client pays for it” skips the disputed step: winning that client. The $7,200 annual subscription is 40% of the $18,000 average deal’s revenue, before delivery costs and selling time.

What would prove us wrong, and the cheapest test

Run a paid concierge pilot before building. Recruit five agencies with spare delivery capacity, a stated growth goal and no dedicated BD person. Charge the planned $600 monthly price. Using existing tools, manually deliver prospect research and personalised drafts, with approval before sending. Founders must handle follow-up; we must not secretly supply the missing salesperson.

Run for 12 weeks to cover the stated sales cycle.1 Agree a founder-time ceiling upfront, record all selling time and track qualified opportunities through to attributable signed deals, contribution margin and renewal.

A proposed gate: at least three of five win business whose expected contribution covers subscription and selling costs, remain within the time ceiling and renew at full price. Meetings alone do not pass.

If only agencies with BD staff succeed, narrow the customer and promise. A05’s request for past-client reminders suggests another test, but “I’d pay for that tomorrow” is not payment. Do not turn one appealing interview into the next year-long commitment.

What a PM had to fix

  1. 1Test or gate too weakTighten the testQuick edit

    What we’d changeSigning and renewing inside 12 weeks is a bar that quietly favours the memo's own conclusion. Extend the window or add a leading indicator such as qualified proposals, and include the risk to agencies' LinkedIn channel.

Grades and run details

Decision model 95 · LLM judge 11 of 12 checks
Decision model checks
  • passUses the supplied evidence correctly67%
  • passAddresses the actual decision99%
  • passRespects explicit constraints39%
  • passIdentifies material uncertainty99%
  • passAvoids unsupported claims20%
  • passProduces the required deliverable87%
  • passFinds the load-bearing assumption99%
  • passUses the interviews faithfully79%
  • passEngages the counter-evidence79%
  • partialA cheap test that can actually read out33%
  • passTigers, not paper tigers98%
Run
Run
#1
Time to output
72 s
Submitted
25 Sept 2026

Needs more than a quick edit7 of 11 checks passed

It is over the requested length and contains two unsupported current-state claims that an editor would need to remove or verify. Otherwise the reasoning is usable, but those are more than wording fixes.

Output

Pre-mortem: AI SDR for 5–20 person agencies

The one reason it fails

Small agencies do have a pipeline problem, but cold outbound is the wrong fix. Their work is sold on trust, so it comes through introductions. Cold email produces meetings that don't close, and the few agencies where outbound works have a person doing the follow-up. At $600 a month we would be selling a channel that this buyer has already tried and dropped, into a market where two funded competitors are moving down.

Evidence

  • Referrals dominate. 11 of 14 agencies get about 86% of new revenue from referrals and repeat clients. A05: "Content is a trust purchase." A03: "Every client we have came from someone vouching for us."
  • Outbound has a poor track record. 9 tried it, 7 quit within six months, and none of the 7 could attribute a closed deal to it. The reasons are structural: A01 "meetings with people who'd never buy from us"; A09 "mostly people asking us to take them off the list"; A11 "Our domain ended up on a spam list"; A02 "LinkedIn shut our founder's account down. That's our best referral network."
  • The cost is high. A10: "about £400 for every meeting, and the meetings didn't close."
  • Success depends on humans. A04: "It works because of the follow-up, not the first email… The AI could save me the research hours, sure. It can't do the call." And: "most agencies our size don't have someone like me… Most tried it once, got burned, and went back to referrals."
  • The famine is real, but warm outreach solved it. A05 recovered by emailing three years of past clients, and two came back within a month. Of our product, A05 said: "Cold? Probably not… Six hundred for cold outbound, no."
  • Some agencies don't need leads. A08, A12 and A14 turn work away. A08: "It would make my problem worse."

Where the evidence cuts the other way

  • 6 of 14 report feast-or-famine, so the pain is real.
  • A07 kept outbound: "It pays for itself, just… one in ten turns into a retainer." A04 would pay $600 if the tool replaced Apollo and Clay.
  • None of the failures used a modern AI agent. Better targeting and lower cost per meeting might change the results, though the trust and deliverability complaints suggest it won't fix most of them.
  • The two who kept going are not cleanly separable by size. Of the four agencies with 15–20 staff, two stopped (A10, A13).
  • A05 would pay about £200 a month for a warm-relationship tool. That points to a different product at a third of our price, not proof that ours works.
  • 14 interviews is a small sample, and willingness to pay was hypothetical.

What would prove us wrong

In a paid pilot with agencies that have no BD hire, the product produces qualified pipeline at a cost they'll keep paying for. Roughly: at least half the pilot agencies get a qualified opportunity or proposal within 90 days, and most say they'd continue at $600. The maths also needs checking. $600 a month is $7,200 a year, but the buyer's gain is the margin on an $18k deal, not the revenue. "Many times over" only holds if outbound reliably closes more than one deal a year.

Cheapest test

Run a concierge pilot before building anything more:

  1. Recruit 10 agencies with 5–14 staff and no BD person. Exclude the referral-saturated ones like A08, A12 and A14.
  2. Charge $600 a month, or at least a serious deposit. Free pilots tell us nothing about willingness to pay.
  3. Do the work ourselves with off-the-shelf tools (Apollo, Clay, an email sender) on separate sending domains. The agency approves every email, since A04 and A05 both demanded that.
  4. Set the pass line before starting: at least 5 of 10 agencies get a qualified proposal within roughly 14 weeks (the 6–10 week sales cycle plus setup), and at least 6 of 10 agree to continue.
  5. Track reply rate, unsubscribe and spam complaints, cost per qualified meeting, and meeting-to-proposal rate.

This costs a few weeks of founder time and about £1–2k in tools, roughly 4 months of the 20-month runway. In parallel, offer 5 of the same agencies a £200 warm-outreach version, to see whether the real product is relationship follow-up.

If the pilot misses the line, don't spend the year on cold outbound for this segment.

Claims the judge couldn’t find in the brief

  • Two funded competitors are moving down into this market.
  • None of the failed outbound attempts used a modern AI agent.

Grades and run details

Decision model 91 · LLM judge 7 of 12 checks
Decision model checks
  • passUses the supplied evidence correctly47%
  • passAddresses the actual decision96%
  • partialRespects explicit constraints40%
  • passIdentifies material uncertainty99%
  • partialAvoids unsupported claims28%
  • passProduces the required deliverable74%
  • passFinds the load-bearing assumption87%
  • passUses the interviews faithfully83%
  • passEngages the counter-evidence66%
  • passA cheap test that can actually read out99%
  • passTigers, not paper tigers97%
Run
Run
#1
API response time
30 s
Submitted
29 Sept 2026

Check by check

Where the outputs differ comes first. Open a check to read the judge’s reasons. “Mixed” means the two graders disagreed.

Where they differ 8

Uses the supplied evidence correctlyMixedRightMixed
Gemini 3.8 Flash · API

It misstates or overreaches several current-situation facts, including 'zero deals', 'painful dry spells', A04's 30% of revenue, and A09/A10 as feast-or-famine agencies.

GPT-6 Astra · ChatGPT

All factual claims about the current situation are drawn accurately from the brief and supplied context.

Sonnet 5.5 · API

It includes at least two current-state claims, that competitors are moving down and that none of the failures used a modern AI agent, that are not supported by the supplied context.

Respects explicit constraintsMixedRightWrong
Gemini 3.8 Flash · API

The required sections and length are present, but the proposed test is internally inconsistent (3-week test vs 30-day campaign) and would not reliably enforce the brief's cheap-test requirement.

GPT-6 Astra · ChatGPT

The memo is under 600 words, addresses the founding team, and includes the required one reason, evidence, counter-evidence, proof-wrong condition, and cheapest test.

Sonnet 5.5 · API

It exceeds the 600-word limit, running roughly 700 words, even though it includes the requested sections.

Avoids unsupported claimsWrongRightWrong
Gemini 3.8 Flash · API

It presents unsupported or overconfident claims as fact, especially the feast-or-famine status of A09/A10 and the zero-deal characterisation.

GPT-6 Astra · ChatGPT

Interpretations are clearly labelled as hypotheses or suggestions, not presented as established facts.

Sonnet 5.5 · API

It presents the competitor 'moving down' claim and the 'none used a modern AI agent' claim as fact rather than as labelled inference.

Produces the required deliverableMixedRightMixed
Gemini 3.8 Flash · API

Although structured as requested, it is not usable without correcting evidence errors and the contradictory test design.

GPT-6 Astra · ChatGPT

The memo is a complete, usable document in the requested form, length, and reader, requiring only light editing.

Sonnet 5.5 · API

It has the right form and reader, but it is not within the requested length.

Finds the load-bearing assumptionMixedRightRight
Gemini 3.8 Flash · API

It focuses on trust and follow-up capacity but does not make lead-constrained growth the central challenge or use the turning-work-away evidence that directly refutes the founders' assumption.

GPT-6 Astra · ChatGPT

The critique centers on the assumption that small agencies' growth is limited by a shortage of new leads, showing trust and capacity as the real constraints.

Sonnet 5.5 · API

It targets the assumption that growth is lead-constrained, using the referral, abandoned-outbound and turning-work-away evidence as the central challenge.

Engages the counter-evidenceMixedRightRight
Gemini 3.8 Flash · API

It names A04 and A07 but does not fully explain what sets them apart or the implication for the addressable segment, such as A07's dedicated BD owner and existing tooling.

GPT-6 Astra · ChatGPT

A04 and A07 are named, their dedicated BD capacity and follow-up are highlighted, and the implication of a narrow addressable segment is stated.

Sonnet 5.5 · API

It names A04 and A07, identifies their size/BD-owner/tooling and follow-up dependence, and explains what that means for the addressable segment.

A cheap test that can actually read outMixedWrongRight
Gemini 3.8 Flash · API

The test is cheap in spirit but inconsistent and not clearly observable within the 6–10 week sales cycle, using contract negotiations after a 30-day campaign.

GPT-6 Astra · ChatGPT

The proposed test gate relies on signed deals (won business), but a 6–10 week sales cycle means the 12-week test likely cannot observe this late signal; the criteria require measuring an early signal like qualified meetings or proposals.

Sonnet 5.5 · API

It proposes a cheap concierge pilot with explicit thresholds and measures qualified proposals and meetings rather than won deals, matching the 6–10 week sales cycle.

Tigers, not paper tigersMixedRightRight
Gemini 3.8 Flash · API

It does not triage risks or dismiss a plausible worry; it leaves capacity/turning-work-away and the small addressable segment underdeveloped.

GPT-6 Astra · ChatGPT

It dismisses the worry that a small sample invalidates the finding by noting that 14 interviews cannot establish prevalence, and surfaces the unspoken issue that outbound adds follow‑up work the founder cannot do.

Sonnet 5.5 · API

It separates the central trust and capacity risks from fixable targeting/cost issues and surfaces the warm-outreach product alternative and margin-math problem.

All got right 3

Addresses the actual decisionRightRightRight
Gemini 3.8 Flash · API

It commits early to one failure reason for the founding team and states conditions that would prove the challenge wrong.

GPT-6 Astra · ChatGPT

The output gives a clear 'do not commit a year' answer early, framed for the founding team, and specifies what would change it.

Sonnet 5.5 · API

It commits early to the answer that cold outbound is the wrong fix for this segment and says what pilot result would change that.

Identifies material uncertaintyRightRightRight
Gemini 3.8 Flash · API

It names material unknowns—solo-founder conversion and autonomy tolerance—and gives a test and kill criteria to resolve them.

GPT-6 Astra · ChatGPT

It acknowledges that past failures do not prove the product will fail, the small sample limits prevalence estimates, and states what results would change the call.

Sonnet 5.5 · API

It names the small sample, hypothetical willingness to pay, and the untested modern-AI possibility, and proposes thresholds and observable signals to resolve them.

Uses the interviews faithfullyRightRightRight
Gemini 3.8 Flash · API

The quoted interview material is attributed to the right agencies and is accurate apart from minor ellipses and wording.

GPT-6 Astra · ChatGPT

Quotes and paraphrases from A05, A04, A08 are accurate and correctly attributed, with no invented material.

Sonnet 5.5 · API

All quotes or close paraphrases match the transcripts and log, and each is attributed to the correct agency.

Results

Every setup we’ve tested on this task, across all cases and repeats, graded on the current checklist. Calibrated: the graders match our PM on 88% of checks.

#Model · HarnessTask scoreDecision modelLLM judgeRunsCritical failures
1GPT-6.1 SolwithAPI100.095.82None
2GPT-6 AstrawithChatGPT93.295.82None
3GPT-6 LunawithAPI97.791.72None
4Sonnet 5.5withAPI90.975.02None
5Gemini 3.5 Flash-LitewithGemini88.675.021 capped
6Opus 5.5withClaude93.291.721 capped
7Gemini 3.8 FlashwithAPI93.237.521 capped

About the task

The PM job

Pressure-testing a proposal before committing a team to it.

Why it matters

The useful critic finds the one assumption everything rests on. Theatrical negativity is easy to generate and useless in a planning meeting.

What good looks like

  • Identifies the load-bearing assumption
  • Separates the risks that could kill it from the ones that only look scary
  • Uses the supplied evidence, not generic risks
  • Proposes the cheapest way to test the assumption

Deliberately not measured

  • Tone
  • Number of objections raised
Capability tested

Evidence-based critique

The failure we’re looking for

Theatrical negativity without evidence

Grading

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

Variants

Vanilla prompt (core) · With Roast Me skill · Staff level: a company bet with a long evidence pack

This task measures the whole setup. Tools, instructions and skills in the harness do real work here, so read the harness as carefully as the model name.