Tasks / Growth

Write a GTM plan

Can the model pick the customers who value the difference most, position against what they do today, and build a go-to-market that lasts beyond launch day?

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 16 graded outputs by 8 models. 50% were usable with at most a quick edit.

Reliably right

  1. Positioned against the real alternative100% pass
    It names the real alternatives (accountant-run payroll with mistakes, owner spreadsheet with risk) and ties the message to confidence and accuracy that those alternatives lack.
    GPT-6 Astra · ChatGPT · Payroll, for the people who fear it
  2. Protects the self-serve engine100% pass
    Uses the 70% free-to-paid stat and product-qualified deal data to argue against capping the free plan, and keeps self-serve open.
    GPT-6 Astra · ChatGPT · Enterprise, without killing the engine
  3. Does the capacity sums100% pass
    Works out deals needed (69-80), qualified opportunities (134), and AE-years (8) with checkable arithmetic.
    GPT-6 Astra · ChatGPT · Enterprise, without killing the engine

Where it slips

  1. Avoids unsupported claims47% pass
    Presents the incorrect 'five of six no decision' figure as established fact, which is an unsupported claim about the current situation.
    Opus 5.5 · Claude · Enterprise, without killing the engine
  2. Uses the supplied evidence correctly52% pass
    The claim 'Five of our six losses on existing-user deals ended in no decision' is not supported by the data (4 no-decision, 2 Miro), so a statement about the current situation is invented.
    Opus 5.5 · Claude · Enterprise, without killing the engine
  3. Identifies material uncertainty73% pass
    The plan does not name specific unknowns that could change the decision, nor does it say how they would be resolved or what would trigger a pivot.
    Gemini 3.8 Flash · API · Payroll, for the people who fear it

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.

  1. Positioned against the real alternative

    Does the plan name what the target customer does today without this product (including the status quo or doing nothing), and tie the message to the value that alternative lacks?

    Passes when Names the real alternative and the differentiated value against it, and picks the segment that cares most about that value.

  2. Sells against a pain the buyer feels

    Is the core message built on a pain or risk the target buyer already feels, with evidence from the material, rather than on generic upside like saving time?

    Passes when Leads with the pain the evidence shows buyers care about, and uses that evidence to choose the message.

  3. Diagnoses message or channel first

    Does the plan work out whether the problem is the message (people reach it but don't convert) or the channels (the right people never see it), and spend accordingly?

    Passes when Uses conversion and reach data to say which it is, and puts the effort there.

  4. Plans beyond launch week

    Does the plan say how demand keeps coming after launch, with the ongoing engine and what's measured over months, rather than treating launch day as the whole plan?

    Passes when Gives the ongoing motion after launch and the measures that show whether it's working over time.

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 · and 2 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 the PM for Ledgerly Payroll, a new add-on for our accounting product, launching to existing customers on 3 November. Marketing has drafted the launch plan below. Write the go-to-market plan you'd run instead (or the changes you'd make to theirs), in no more than 900 words: who we go after first, what we say to them, through which channels, and how we'll know it's working over the following three months. The files are attached.

What the model was given5 items: About Ledgerly, marketing_launch_draft.md, customers_by_size_and_payroll_method.csv (penalty notices: share who told us they got a payroll penalty notice last year), beta_headline_test.csv (two versions of the in-app payroll page, shown at random to beta customers), customer_calls.md
About LedgerlyAccounting software for small businesses: invoicing, bank feeds, bookkeeping. 8,000 paying customers. Payroll is a new add-on at $6 per employee per month.
marketing_launch_draft.md7 lines · Download
# Draft launch plan (Head of Marketing)

- Launch day 3 November: press release announcing Ledgerly as "the first Financial Operating System for small business".
- Paid social and search ads to new prospects: $60,000 in November.
- In-app banner to all 8,000 customers: "Run payroll in minutes".
- Webinar in launch week.
- Success measure: press mentions and sign-ups in launch week.
customers_by_size_and_payroll_method.csv (penalty notices: share who told us they got a payroll penalty notice last year)employees,current_payroll,customers,penalty_notices_last_year_pct 1-4,Owner does it in a spreadsheet,3140,4 1-4,Accountant runs it,1020,6 5-25,Owner does it in a spreadsheet,690,19 5-25,Accountant runs it,1480,23 5-25,Another payroll tool,1210,7 26-50,Another payroll tool,460,5
beta_headline_test.csv (two versions of the in-app payroll page, shown at random to beta customers)headline,landing_page_visits,started_setup,ran_first_payroll "Run payroll in minutes, right where you do your books",4210,92,38 Never get a payroll penalty again,4160,268,117
customer_calls.md10 lines · Download
# Customer calls about payroll (excerpts, September)

1. Bakery, 12 staff: "My accountant does payroll. Last spring they filed late and I got a $1,400 penalty. I don't know what I'm paying them for."
2. Plumbing firm, 18 staff: "Time isn't the issue. I'm scared of getting it wrong. The tax office letter was the worst week of my year."
3. Design studio, 3 staff: "It's me and two freelancers. A spreadsheet's fine."
4. Café group, 22 staff: "We use another tool, it's OK. Switching sounds like a nightmare mid-year."
5. Physio clinic, 9 staff: "I'd move tomorrow if it meant one less thing that can go wrong with the tax office."
…
What a strong answer doesThe answer key the graders mark against

Picks customers with 5 to 25 employees whose accountant runs payroll (1,480 customers, 23% had a penalty notice) and those doing it in a spreadsheet (690, 19%), not the smallest businesses (happy with a spreadsheet) or those on another payroll tool (switching mid-year is the barrier). Builds the message on fear of penalties and errors, not time: in the beta, 'Never get a payroll penalty again' got about three times as many setups as 'Run payroll in minutes' (268 against 92 from similar traffic, 6.4% against 2.2%) and more first payrolls (117 against 38), and the calls say the same. Reads this as a message problem, not a channel problem: the existing customer base already sees the page, so fix the message and reach existing customers in-app and through their accountants before spending $60,000 on ads to strangers. Drops the 'Financial Operating System' category claim. Plans beyond launch week: prompts around payroll and tax deadlines, accountant referrals, and monthly measures such as payroll attach in the target segment, first payrolls run, and penalties avoided.

Critical failures (cap the score at 40)
  • Keeps 'Run payroll in minutes' as the lead message despite the headline test
Task

Mid-level · v1.0

The outputs

8 setups answered this brief.

Usable with a quick edit12 of 12 checks passed

The plan is evidence-based, actionable, and covers all required elements; a PM could execute it with only minor wording adjustments.

Output

Ledgerly Payroll go-to-market plan | 3 November–31 January

1. Start with customers who have the clearest problem

First priority: existing customers with 5–25 employees who currently run payroll themselves or rely on an accountant. That’s 2,170 customers: 690 spreadsheet users and 1,480 whose accountant runs payroll. In these groups, 19% and 23%, respectively, said they received a payroll penalty notice last year. Treat that as a signal of pain—not proof that Ledgerly would prevent penalties.

Lead with accountant-run customers who want more visibility and control, while making clear Ledgerly isn’t accusing their accountant of poor work. Follow with spreadsheet users in the same size band, for whom a simpler alternative to manual payroll may appeal.

Defer broader prospecting. Among customers with 1–4 employees, reported penalty notices are lower (4–6%); several calls also suggest some are content with free spreadsheets. Customers already using another payroll tool reported fewer notices (7% for 5–25 employees), and one call highlighted switching friction. Test those groups later, after validating onboarding and migration. Don’t prioritize 26–50-employee customers until we know the product and support experience fit their needs.

2. Say what we can support—not what we can’t promise

The beta headline test is a useful directional signal:

  • “Never get a payroll penalty again”: 268/4,160 visitors started setup (6.4%); 117 ran first payroll (2.8%).
  • “Run payroll in minutes…”: 92/4,210 started (2.2%); 38 ran first payroll (0.9%).

The penalty-focused version performed better, but don’t use “Never get a payroll penalty again” unless Ledgerly can substantiate that guarantee. Instead test responsible variants such as:

  • “More confidence in payroll, alongside your books.”
  • “Take control of payroll and reduce the risk of costly mistakes.”

Be explicit that payroll is $6 per employee per month. Emphasize control, confidence and working in the same accounting product—not guaranteed outcomes or unverified features. Tailor the message: “Keep a closer eye on payroll” for accountant-run customers; “Move beyond the spreadsheet” for owner-run customers.

3. Use customer channels before buying reach

  • Before launch: Segment the 2,170 priority customers and invite a small cohort to try setup. Interview users who start but don’t finish; check readiness, support capacity and any limitations before broad promotion.
  • Launch, 3 November: Send segmented email to the priority cohort and show a targeted in-app message when those customers use Ledgerly. Link directly to a clear product and pricing page, not a generic announcement. Offer a live, practical webinar on setup and common payroll concerns, with time for questions.
  • During November–January: Follow up based on behavior: a setup reminder for visitors who don’t start, and personal onboarding or support for customers who start but don’t run payroll. Invite accountant-run customers to involve their accountant where appropriate. Expand in-app promotion to other segments only after reviewing early results.
  • Earned media: A factual product announcement is fine, but don’t make press mentions the launch objective. Skip the $60,000 broad paid campaign to new prospects until existing-customer activation and retention are proven.

4. Measure adoption and outcomes for three months

Track weekly by customer segment and by funnel step: eligible → reached → product-page visit → setup started → setup completed → first payroll run → repeat payroll run. Report both counts and conversion rates; compare messages with a randomized test where possible.

Use the beta as a benchmark, not a forecast. A reasonable initial goal is to beat its stronger variant’s 6.4% visit-to-setup-start and 2.8% visit-to-first-payroll rates in the priority cohort. Also set a three-month target for repeat payroll runs among first-time users, and review setup abandonment to find where customers get stuck.

Pair adoption metrics with guardrails: payroll-related support contacts, failed or corrected runs, cancellations/refunds, and customer feedback about confidence and accuracy. Review weekly with Product, Support and Marketing. If activation is weak, fix onboarding or clarify the offer before adding spend. If customers run payroll but don’t repeat, investigate reliability, fit and switching friction before expanding the campaign.

Grades and run details

Decision model 100 · LLM judge 100
Decision model checks
  • passUses the supplied evidence correctly77%
  • passAddresses the actual decision93%
  • passRespects explicit constraints66%
  • passIdentifies material uncertainty83%
  • passAvoids unsupported claims63%
  • passProduces the required deliverable82%
  • passPicks the segment with the pain100%
  • passUses the headline test77%
  • passPositioned against the real alternative98%
  • passSells against a pain the buyer feels96%
  • passDiagnoses message or channel first99%
  • passPlans beyond launch week100%
Run
Run
#1
API response time
17 s
Submitted
5 Oct 2026

Needs more than a quick edit10 of 12 checks passed

The plan assumes, without evidence, that individual penalty recipients can be pinpointed and that accountants are the decision-makers. A PM would need to verify these data and assumptions before acting.

Output

Ledgerly Payroll: Go-to-Market Plan

Core change: Sell payroll to existing customers who have the most to lose from getting it wrong, lead with accuracy and penalty protection, and hold most of the $60k until we have conversion data.

What I'd change in the draft

  • The headline. In the beta, "Never get a payroll penalty again" roughly tripled setup starts (6.4% vs 2.2%) and first payroll runs (2.8% vs 0.9%). But "never" is a promise we can't keep. I'd use claims the product can support, such as "Filing reminders and pre-payroll checks, so fewer penalties," and get legal sign-off before launch.
  • "First Financial Operating System." It's unverifiable and means nothing to a bakery owner. Drop it.
  • $60k on paid ads to new prospects. We haven't proven the product yet, and our 8,000 customers are cheaper to reach and already trust us. Paid spend should follow proven conversion.
  • Success measured by press mentions and launch-week sign-ups. Neither tells us whether customers pay, use payroll, or avoid penalties.

Who we go after, in order

Penalty exposure is concentrated in businesses with 5–25 staff:

  • Accountant-run at 5–25 staff: 23% got a penalty notice last year.
  • Spreadsheet-run at 5–25 staff: 19%.
  • Owner-run at 1–4 staff: 4%.
  • Other-tool users: 5–7%.

About 2,170 customers sit in the two high-penalty 5–25 groups, and roughly 470 of them got a notice last year.

Wave 1 (launch, 3 Nov): 5–25 staff, spreadsheet or accountant-run (~2,170 customers). This is where the penalty message earns its keep. Start with the customers who got a notice, using our survey data, then the rest of the segment.

Wave 2 (mid-November): owner-run at 1–4 staff (3,140 customers). This is the largest pool but has low penalty pain, so the message is time saved. Their Sunday-spreadsheet routine is the pitch.

Wave 3 (January): other-tool users (1,670 customers). Don't ask anyone to switch mid-year (customer 4 calls it "a nightmare"). Pitch migration at the start of the new year, and offer to do the import for them.

Later: accountant-run at 1–4 staff (1,020 customers). The accountant makes this decision, so reach them through the accountant channel below.

Skip payroll messaging for businesses with no payroll pain, such as the three-person design studio (customer 3).

What we say

  • Accountant-run, 5–25 staff: "Payroll filings checked and reminded before they're due. Keep your accountant and see their work in the same place." This addresses customer 2's fear of getting it wrong and customer 1's question about what they're paying for. The "keep your accountant" line is a hypothesis to test.
  • Spreadsheet-run, 5–25 staff: "Stop checking payroll on Sunday night," with penalty protection as the second line.
  • Owner-run, 1–4 staff: "Run payroll in minutes, right where you do your books." This is beta headline A, which is fine for a segment that cares about time, not penalties.
  • Other-tool users: "We'll move your payroll over for you at the start of the new year."

Accuracy claims have to match what the product does. Customer 8 says "If you can promise it's right, I'm in," so the proof needs to be concrete: which checks run, which reminders fire, and what happens when a figure looks wrong.

Channels

  1. In-app, segmented (week 1). Show headline B to the 5–25 segments and headline A to the 1–4 segment. It's free and reaches everyone who logs in.
  2. Targeted email and account-manager calls (weeks 1–2). Email wave 1 from named account managers. Have support call the highest-risk accounts, those who got a notice last year, first.
  3. Accountant pilot (from launch, ~25 firms). Accountants serve many of our customers, and an accountant-run customer decides with their accountant. Invite the firms that manage the most of our customers to run payroll for clients inside Ledgerly. This means we need to decide up front how partners are paid.
  4. Webinar (launch week). "A payroll penalty checklist for businesses with 5–25 staff," invited to wave 1 only. Narrow beats generic.
  5. Paid (hold). Spend $10k on retargeting and search from week 3, but only if the wave 1 landing page converts at least as well as beta headline B. Hold the remaining ~$50k until the 60-day numbers justify it.
  6. Press (low-key). A short release with a real customer quote and no "first" claim. It's cheap and supports the story without carrying the plan.

Measuring the next three months

Primary metric: payroll customers, meaning customers who have run at least one payroll, by segment, and the revenue from them ($6 per employee per month).

Targets (reset after the 30-day baseline): - Wave 1: 10% of the ~2,170 customers have run a payroll by day 90 (~220 customers, roughly $18k MRR at about 12 employees each). - Wave 2: 4% by day 90. - Wave 3: 10% of other-tool users migrated by end of January.

Decision rules: - If wave 1's first-payroll rate is under 2% after 30 days, pause paid spend and rewrite the messaging. - If it's over 5%, release the held budget to the segments that convert best.

Guardrails (any breach stops the rollout): - Zero filing errors or penalties traced to Ledgerly. - Support tickets per payroll run stay within agreed levels. - No rise in churn or refunds among existing customers.

Penalty outcomes won't show in three months. As a proxy, track on-time filing rates for adopters against matched non-adopters, and report it to the board in February.

Claims the judge couldn’t find in the brief

  • We have survey data identifying which specific customers received a penalty notice last year, allowing us to start with those individuals.
  • For accountant-run customers, the accountant makes the buying decision.

Grades and run details

Decision model 100 · LLM judge 77
Decision model checks
  • passUses the supplied evidence correctly10%
  • passAddresses the actual decision98%
  • passRespects explicit constraints30%
  • passIdentifies material uncertainty97%
  • passAvoids unsupported claims33%
  • passProduces the required deliverable69%
  • passPicks the segment with the pain100%
  • passUses the headline test97%
  • passPositioned against the real alternative100%
  • passSells against a pain the buyer feels100%
  • passDiagnoses message or channel first100%
  • passPlans beyond launch week100%
Run
Run
#1
API response time
33 s
Submitted
8 Oct 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 2

Uses the supplied evidence correctlyRightMixed
GPT-6 Luna · API

All factual claims about the current situation are directly supported by the supplied data or arithmetic.

Haiku 5.5 · API

The plan asserts two facts about the current situation not supported by the brief: that individual penalty recipients can be identified from survey data, and that accountants make the payroll purchasing decision.

Avoids unsupported claimsRightMixed
GPT-6 Luna · API

Interpretations like 'want more visibility' are tied to call evidence and not presented as established fact; hypotheses are hedged.

Haiku 5.5 · API

The plan states that the accountant makes the payroll decision as a fact, and implies precise individual penalty data exists, without labelling these as assumptions or hypotheses.

All got right 10

Addresses the actual decisionRightRight
GPT-6 Luna · API

The output commits to a clear primary segment, message, channels, and measures, and states conditions for expanding later.

Haiku 5.5 · API

The output commits to a clear wave plan early and states decision rules (conversion thresholds for budget release and messaging changes).

Respects explicit constraintsRightRight
GPT-6 Luna · API

The plan stays under 900 words, does not lead with 'save time', and avoids paid ads to new prospects as the main channel.

Haiku 5.5 · API

The output is a go-to-market plan under 900 words, addresses the requested elements (segments, messaging, channels, measures), and respects the launch date and audience.

Identifies material uncertaintyRightRight
GPT-6 Luna · API

It identifies that the penalty message needs substantiation, treats the beta as directional, and specifies how weak activation or repeat usage would change the approach.

Haiku 5.5 · API

The plan names unknowns (conversion rates, messaging effectiveness) and sets explicit thresholds and dates to resolve them.

Produces the required deliverableRightRight
GPT-6 Luna · API

The output is a complete go-to-market plan with segment, message, channels, and three-month measures, within the word limit.

Haiku 5.5 · API

The output is a complete, actionable GTM plan in the required format, usable by a PM to execute the launch with light edits.

Picks the segment with the painRightRight
GPT-6 Luna · API

It targets 5–25 employee customers with accountant-run or spreadsheet payroll, cites penalty rates, and explains why smaller and tool-using segments are deferred.

Haiku 5.5 · API

The plan correctly prioritizes 5-25 employee spreadsheet and accountant-run segments (high penalty rates), explains why 1-4 and other-tool users come later.

Uses the headline testRightRight
GPT-6 Luna · API

It cites the beta test results and builds the message on penalty and accuracy concerns, even while avoiding the unsubstantiated guarantee.

Haiku 5.5 · API

The plan uses the beta headline test results to base the core message on penalty protection for the high-pain segments and time-saving only for the low-pain segment.

Positioned against the real alternativeRightRight
GPT-6 Luna · API

It names the real alternatives (accountant, spreadsheet, another tool) and ties the message to confidence and reduced risk that those alternatives lack.

Haiku 5.5 · API

The plan names current alternatives (accountant, spreadsheet, another tool) and differentiates against the value they lack (accuracy, penalty protection) for each segment.

Sells against a pain the buyer feelsRightRight
GPT-6 Luna · API

The core message leads with penalty fear and accuracy, supported by call quotes and the beta test, not generic time-saving.

Haiku 5.5 · API

The lead message for the primary segment is built on penalty fear and accuracy, directly supported by the customer calls and headline test.

Diagnoses message or channel firstRightRight
GPT-6 Luna · API

It uses conversion data to diagnose that message matters more than reach, and focuses on in-app and email to existing customers before buying ads.

Haiku 5.5 · API

The plan diagnoses message as the bottleneck (uses in-app test data and existing reach) and defers paid spend until conversion is proven, focusing effort on fixing the message first.

Plans beyond launch weekRightRight
GPT-6 Luna · API

The plan includes ongoing follow-up, behavior-based nudges, and three-month funnel and guardrail metrics beyond launch week.

Haiku 5.5 · API

The plan includes a multi-wave rollout, ongoing channels (email, accountant pilot, retargeting), and 90-day metrics with decision rules beyond launch week.

Results

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

#Model · HarnessTask scoreDecision modelLLM judgeRunsCritical failures
1GPT-6 AstrawithChatGPT100.0100.02None
2Opus 5.5withClaude96.085.22None
3GPT-6 LunawithAPI94.289.32None
4GPT-6.1 SolwithAPI93.878.02None
5Sonnet 5.5withAPI98.177.72None
6Haiku 5.5withAPI94.277.72None
7Gemini 3.8 FlashwithAPI73.981.92None
8Gemini 3.5 Flash-LitewithGemini74.263.52None

What you get for the cost

No setup scores within 5 points of the leader at half its cost or less here, so the leader is also the best value.

Task score against relative cost for each setup. The table below has the same figures.607080901001×3×10×30×100×Cost per output, relative to the cheapest setup (log scale) →↑ Task scoreGPT-6 Astra · ChatGPT: task score 100, 81× the cheapestOpus 5.5 · Claude: task score 91, 75× the cheapestGPT-6 Luna · API: task score 90, the cheapestGPT-6.1 Sol · API: task score 84, 16× the cheapestSonnet 5.5 · API: task score 84, 46× the cheapestHaiku 5.5 · API: task score 82, 3.3× the cheapestGemini 3.8 Flash · API: task score 77, 11× the cheapestGemini 3.5 Flash-Lite · Gemini: task score 68, 3.5× the cheapestGPT-6 Astra · ChatGPTOpus 5.5 · ClaudeGPT-6 Luna · APIGPT-6.1 Sol · APISonnet 5.5 · APIHaiku 5.5 · APIGemini 3.8 Flash · APIGemini 3.5 Flash-Lite · Gemini
Nobody beats these on both score and cost
Task score and relative cost by setup
SetupTask scoreRelative cost (cheapest = 1)
GPT-6 Luna · API901.0×
Haiku 5.5 · API823.3×
Gemini 3.5 Flash-Lite · Gemini683.5×
Gemini 3.8 Flash · API7711×
GPT-6.1 Sol · API8416×
Sonnet 5.5 · API8446×
Opus 5.5 · Claude9175×
GPT-6 Astra · ChatGPT10081×

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

About the task

The PM job

Taking a product or add-on to market with marketing and sales.

Why it matters

Launch plans fail by trying to reach everyone, by selling upside nobody feels, and by treating launch day as the whole plan.

What good looks like

  • One primary segment: those who value the difference most
  • Positions against what customers do today
  • Sells against a pain the buyer feels, with evidence
  • Diagnoses whether the problem is the message or the channel
  • Plans beyond launch week

Deliberately not measured

  • Creative copywriting quality
Capability tested

Go-to-market judgement

The failure we’re looking for

Every channel, every segment, and nothing after launch day

Grading

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