Validate a SaaS Idea
Validate a SaaS idea before investing months building it.
Most SaaS ideas fail quietly because founders build before they test. This Cookbook runs a structured validation sprint: sharpen the problem, map the buyer, stress the market, rank your riskiest assumptions, and design experiments that produce real evidence. Enter what you actually know today once. Every step is told not to invent facts, market sizes, or interview quotes. What you approve carries forward until a final memo forces an honest go, revise, or stop decision — before you write production code.
- Made for
- Founders and small teams validating a B2B SaaS idea before building
- You leave with
- validation brief, ICP, competitor map, assumption stress test, MVP scope, experiment plan, and go/revise/stop memo
- Time
- about 180 minutes
- Steps
- 12 guided steps
- Requires
- any AI you already use (a free chat account is usually enough); sample answers included where available
- Completed runs
- 312 (demo metric)
- Average rating
- 4.7 / 5 (demo metric)
Also in Start Here Selected by SousMeow Recently Added Deep Workflows
Have an account? Sign in.
Stages
5 stages in sequence-
12 steps
Frame the Bet
Sharpen the problem and map how often it hurts.
-
22 steps
Know the Buyer
Define who buys and how you will reach them.
-
32 steps
Map the Market
Place competitors and articulate your wedge.
-
43 steps
Stress the Assumptions
Rank risks, gather evidence rules, and set kill criteria.
-
53 steps
Decide and Test
Scope the MVP, plan experiments, and record the decision.
Workflow steps
12 steps · completed in orderStage 1: Frame the Bet
-
1
Problem Validation Brief
Sharpen the problem, buyer pain, and the validation questions worth answering.
Every later Recipe quotes this brief as ground truth. A vague problem statement is why founders run twelve interviews and still cannot explain what they are testing.
- Problem is specific to a moment You can picture the Tuesday when reconciliation breaks, not a generic "inefficiency" claim.
- Founder insight stays anecdotal Personal stories are labeled as yours, not presented as industry statistics.
- Questions are falsifiable Each validation question could fail based on interview answers.
- No invented TAM or metrics Nothing claims market size, churn rates, or survey percentages not in your Pantry.
-
2
Pain Frequency Snapshot
Map when the pain fires, how often, and what evidence you still need.
A painful problem that fires once a year is a different business than one that fires every month-end. This snapshot stops you from building for imaginary urgency.
- Month-end close appears if relevant For finance/ops pains, the snapshot names when in the calendar the pain spikes.
- Gaps are named honestly Missing invoice counts or hours are listed as gaps, not filled with plausible numbers.
- Table rows match your workaround Trigger moments trace to steps in your Pantry workaround, not generic SaaS pains.
- Interview prompts are usable verbatim You could paste each prompt into a Calendly call agenda without editing.
Stage 2: Know the Buyer
-
3
Ideal Customer Profile
Define who is in-segment, who is out, and how they recognize the problem.
Broad ICPs produce noisy interviews. Tight disqualifiers save weeks of conversations with merchants who will never pay for reconciliation software.
- Revenue band matches Pantry ICP honors the buyer range you entered, not a broader "all Shopify stores" definition.
- Disqualifiers are concrete At least one disqualifier names a segment you might wrongly chase (e.g., pre-revenue, EDI-heavy).
- Trigger scene is recognizable Ops/finance lead would say "that is my month-end" after reading the day-in-the-life section.
- Competitors inform stack context ICP mentions realistic adjacent tools from your competitor list, not invented logos.
-
4
Buyer Access Plan
Plan how to reach in-segment buyers for interviews and pilots on your primary channel.
The best validation brief is useless if every conversation is with friends or wrong-fit founders. This plan turns your reach channel into a calendar.
- Channel matches Pantry selection Primary path uses the reach channel you picked, not a random growth hack.
- Tone is diagnostic Outreach asks about workflows, not "book a demo" spam.
- Calendar fits decision deadline Two-week sprint plus follow-ups can repeat before your decision date.
- Success signal is restated measurably Precursor metrics tie to 5 paid pilots or your exact Pantry success signal.
Stage 3: Map the Market
-
5
Competitor Landscape
Map alternatives from your list and place where the idea sits in the stack.
Founders often discover they are building a feature of an incumbent, not a company. Honest placement now prevents a six-month misfire.
- Every Pantry competitor appears Each line from known_competitors has a row, including manual bookkeeper/workaround.
- Overlap vs gap is honest Bookkeep/A2X-style tools are credited for what they actually do before stating the gap.
- Unknowns section exists Market share and pricing are flagged unverified, not stated as facts.
- Category name fits Shopify ops pain Placement reads as supplier invoice reconciliation, not generic "fintech."
-
6
Differentiation Hypothesis
State the wedge, why incumbents miss it, and how you will falsify the claim.
"Better UX" is not differentiation. A falsifiable wedge lets you test whether buyers see invoice-to-order matching as a separate job from bookkeeping.
- Wedge mentions invoice-to-order matching Core claim reflects supplier reconciliation, not generic automation.
- Tests could fail Each falsification test has a clear fail condition, not a vanity metric.
- Price ROI is framed as hypothesis $79/month (or your price) is tied to buyer math to verify, not asserted as validated WTP.
- Incumbent bullets cite landscape gaps Why Bookkeep/A2X/Ramp miss the wedge references the competitor table, not strawmen.
Stage 4: Stress the Assumptions
-
7
Riskiest Assumptions Register
Rank the assumptions that would kill the idea if wrong.
Building tests the riskiest assumption first is cheaper than building the whole MVP. This register decides what evidence matters most.
- Feasibility assumptions included At least one row covers invoice matching or data feasibility, not only "people want this."
- Top 3 map to kill scenarios Each priority assumption says what happens if it fails (kill, revise, or narrow).
- Founder anecdote flagged $14K duplicate story or similar is listed as unverified anecdote, not customer proof.
- Channel assumption present Reach channel from Pantry appears as testable row with risk rationale.
-
8
Assumption Evidence Matrix
Pair each assumption with evidence you have, need, and pass/fail thresholds.
Assumptions without pass/fail thresholds drift forever. This matrix is the scoreboard for your validation sprint.
- Every ranked assumption has a row Matrix covers the full register, not just top 3.
- Pass and fail thresholds differ Each row has distinct pass vs fail, not vague "learn more."
- Evidence we have cites Pantry only Current evidence column does not claim interviews you have not run.
- Priorities link to access plan 10-day actions mention outreach or teardown calls from your buyer access plan.
-
9
Kill Criteria
Define stop, revise, and continue rules before optimism bias sets in.
Kill criteria written after failed experiments are rationalizations. Writing them now protects months of runway.
- Stop rules cite matrix fails At least three stop rules mirror fail thresholds (volume, match rate, WTP).
- Revise paths are actionable Each revise rule names a concrete pivot (segment, price, channel), not "keep learning."
- Continue rules reference success signal Green light includes your Pantry success signal verbatim or equivalent.
- Deadline hygiene is explicit Decision deadline triggers revise if cells are empty, not automatic go.
Stage 5: Decide and Test
-
10
MVP Scope Recommendation
Recommend the smallest build that tests feasibility and willingness to pay.
Over-scoped MVPs validate nothing except your ability to ship late. This scope ties every feature to a ranked assumption.
- In scope tests match feasibility Core MVP proves invoice ingest and line matching, not full AP automation.
- Out of scope lists tempting traps Bookkeep sync, EDI, and payments are explicitly deferred if not in v1.
- Concierge bias is explicit Plan allows human-in-loop matching before automated scale.
- Eng gated on kill criteria Build commitment waits on manual match test passing continue rules.
-
11
Validation Experiment Plan
Sequence interviews, tests, and pilots against your decision deadline.
Unscheduled validation slips until "after this feature." A dated experiment plan is how advanced founders treat discovery like delivery.
- Four experiments cover demand and feasibility Plan includes interviews, manual match test, outreach, and paid concierge pilot.
- Timeline ends on decision deadline Milestones anchor to your Pantry decision date, not vague "soon."
- Outreach experiment cites access plan Channel tactics match buyer access plan, not new channels.
- Success metrics tie to kill criteria Experiment pass/fail uses thresholds from kill criteria and matrix.
-
12
Go / Revise / Stop Memo
Synthesize evidence and force an honest go, revise, or stop decision.
The memo is the deliverable investors and co-founders actually want: a decision with criteria, not another brainstorm doc.
- Recommendation matches evidence honesty If experiments are unrun, memo defaults to REVISE, not false GO.
- All three decision paths defined GO, REVISE, and STOP each have triggers tied to kill criteria.
- Success signal appears in GO path Continue option references 5 paid pilots or your exact Pantry success signal.
- Signed decision block is present Memo ends with checkboxes, owner, and decision deadline date.
Your information
10 details, answered once- Idea name Working name for the product concept. Recipes use it as a label, not a claim it exists in market.
- Problem statement What goes wrong today, for whom, and what it costs them in time or money. Two to four honest sentences.
- Target buyer Role, company type, and revenue band if known. Say what you are guessing.
- Current workaround What they do today instead of using your product. Include tools and manual steps.
- Founder insight Why you believe this problem is real. Personal experience, customer conversations, or domain exposure. Label anecdotes as yours, not market facts.
- Known competitors One name per line. Include do-nothing and manual processes if relevant. Recipes will not invent competitors beyond this list.
- Price hypothesis (USD/month) What you would charge if the product existed. Recipes treat this as a hypothesis to test, not validated WTP.
- Primary reach channel Where you will find buyers for validation interviews and pilots. The access plan Recipe builds from this choice.
- Success signal One measurable outcome that would prove validation succeeded. Be specific enough to falsify.
- Decision deadline When you will force a go/revise/stop call. Recipes anchor experiment timelines to this date.
Finished files
What you export when every step is approved
One Markdown file per approved step, plus kit.html (opens in any browser, offline)
and a README of what you approved.
kit.htmlOffline HTML reader for every approved step01-problem-validation-brief.mdProblem Validation Brief02-pain-frequency-snapshot.mdPain Frequency Snapshot03-ideal-customer-profile.mdIdeal Customer Profile04-buyer-access-plan.mdBuyer Access Plan05-competitor-landscape.mdCompetitor Landscape06-differentiation-hypothesis.mdDifferentiation Hypothesis07-riskiest-assumptions-register.mdRiskiest Assumptions Register08-assumption-evidence-matrix.mdAssumption Evidence Matrix09-kill-criteria.mdKill Criteria10-mvp-scope-recommendation.mdMVP Scope Recommendation11-validation-experiment-plan.mdValidation Experiment Plan12-go-revise-stop-memo.mdGo / Revise / Stop MemoREADME.mdManifest of what you approved