Milestone map
Milestone map
3 milestones
Problem interviews done
2–3 weeks
Conduct at least 5 structured problem interviews with people who match your target customer profile. A problem interview is NOT a pitch — you are not describing your solution; you are asking about the customer's current life, their frustrations, and what they have already tried. The interview must explore whether the problem you are planning to solve is real, frequent, and important enough for the customer to pay to fix. Keep a written record of each interview: who you spoke to, what they said about the problem, and how they rated its severity. The key output is an honest assessment of whether the problem is validated — including a clear-eyed view of any evidence that contradicts your assumption.
Proof required
Submit: interview notes or transcripts from at least 5 problem interviews, each identifying the interviewee's role and context (no full names needed), the specific problem statements they described, how they currently address the problem, and their approximate rating of how painful or important it is. Include a 100+ word synthesis assessing whether the problem is validated, partially validated, or invalidated by the interviews.
What gets checked
- At least 5 interviews with distinct individuals who match the target customer profile — 5 interviews with the same person, with close friends or family who will say yes to anything, or with people who do not match the customer profile do not pass; the interviewee's role or context must be described for each
- Interview notes capture the customer's own language, not the founder's framing — a note that says 'they confirmed the problem exists' is a summary; a note that says 'they described spending 2 hours per week manually reconciling spreadsheets and called it their biggest operational headache' is evidence
- Synthesis must include any contradictory evidence — if even one interviewee said the problem is not serious, not frequent, or already solved by an existing tool, that must appear in the synthesis; a synthesis that only reports positive signals is incomplete
Resources
Foundation: start here · Depth: go deeper · Mastery: for the dedicated
Foundation
Depth
Mastery
What a verifier looks for
- Open the interview notes and check for the customer's own language — phrases like 'they confirmed the problem exists' or 'they agreed it was important' are the founder's summary, not evidence; look for direct quotes or paraphrases of what the interviewee said in their own words
- Check the synthesis for contradictory evidence — ask 'did any interviewee say the problem was not serious, not frequent, or already solved?' If yes, it should be in the synthesis; if the founder cannot recall any contradictory signal from 5 interviews, probe — in genuine customer discovery some interviews always produce unexpected responses
- Verify the interviewees match the target customer profile — ask who the target customer is and how each interviewee was recruited; if all 5 were personal contacts who already know about the founder's idea, the interviews were likely contaminated by courtesy bias; the validation signal is weaker than it appears
Solution validation
2–3 weeks
Test whether customers are interested in YOUR solution, not just the problem. Show at least 5 target customers something concrete — a landing page, a mockup, a prototype, or a live demo — and record their responses. A solution validation is NOT asking 'would you use this?'; that question produces courtesy responses. Instead, observe whether they ask follow-up questions, try to use it, ask about pricing, or ask how to sign up. Those behaviours are signals; words alone are not. The output is a written record of what you showed, who responded how, and what changed in your solution concept as a result of what you learned.
Proof required
Submit: (1) a link or screenshot of the validation artifact (landing page, mockup, prototype, or demo), (2) notes from at least 5 solution conversations documenting who you spoke to and their behavioural response (what they did, asked, or requested — not just what they said they would do), and (3) a 100+ word note on what you changed or decided to cut from the solution based on what you observed.
What gets checked
- Validation artifact exists and is linked or shown — a description of what you plan to build does not pass; a Figma mockup, a live landing page URL, or a video of a working prototype does
- Response notes capture behaviour, not promises — 'they said they would use it' does not pass; 'they immediately asked whether it integrated with Slack and tried to click the sign-up button' does; verbal commitments without behavioural signals are weak evidence
- Something changed in the solution as a result — if the solution concept after milestone 2 is identical to the concept before, either the validation produced no useful signal (which itself is a finding worth reporting) or the founder did not listen; at minimum, name one feature that was added, cut, or changed based on what customers showed you
Resources
Foundation
Depth
Mastery
What a verifier looks for
- Open the validation artifact — a landing page URL, Figma link, or prototype video; if the artifact does not exist or is not shown, milestone 2 has not happened regardless of what the notes say
- Check that response notes record behaviour, not promises — 'they said they would use it' is not enough; ask 'what did they actually do during the demo?' or 'did anyone ask how to sign up or ask about pricing?'; those are the signals that matter
- Verify something changed in the solution — if the founder says the validation confirmed everything they already thought, ask 'what was the most surprising thing a customer told you?' A genuine validation always surfaces at least one surprise; if nothing surprised the founder, either the interviews were not genuinely exploratory or the responses are being filtered
Pre-sales confirmed
1–2 weeks
Secure at least one genuine pre-sale or letter of intent from a target customer BEFORE building the full product. A pre-sale is not 'they said they would pay when it's ready' — that is a courtesy commitment that costs nothing and predicts nothing. A genuine pre-sale has at least one of the following: a financial deposit (any amount), a signed letter of intent, a purchase order, or an email explicitly committing to a specific price and timeline in writing. The distinction is commitment cost: a customer who has paid $1 is more predictive of a real customer than a customer who said 'sounds great, keep me posted.'
Proof required
Submit: (1) evidence of at least one genuine pre-sale — a Stripe payment screenshot, a PayPal receipt, a signed letter of intent, or an email stating 'I will pay $X for Y by Z date', and (2) a 100+ word account of how the pre-sale was secured — what you offered, what you asked for, and how the customer responded.
What gets checked
- Pre-sale has a financial or written commitment — verbal commitments and 'interested' signups do not pass; the evidence must show money moving or a name on a document with a specific price and delivery expectation
- Pre-sale is from a genuine target customer — a family member pre-paying as a favour, a friend who pre-buys to be supportive, or the founder themselves 'pre-buying' does not pass; the customer must match the target customer profile defined in milestone 1
- Account of how the pre-sale was secured is specific — 'I asked and they agreed' does not pass; 'I offered a founding member discount at $29/month versus a planned $49/month, explained what the first version would include, and they paid immediately via Stripe' does; the account must reveal enough about the conversation to assess whether the commitment is real
Resources
Foundation
Depth
Mastery
What a verifier looks for
- Open the pre-sale evidence — a Stripe payment screenshot, PayPal receipt, signed LOI, or email commitment; if the only evidence is email signups or verbal promises, milestone 3 has not been completed
- Verify the customer is a genuine target — ask who the pre-sale customer is and how they were recruited; if the customer is a friend, family member, or anyone with a personal relationship that creates social obligation to support the founder, the pre-sale signal is weak
- Check the price tested matches the planned price — if the founder plans to charge $99/month but the pre-sale was at $5, the validation gap is significant; ask 'what price did you quote, and how does that compare to the price you plan to charge at launch?'