The Pre-PMF Validation Checklist:
Prove Demand Before You Build
Indie founder? Pre-PMF? Here's the scorecard. The three signals that actually predict pull, the four traps that look like validation but aren't, and a one-page checklist you can fill out in 15 minutes — before you write a line of code.
What Good Validation Actually Looks Like (And Why Most Founders Skip It)
You're pre-product-market fit. You don't have a team full of analysts. You don't have a market research budget. What you do have is a single founder, a 60-second attention span for any given tactic, and an idea you need to either kill or confirm before you spend the next six months on it.
The temptation is to skip straight to building. Momentum feels productive. Coding feels tangible. But the founders who get to a real business — versus an expensive hobby — do one thing differently: they collect evidence from real, paying-able humans before they ship.
The good news: pre-PMF validation doesn't need a research team. It needs a clear set of signals, an honest scorecard, and the discipline to act on the result. That's what this article is for.
Three Signals That Actually Predict Pull
Most "validation" frameworks list six or ten signals. That makes them feel rigorous. It also makes them impossible to apply when you're one person on a Sunday afternoon. Here are the only three that matter pre-PMF.
Signal #1 — Active problem language
Real demand shows up in the words your target market already uses. Search Reddit, Hacker News, niche Slack groups, support forums, Twitter complaints. If you can't find at least 20 recent public posts where people describe the problem in their own words, your ICP either doesn't exist at the scale you need or doesn't yet care enough to voice it. Both are real answers.
Signal #2 — Current spend or DIY effort
People pay for pain. They also spend hours per week hacking around it with spreadsheets, manual processes, multiple tools duct-taped together. Either is a signal. If your target segment isn't paying for something related to the problem and isn't doing it manually themselves, they likely won't pay for your solution either.
Signal #3 — Unprompted buying intent
The strongest signal pre-PMF is when a stranger asks you a buying question without you prompting them. "How much would it cost?" "When can I try it?" "Does it integrate with X?" These are real. They happen when you share your idea with prospects, not when you pitch them. The distinction matters: don't sell. Share and observe.
If two of three signals are positive, validate deeper. If only one is positive, reframe and re-test. If zero are positive, kill fast.
Four Traps That Fake Validation (Pre-PMF Founders Fall Into All of Them)
The four traps below sound like validation. They aren't. They feel productive because they produce positive signals — but those signals have nothing to do with whether your specific idea has demand.
"Friends and family love it"
They want you to succeed. They won't tell you it's bad. They're also almost certainly not your target customer. Replace this with conversations with five strangers who match your ICP — same outcome, opposite meaning.
"The landing page converted"
Curiosity is not pull. A landing page measures whether your headline was interesting, not whether anyone will pay. It's a useful early signal but never sufficient on its own — especially without a real traffic sample behind it.
"AI said the market is big"
A market research tool that hallucinates ~20% of its data points will tell you the TAM is $10B. It will be confidently wrong. Pre-PMF, what matters isn't TAM — it's whether your specific ICP feels the specific problem acutely enough to act.
"I'd use this myself"
You are not your customer. You're a founder with unlimited context on the idea, infinite sunk-cost optimism, and strong hypothetical-bias ("I'd totally pay for this"). Test with people who have none of those three.
Each trap outputs a signal that feels like validation. None of them measure real demand from your real ICP. Score 0 points for any of them on the checklist below.
The One-Page Pre-PMF Validation Checklist
Print this. Fill it out before you start writing your first feature. It takes 15 minutes if you've done the homework, longer if you haven't — and the homework is the work.
Step 1 — Problem language exists (Signal #1)
- I can find at least 20 recent public posts where real people in my ICP describe this problem in their own words.
- The language they use matches what I'd put on a landing page (without paraphrasing / inventing).
- At least three of these posts are from the last 90 days (i.e. the problem isn't stale).
Step 2 — Current spend or DIY effort (Signal #2)
- People in my target segment are currently paying for something related to the problem — even if it's an inferior tool or a manual workaround.
- I can point to at least three specific competitors, alternatives, or DIY solutions the ICP uses today.
- I can estimate what they spend (money or hours) per month on the workaround.
Step 3 — Unprompted buying intent (Signal #3)
- I've talked to at least 10 ICP-fit strangers about the problem (not the solution).
- At least 2 of them asked a buying question without me prompting it.
- I've sent the message I'd actually send (cold, personalized, problem-led) — not a survey.
A score of 3 doesn't guarantee success — nothing pre-build does. But it dramatically improves your odds because every "yes" is grounded in a signal from a real human outside your head.
What to Do After the Checklist
If you scored 3 of 3: build the smallest thing a real ICP would pay for and ship it to the ten strangers you already talked to. Use the Demand Discovery framework to keep scoring demand as you ship.
If you scored 2 of 3: you need real outreach, not more research. The third signal — unprompted buying intent — only comes from conversations with strangers who don't know you. You can run those manually: identify 30 ICP-fit prospects, send problem-led cold messages, log every reply, and grade the signal. Or you can run them through a system that does this for you.
If you scored 1 of 3 or 0 of 3: don't build yet. Reframe the problem, switch ICPs, or move on. A pre-PMF kill is an 18-month and $50k+ shortcut, not a failure.
The cheapest next step is to run the same scoring on a structured demand report for your idea — let AI map the language, the spend, and the active buyers before you spend weeks doing it by hand.
Score your idea before you build it
Get a free market research report on your startup idea — Demand Score, competitor landscape, and active problem language in minutes.