Startup Idea Validation Checklist: 12 Points Before You Build (Free)

Most startup ideas don't die because the founder lacked skill — they die because the founder built the wrong thing confidently. Validation is the process of lowering that confidence until it matches reality, before reality sends you an invoice.

This is the checklist we use. Twelve points, each answerable with evidence rather than optimism. Work through it in order: the early points are cheap to test, and most ideas fail before point 6.

What does it mean to validate a startup idea?

Validating a startup idea means producing evidence, not opinions, that strangers have the problem, want your solution, and will pay for it. A landing page that converts at 5%, ten interviews describing painful workarounds, or three pre-orders are evidence. Compliments from friends, upvotes from other founders, and survey results from people who never pay are not.

CB Insights' running post-mortem dataset — most recently 431 VC-backed shutdowns — keeps pointing at the same root cause: 43% of failures cite poor product–market fit. Every point on this checklist exists to detect that failure mode early, while it still costs an afternoon instead of a year.

The 12-point startup idea validation checklist

Problem (points 1–4)

1. Can you name the specific person who has this problem? Not a demographic — a real persona. "Freelance graphic designers who invoice more than 10 clients a month," not "small businesses." If you can't name the person, you can't find them later to sell to them.

2. Do people already spend money or serious time solving it? Look for existing paid tools, consultants, spreadsheets held together with hope, or manual workarounds. Competition is evidence of demand; a problem with zero existing solutions is usually a problem nobody has.

3. Have you observed the problem firsthand? Ideally you've lived it, watched a friend wrestle with it, or seen it in a community you're part of. Secondhand problems ("I read that doctors struggle with X") need triple the interview effort to validate.

4. Would your target user call this a top-3 frustration? Interview ten people in your target market (these questions keep the answers honest). Ask them to list their three biggest pains in the relevant area. If your problem doesn't come up unprompted, it's not painful enough to build a business on.

Market (points 5–7)

5. Can you size the reachable market in one sentence? "There are roughly 500K freelance designers in English-speaking markets, most reachable through design communities and newsletters." If your market math requires three assumptions stacked on each other, the market is probably imaginary.

6. Is the market growing, stable, or shrinking? Stable or growing is fine. Shrinking markets can work for lifestyle businesses but fight you on pricing and retention. Know which you're walking into.

7. Is the problem frequent or expensive? Products attached to frequent problems (daily/weekly) or expensive ones (thousands of dollars at stake) get used and paid for. Problems that occur once a year and cost $20 are brutally hard to monetize.

Distribution (points 8–10)

8. Can you name one channel where your users already gather? Niche subreddits, Discord servers, industry newsletters, trade shows. "We'll do content marketing" without a specific wedge is a hope, not a channel.

9. Can you reach 100 potential users for under $100? Run the math: niche newsletter sponsorship, targeted ads, direct outreach in communities. If customer acquisition costs more than your expected first-year revenue per customer, the model breaks before it starts.

10. Do you have an unfair advantage in this channel? A personal brand in the niche, a co-founder with an existing audience, professional credibility. Founders without a distribution advantage need to build one deliberately — see our framework for generating ideas from positions of advantage rather than thin air.

Business model (points 11–12)

11. Can you state your pricing in one sentence, and is it defensible? "$49/month because comparable tools charge $40–80 and we save about 5 hours." If you can't justify the number against time saved or money earned, expect a race to the bottom.

12. Does the unit economics work at small scale? At 100 customers, does revenue cover your costs and pay you something? Micro-SaaS ideas pass this at 100 customers; venture-scale ideas don't need to — but you must know which game you're playing. Our micro-SaaS ideas for solo founders breakdown covers the math.

How to score the results

Count your passes honestly — evidence required for each check, or it doesn't count.

The clone test: one extra check if you build with AI

With Claude, Cursor and no-code tools, a solo founder can ship an MVP in a weekend. So can everyone who reads the same thread you did. Before you build, ask one more question the 12 points only imply: if someone shipped a copy of this next weekend, what would they still not have?

Good answers are things an AI-built clone can't reproduce: an audience or channel you already reach (point 9), data that compounds as customers use the product, domain knowledge from years inside the problem (point 4), or a workflow so specific that a generic copy fits nobody. "I'll build it faster" or "mine will have better UX" are not answers. If you can't name one, the idea may still be worth a weekend, but treat it as a learning project, not a business.

Score your idea in 3 minutesOur free AI validator interviews you, researches real competitors, and scores founder fit, Why Now, tarpit risk, and more.

What to do after the checklist

Passing all 12 points means you're cleared for the cheapest real-world test: a one-page landing page describing the product, sent to 100 people from point 9's channel. Signup rate above ~3% is genuine demand; below 1% means the idea or the messaging needs work.

Score a few real candidates with this checklist and you'll want a steady supply of them. Pikari delivers fresh, curated startup ideas every day — matched to your skills, the tools you build with, your budget, and your time, with the problem, audience, and a realistic validation angle included — so there's always a fresh contender to run through points 1–12. You can try it without signing up. Your first week is free, no card required; see what's included.