← Validate a Startup IdeaLibrary Log in Sign up free

Spoke · Updated May 2026 · 7 min read

Validate before you build

A pre-MVP checklist before you write a line of code. 7 GO signals, 5 WAIT signals, and a 7-day plan to find them.

Short answer

Should I validate before writing code?

Yes, because a fast MVP usually is not fast. The median solo-founder MVP ships in three to six months rather than three to six weeks, and if the idea was wrong you spent a quarter of the year learning it.

Score the idea in nine questions →

The short version

The median solo-founder MVP ships in 3 to 6 MONTHS, not 3 to 6 weeks
The trade is not validation against speed. It is a week spent learning the idea is wrong against six months spent learning the same thing
Momentum on the wrong idea is the most expensive feeling in startups
There are very few genuinely novel ideas. “Nobody can imagine this” almost always means the right customer has not been found yet
Pre-code validation is about what can be falsified WITHOUT a product, not about building a smaller product

Why validate before coding

The most expensive code is the code you write before you know who's going to use it. Every founder eventually learns this — the question is whether you learn it in week one or month six.

Pre-MVP validation isn't a delay. It's the cheapest, fastest filter you have. A week of structured conversations costs you nothing but calendar time — and the answer it gives is more honest than the answer you'll get after you've already shipped a prototype and grown attached to it.

The shift in 2026 isn't "skip validation." It's "compress validation." The same Mom Test–style questions that took 8 weeks of recruiting in 2010 can run in days now — through structured AI sessions, synthetic personas grounded in real public data, and faster outreach tooling. The framework is the same. The cycle time isn't.

7 signals that say build it

Hit 5 of these and the idea is worth building. Hit 7 and you're probably already late to ship.

1. 5+ specific people named

You can name 5+ specific humans (with first/last names or LinkedIn URLs) who have the problem severely enough to have tried solving it.

2. Unprompted second sentence

When you describe the problem, prospects keep talking — adding context, frustrations, workarounds — without you asking another question.

3. Existing budget

Prospects are currently spending money or significant time on workarounds. Money is moving — you're just redirecting it.

4. Tried 2+ alternatives

Real prospects have tried at least two other solutions and abandoned them for specific reasons you can articulate.

5. Recent pain (< 30 days)

The pain happened in the last 30 days, not "back in 2023." Recency = active need.

6. Strong commitment signal

A pre-order, deposit, calendar booking, or LinkedIn intro to their boss who controls budget. Verbal "yes" doesn't count.

7. Repeatable acquisition path

You know specifically where to find the next 100 prospects: a subreddit, a LinkedIn search, a conference attendee list, an industry directory.

5 signals that say wait

These don't mean "kill the idea" — they mean "the idea isn't ready yet." Fix the gap, then re-validate.

⚠ Polite enthusiasm only

"That sounds great!" without follow-up questions or specific stories. Politeness is not a buying signal.

⚠ Hypothetical answers

Prospects answer in the future tense ("I would use it") instead of past tense ("last week I tried..."). Hypothetical = no signal.

⚠ No existing spend

Nobody is currently paying for an alternative — even a workaround. Markets without existing spend are 10× harder to monetize.

⚠ You can't describe the buyer

When asked "who specifically?", you say "small business owners" or "founders." Too broad = no ICP = no go-to-market.

⚠ Can't name 100 prospects

You don't know where the next 100 customers come from after the first 10. No path = no business.

The 7-day plan

A concrete week-long sprint that ends with a GO/WAIT/NO-GO decision. No coding required.

Day 1

Voice intake + problem hypothesis

Run a 15-minute voice session with GoNoGo (or write a 1-page problem hypothesis). Output: a falsifiable claim about who has the problem and how badly.

Day 2–3

Find 30 ICP prospects

LinkedIn Sales Nav, niche subreddit, Slack community, conference attendee list. Build a 30-row spreadsheet: name, role, contact, why they'd care.

Day 4–5

Cold outreach + first 10 conversations

Message 30, accept 25 will ignore. The 5 who reply are the signal. Run 30-min conversations using Mom Test–style questions about past behavior.

Day 6

Ask for commitment

For the prospects who lit up, ask for the strong signal: deposit, calendar slot, intro to boss. Anything less is a soft no.

Day 7

GO/WAIT/NO-GO decision

Score against the 7 GO signals. 5+ = build. 3–4 = WAIT and re-validate the gap. 0–2 = NO-GO, pivot the customer or problem before reattempting.

Compress the week to 15 minutes

GoNoGo runs the framework above as a single 15-minute voice session: problem hypothesis, market check, competitor scan, synthetic focus group, GO/NO-GO score with reasoning. It doesn't replace the 10 real conversations on Day 4–5 — it replaces the weeks of throat-clearing that come before them.

Run my 30-min validation →

Free · No credit card · up to 20 reports

Frequently asked questions

Why validate before coding instead of just shipping a fast MVP?+

Because a "fast" MVP is rarely fast. The median solo founder MVP ships in 3–6 months, not 3–6 weeks. If the idea was wrong, you just spent a quarter of your year learning that. Pre-code validation costs days and answers the same question — without the sunk-cost bias that makes founders defend bad ideas after they've already built them.

Doesn't this slow down momentum?+

Momentum on the wrong idea is the most expensive feeling in startups. The right framing isn't "validation vs speed" — it's "spend a week learning the idea is wrong, or spend six months learning it." Validation that takes 5–7 days is faster than any MVP and gives you a higher-confidence answer than "people clicked the landing page."

What if my idea is so novel that nobody can validate it?+

There are very few genuinely novel ideas. "Nobody can imagine this" almost always means "I haven't found the right ICP yet." If you truly have an idea so novel that the target customer can't evaluate it, you have a market-creation problem, not a validation problem — and those need 10× the runway of a normal startup.