What actually counts as traction for a startup?

Build an hour · Evidence in hand: two to four weeks

The question underneath the question

A founder asks whether ten customer interviews count as traction. Then whether an MVP counts. Then whether one pilot commitment counts. Coaches field this question more than any other, and no answer ever satisfies, because it is not the real question. The real question is whether they can raise money yet.

Four other questions arrive with it, wearing different clothes.

The first is opinion-seeking dressed as validation. One founder describes working through friends, mentors, strangers in startup forums, and finally his own mother, asking whether the idea was good and whether they would buy it. He came away with contradictory answers, a folder of feedback, and no momentum.

The second is fear of building the wrong thing. Most first-time founders build before they validate, spend months and money doing it, and arrive at the raise without the traction that would have funded it. The framing that lands is before you write a line of code.

The third is the traction gate. Investors ask about traction before they will engage, so founders show up wanting a number rather than an insight. "How do I get my first hundred users" is the same anxiety with a target attached.

The fourth is willingness to pay. Everyone knows to ask it. Almost nobody knows how to test it.

All four have the same answer, and the rest of this is that answer in the form you can execute. The exact fields, the scoring, where to stand it up, and who to send it to. An afternoon to build, two to four weeks to have evidence in hand. That is faster than the next round of coffees you were going to schedule anyway.

How do you get evidence of traction quickly and cheaply?

You stop asking people what they think and start recording what they do. That is the whole method, and everything below is the mechanics of it.

Here is why the distinction is worth the trouble. Collecting opinions until you feel safe is not validation. It is outsourcing belief to people with no stake in the outcome, and they will say yes to be polite, because the social cost of discouraging you is higher than the social cost of being wrong. Then watch what an investor does with those conversations. They ask about traction, and immediately test the quality of it. Is it in the bank. Is it recurring. Is it from people you didn't already know. Encouraging conversations survive none of those three. All hat, no cattle.

Behavior survives all three, and you can capture behavior before a product exists. Somebody choosing one urgency band over another, naming a price, or telling you what they already pay for is doing something, on a date, under their own name.

So the better question is not how much traction is enough. It is what does my traction prove. You do not have a traction problem. You have an evidence problem.

A rookie goes all in and finds out in month nine. Run the cheap experiment first.

What the six categories are for

Every question in the instrument belongs to one of six categories, and each category exists to defend a specific claim later.

Contact. Name and email. This is what turns a count into a list. Anonymous survey results prove nothing to an investor because nobody in them can be called back, and a claim nobody can check is an opinion again.

Segmentation. Who answered. Without it you cannot separate the operators who have the problem from the investors and the merely curious, and encouraging noise from the second group will quietly inflate everything else.

Problem and urgency. What is broken and how much it costs them right now. Urgency is what separates a real problem from an interesting one, which is why it carries the most weight in the score.

Solution fit. What they do about it today. Somebody maintaining a spreadsheet has already spent hours; somebody paying a competitor has already spent money. That is behavior that has already happened, which is stronger than anything they say about the future.

Qualification. Budget and timeline, captured as bands. Authority is deliberately out of scope here. You are measuring demand, not closing a deal, and asking who signs makes a diagnostic feel like a sales call.

Open and commitment. The free text, and the door to a next step. This is where the deposit-grade evidence starts, and it is the only field where you learn something you didn't think to ask.

The instrument

Two lead fields, then six questions. Lead fields go first so partial completions still convert.

1. Name. Short text.
2. Email. Validated. Unvalidated email is how a list becomes uncontactable.
3. Which best describes you? Single-select radio: founder, operator, investor, curious. Unscored. Drives branching and list tags.
4. What problem does this solve for you today? Single-select from four to six named pains, plus "other" free text. Single-select, not multi. Multi-select destroys the scoring.
5. How are you handling it now? Single-select: nothing, manual, spreadsheet, a competitor, built our own. "Nothing" and "built our own" are the two extremes worth knowing.
6. How urgent is this? Single-select, ordinal, three or four bands. Weighted heaviest.
7. What would you pay? Single-select price bands. Never open text. Open text gets you a number nobody committed to.
8. Anything we're missing? Optional long text, unscored.

Only questions 4 through 7 carry points. That score is the readiness tier, and the tier drives both the results page and the segments.

Six questions, not thirty. This is a signal capture, not a survey. Every question past six costs completions and buys nothing you will act on.

How to field it

Building the instrument is the easy half. Twenty responses is the bar, and getting twenty of the right people to answer is where most of these die in a drawer.

Three ways to stand it up. A page on a site you control, which is the only option where the list is unambiguously yours. A scoring tool, which is the fastest: ScoreApp is built for exactly this shape, handles the branching, scoring, and segments without you writing anything, and has a free tier with paid plans priced by monthly responses. Or a plain form builder, if you are willing to do the scoring yourself in a spreadsheet.

Distribution is the part that actually decides whether this works. Do not broadcast it. Send it to twenty named people who match your pain list, one message at a time, and ask them directly. A form posted to a general audience collects curious people, and curious people inflate the count while proving nothing, which is the entire reason question 3 exists.

After that, the cheapest distribution available to a founder with no audience is somebody else's. A borrowed list, a newsletter, a network that already trusts the person handing you over.

What you end up with

A dated list of named people, each carrying an urgency signal and a price signal. On a slide it reads: 214 respondents since March 3, 41 in the top urgency band, 29 of those at $200 a month or higher, and none of them investors. That is countable, dated, attributable, and segmented. It is not a pile of nice emails.

It also answers the question the founder actually asked. Ten interviews do not count. Forty-one named operators who told you on a date that this is urgent and priced it do count, and they count exactly as much as the next thing you do with them.

The honest caveat

Stated willingness to pay is weaker evidence than a deposit. Say so before an investor says it to you, because volunteering the weakness in your own evidence is the fastest way to be believed about the rest of it. You strengthen it three ways: a small pre-order, a paid pilot, or a scheduled call the person showed up for. Attendance is behavior. Intent is not.

Keeping the list alive

A dead list is one that hears from you at launch. Write to the top tier every two weeks with what you built and what you got wrong. An engaged list writes back, and what it writes back is more evidence.

Next action

Build the eight fields today. Send it to twenty people who match your pain list, and date the file. Give it two weeks, then send it to twenty more.