The 90-Day Rule — Click Science Ventures

Every idea that enters this studio gets the same deal: ninety days to prove that real people want it, with real tests and real money. Prove it, and we build. Don't, and the idea gets retired, written up, and shelved without ceremony.

No exceptions for ideas we love. Especially no exceptions for ideas we love.

This essay explains the rule, the four phases inside it, what actually counts as proof, and why a clock beats conviction every time. It's the most copied-sounding and least copied thing we do, because the rule is easy to admire and miserable to obey.

Why a clock

First, the uncomfortable premise the rule is built on: your judgment about your own ideas is the least reliable instrument you own. Ours too. Everyone's. The founder who "just knows" the market wants something is the single most common failure pattern in small business, and it fails expensively, because unvalidated conviction doesn't fail fast. It fails after the build, after the launch, after the year of pushing. Conviction is a loan against future evidence, and most of those loans default.

The standard fix is research. Surveys, interviews, competitive decks. We do some of it. But research has a dirty secret: it produces opinions about demand, and opinions about demand are not demand. People are generous in surveys and stingy with credit cards. The only honest instrument is the market itself, asked directly, with something at stake.

So the rule does two jobs at once. The tests-with-real-spend half forces every idea to face the market instead of our enthusiasm. The ninety-days half forces a verdict. Without a deadline, validation becomes a hobby. You can test forever, reading ambiguity as encouragement, moving the goalposts a little every month. The clock makes ambiguity a verdict too: if ninety days of honest effort can't surface clear demand, that absence is the data.

Why ninety, and not thirty or a year? Because the number has to be long enough to run real tests through real channels, search takes weeks to read, paid tests need iterations, and short enough that the cost of being wrong stays trivial. Ninety days of one operator's attention is a cost we can pay dozens of times a year. A wrong build is a cost we can't. The exact number matters less than its fixedness. A deadline you can extend isn't a deadline. It's a mood.

One honest carve-out: the clock measures demand proof, not bureaucratic reality. A venture that needs licensing or regulatory setup can prove its demand inside ninety days while the paperwork runs its own clock. We hold the demand question and the paperwork question separately, because confusing them lets the slowest process excuse the weakest evidence.

The four phases

Inside the ninety days, the work runs in four phases. The ledger version, which also lives on our Studio page:

Days 1 to 10: Signal check. Before we spend a dollar, the idea has to show up in instruments we trust. Search volume with history and trajectory, not a spike. People asking for the thing in public, in their own words, in communities we can name. Marketplace gaps where buyers are visibly making do with bad options. We're hunting for evidence that demand already exists, because we don't invent demand, we find it. Most ideas die right here, in the first ten days, at a cost of approximately nothing. That's the system working.

Days 11 to 45: Live tests. Now the idea faces strangers. A real offer, in market: landing pages with honest copy, small paid campaigns, posts where the audience actually lives. We're not measuring applause. We're measuring costly actions: signups with a working email, waitlist joins that survive a confirmation step, preorders where the model allows it. The defining property of this phase is that every test is something a real person did with something at stake, even if the stake is just their attention and inbox.

Days 46 to 80: Economics. Demand existing isn't enough. Demand has to be reachable at a price that works. This phase asks the unglamorous questions: what does it cost to acquire one of these customers through a channel we can actually scale? What will they pay, and how often? Does the math hold without a paid-acquisition budget the business can't support? An idea can pass phase two and die here, and several of ours have: real demand, unreachable economics. That's not a sad ending. That's a cheap one.

Days 81 to 90: The call. The last ten days are for the decision, made against criteria we wrote down before the test started. This sequencing matters more than anything else in the rule. Deciding what proof looks like in advance is the only defense against the most seductive failure mode in validation: redefining success after the results arrive. The call has two outcomes. Build, which means the idea earns a place on the stack and a real launch. Or retire, which means a written log entry, what we tested, what we saw, why it wasn't enough, and a clean shelf. Retired ideas aren't deleted. Demand shifts, and a logged idea with old test data is an asset.

What counts as proof

The standards, stated plainly, because vague standards are how clocks get cheated:

Proof is behavioral, never verbal. What people did, not what they said they'd do.

Proof is costly. The person gave up something: money, a real email, a slot in their day. Free applause is excluded by design.

Proof is attributable. We can trace the signal to a channel we could actually operate at scale. Demand we can't reach repeatably is trivia.

And proof is priced. Somewhere in the ninety days, evidence has to appear that the unit economics can clear, fast, without a fund subsidizing the gap. Built for cash flow isn't a slogan we apply after launch. It's a test criterion.

What doesn't count, no matter how good it feels: encouraging conversations, competitor existence ("the market is validated!"), press interest, and our own mounting excitement. Especially that last one. Excitement compounds inside a team exactly like interest on debt, and the rule exists to mark that debt to market every ninety days.

The portfolio is what survived

Here's what the rule produces over time, and why we'd keep it even if it were twice as annoying.

It produces a true portfolio. Every venture on our ledger passed the same public test, which means the portfolio needs no narrative defense. When someone asks why we believe in a venture, the answer isn't a vision statement. It's a file of dated evidence.

It produces cheap mistakes. We've retired far more ideas than we've built, and the total cost of every retirement combined is less than one bad build would have been. In an asset class with no fund behind it, that arithmetic isn't a preference. It's survival.

It produces speed where speed is safe. Counterintuitively, the rule makes us faster, not slower. Most studios drift because every idea stays half-alive, splitting attention forever. A hard verdict every ninety days clears the deck. The ideas that pass get an undivided build. The ones that don't stop costing attention completely.

And it produces something harder to name: institutional honesty. A studio that has to retire its own ideas in writing, on a clock, cannot sustain the small self-deceptions that kill small companies slowly. The rule isn't really about validation. It's about making truth-telling structural, so it doesn't depend on anyone's mood, including ours.

The deeper logic of the studio runs through this rule. Cheap building, the thing that created the small-cap internet, has a shadow side: it makes it cheap to build the wrong thing, and the wrong thing's true cost was never the build. It's the year of operating attention that follows. The 90-Day Rule is how we ration the only resource that doesn't compound back: time on the wrong idea.

Every idea gets ninety days. The portfolio is what survived.

This is the small-cap internet. We operate here.

Click Science Ventures is a bootstrapped micro venture studio in Fishers, Indiana. Built for cash flow, not fundraising.