$160 of media at $0.82 a click
195 of 191 clicks needed
About 195 clicks clears the 191 needed to reach a decision on conversion.
How the test works
Know what the numbers mean.
Build the page with us, put it in front of the right people, and get a report that turns their actions into a clear next move. The page measures interest before the product exists.
How much traffic a question needs
Before you launch, set a stop line for an offer that is not working and a target for one worth pursuing. The distance between those two numbers determines how many visits the question needs.
| Stop line to target | Clicks needed | At this CPC | Media required |
|---|---|---|---|
| 2% 5% | 191 | $0.80 | $153 |
| 2% 5% | 191 | $1.60 | $306 |
| 3% 8% | 104 | $1.40 | $146 |
| 2% 3% | 1,398 | $0.80 | $1,118 |
Closer thresholds need more traffic. The final row shows the trade-off: separating 2% from 3% takes far more visits than separating it from 5%.
Try your own numbers
2% → 5% needs 191 clicks, about $153 of media at $0.80 a click.
Sizing uses 80% statistical power and a one-sided 5% significance level. These settings stay fixed so the budget cannot move the standard after results arrive.
When zero leads means stop
After enough valid traffic, zero leads is a negative result. Thin or failed delivery is no result. For an idea that genuinely converts at 5%, this is the chance that a smaller run still returns nothing:
35.8%chance of zero leads on 20 clicks, for an idea that really converts at 5%
4.6%chance of zero leads on 60 clicks, for an idea that really converts at 5%
0.2%chance of zero leads on 120 clicks, for an idea that really converts at 5%
See the risk
35.8% chance of zero leads after 20 clicks
If the idea really converts at
Why it matters
At 20 clicks, a good idea can still show zero leads. Thin traffic gets insufficient evidence. Enough valid traffic can earn a negative result. Code enforces the line.
Five bands tell you what to do next.
Your report leads with one of five decisions. Evidence strength stays underneath as a caveat about how much the numbers support. Thin evidence never becomes a negative verdict.
- KillStop this idea.The negative result is clear enough to stop spending on this proposition.
- ReviseChange the offer.The tested offer missed the stop line. Change the proposition before you buy more visits.
- HoldDo not build yet.The result is unresolved. Gather the evidence the report asks for, or revise the offer before another run.Current band
- PromisingBuild the smallest version.Interest cleared the target but not the build zone. Use a small build to test whether leads will pay.
- BuildBuild and start charging.Interest cleared the build zone. Use payment and retention to test the business behind it.
Worked exampleMove the numbers
9 leads from 187 clicks
Observed interest was 4.8%, with a 95% interval of 2.6% to 8.9%, against a 2% stop line and a 5% target. Report band: Hold.
Clicks
Try zero leads at 187 clicks, then at 800.
The range shows uncertainty around the measurement. It qualifies the decision; it is not a second verdict scale.
Before you approve a budget
We price the traffic your question needs at its estimated click cost. When you unlock and set a budget, you see whether the budget supports the full question, a narrower reading, or only part of the evidence.
$120 of media at $1.40 a click
85 of 104 clicks needed
Too thin for a conversion verdict. It can still measure reach, engagement by message and the cost of a visit, so that’s the narrower question we report.
$110 of media at $6.00 a click
18 of 191 clicks needed
About 18 clicks is too thin to answer the question. The estimate is about $1,146 of media. Any increase needs your approval.
AlwaysYou see this before you pay
The answer is part of the quote, not a footnote in the report. If no live test can answer the question honestly, we say so before any ad spend.
What this does not tell you
Being straight about the boundary is the product, so here it is in one place.
- Interest is not revenue. We measure whether people click and hand over an email address for your proposition. Nobody is charged, so nothing here proves willingness to pay.
- One page, several messages. Entry budgets can compare message click-through. They usually cannot read landing-page conversion variants.
- One audience, one geography. A result travels no further than the audience it was bought against.
- Paid traffic is cold traffic. It says nothing about what your existing audience, network, or search demand would do.
- A verdict is about this proposition, not your idea for all time. A different offer, price, or audience is a different experiment.
Find out if your idea has demand.
You could have a page in front of people today. Building and publishing are free.
No card needed. $50 when you’re ready for ads and results.