UserProofUserProof / Funnel calculator
Funnel calculator: see where your signups leak and which step to fix first
For founders with a signup funnel and numbers for each step: in the example below, 31 of 8,000 visitors pay (0.39%). The biggest loss is between the home page and the Sign up click: 7,040 of 8,000 visitors never click. Keeping 10% of them adds about 22.7 paying customers, more than any other step (the last step adds 20.9).
In our scan of 378 YC startup homepages on October 28, 2026, 122 (32%) showed no sign-up, demo or get-started button on the first phone screen. On those sites a phone visitor has nothing to click without scrolling.
Your funnel
Example result
Fix "Visited the home page → Clicked Sign up" first. 7,040 of 8,000 people leave there. If 10% of them stayed, about 22.7 more people would reach "Paid" (31 now, 53.7 then), worth about $909 at $40 each.
31 of 8,000 people made it from the first step to the last: 0.39%.
| Step | People | Kept from the step before | Left here | Extra at the end if 10% of them stay |
|---|---|---|---|---|
| 1. Visited the home page | 8,000 | start | start | start |
| 2. Clicked Sign up | 960 | 12.0% | 7,040 | +22.7 (most) |
| 3. Submitted the sign-up form | 610 | 63.5% | 350 | +1.8 |
| 4. Finished setup | 240 | 39.3% | 370 | +4.8 |
| 5. Paid | 31 | 12.9% | 209 | +20.9 |
How to read the result
- Kept from the step before: the step rate. 610 of 960 people who clicked Sign up submitted the form: 63.5%.
- Left here: people who reached the step before and didn't do this one. The step that loses the most people is not always the one to fix first.
- Extra at the end: how many more people would reach your last step if the share you set (10% by default) of this step's leavers stayed and every later step kept its rate. The step with the most is marked, and the sentence above the table names it.
- Value: if you fill in what one person at the last step is worth (a month of your price, an order's margin), the sentence turns the extra people into dollars.
The ranking says where the most finishers are, not how hard each fix is. Keeping 1 in 10 of the visitors who leave the home page can take one clearer button; keeping 1 in 10 of the trial users who don't pay can take weeks. Weigh the size against the work.
How it works
For each step, the step rate is people at this step ÷ people at the step before, and the overall rate is last step ÷ first step. When a share s of a step's leavers stay, its rate goes from r to r + s × (1 − r). Every later step keeps its rate, so the people at the end grow by end × s × (1 − r) ÷ r.
Worked example, step 2: 960 of 8,000 visitors clicked Sign up, so r = 0.12. With s = 0.10 and 31 at the end: 31 × 0.10 × 0.88 ÷ 0.12 = 22.7. Step 5: 31 of 240 paid, r = 0.129, so 31 × 0.10 × 0.871 ÷ 0.129 = 20.9. Steps 3 and 4 add 1.8 and 4.8.
Why not "raise every step by 10%"? A 10% relative lift at any step multiplies the end of the funnel by 1.1, whichever step it is, so it can't tell you where to start. Keeping a share of the people who leave can.
Where the home page loses people
Some of the loss between the home page and the first click is a page problem you can see. In our scan of 378 YC startup homepages on October 28, 2026, 122 (32%) showed no sign-up, demo or get-started button on the first phone screen. The same scan covered 493 YC startup homepages in all (the report was published from the first 378), and 33 of the 493 (6.7%) had at least one click that did nothing. Check yours with the YC homepage report's 10 checks and the dead click finder.
Limits: what it can't tell you
- Counts have to chain. Each step should count people who also did the step before, in the same date range, from the same tool. Mixing a week of page views with a month of signups gives step rates that mean nothing; the calculator stops when a step has more people than the one before it.
- Why they leave. The numbers say where people leave, not why. For that you watch the people who left at that step: a dead button, a form error, a price they didn't expect.
- Later steps can shift. The math keeps every later step's rate fixed. People you win back at an early step may be less keen than the ones who came through anyway.
Find out why they leave at that step
The calculator shows the step. UserProof records real visits, finds what stops people at that step (dead clicks, rage clicks, errors on click, slow responses) and gives Claude Code, Codex or Cursor the one fix with the file and line, then checks whether it worked. Free for 3 sites and 5,000 recorded visits a month; Pro is $29 a month (pricing).
Related: 378 YC homepages checked on a phone, how common dead clicks are on 493 startup homepages, the session replay cost calculator.
Sources
- UserProof scan of 378 YC startup homepages on a phone and a desktop, October 28, 2026: the report and its method.
- The same scan, all 493 YC startup homepages that loaded, read for dead clicks, October 28, 2026: the report and its method.
- The example funnel is made up to show the math; it is not a benchmark. The formula is the one under "How it works", and the calculator's code is the open file funnel-calc.js.
Cite this
UserProof, "Funnel calculator," https://getuserproof.com/funnel-calculator, updated November 3, 2026. Extra people at the end when a share s of one step's leavers stay: end × s × (1 − r) ÷ r, where r is that step's rate.
FAQ
What is the formula to calculate conversion rate?
Conversion rate is the number of people who did the step divided by the number who could have, times 100. For a funnel, divide each step by the step before it for the step rate, and the last step by the first for the overall rate. In the example, 960 of 8,000 visitors clicked Sign up: 12.0%.
What is a conversion funnel?
The steps a visitor takes from arriving to doing the thing you want, such as visit, click Sign up, submit the form, finish setup, pay. Each step keeps some of the people from the step before; the funnel shows where the rest leave.
Is a 12% conversion rate good?
It depends on the step. 12% of home page visitors clicking Sign up and 12% of trial users paying are different numbers with different norms. We don't publish an average funnel because we haven't measured one across enough sites; compare a step with your own last month instead.
Which funnel step should I fix first?
The one where keeping a few more of its leavers adds the most people at the end. Keeping the same share of leavers adds the most at the step with the lowest step rate, because that step loses the largest share of the people who reached the step before it. The calculator ranks this for you; how hard each fix is stays your call.
Does the calculator store or send my numbers?
No. The calculator runs in your browser, makes no requests and saves nothing. The page itself has the same visit analytics as the rest of this site, and the form and the result are marked so that analytics never records what you type or what the result shows. Reload the page and it starts from the example again.