What the Funnel Conversion Analyzer does
This analyser takes the number of people reaching each step of a funnel and returns the step-to-step conversion rate, the overall rate from the first step, and the drop-off at every stage - for one segment or several side by side. It then sizes the opportunity: how many extra final conversions, and how much value, improving each step would be worth.
Counts are pasted or imported from CSV and processed in your browser only.
How to use it
- Paste a header row and one line per step, in order: the step name followed by a count for each segment, for example Step,Desktop,Mobile.
- Optionally add a column named Target % with the step rate you are aiming for at each step.
- Choose how to size opportunities: recover a share of each step's drop-off, or reach the target rates. Enter the value of one final conversion to express it in money.
- Read the bar chart and table for the first segment, then the opportunity table, which is sorted by extra conversions.
- With two or more segments, check the comparison table for steps where a segment differs from the first by more than chance would explain.
Reading the results
The step with the lowest rate is not automatically the best one to work on. The opportunity table compares like with like: the extra final conversions each realistic improvement produces, holding the other steps constant.
Recovering 10% of the drop-off is worth more at a leaky step than at a healthy one: at a 25% step it lifts the rate to 32.5% (a 30% relative gain), at a 90% step only to 91%.
Segment differences carry a p-value from a two-proportion z-test. Strong evidence of a difference tells you where to look; it does not tell you why the segments differ.
Worked example: a four-step trial funnel
10,000 visitors reach the landing page, 4,000 sign up, 1,000 activate and 250 pay. Step rates are 40%, 25% and 25%; overall conversion is 250 / 10,000 = 2.5%.
Recovering 10% of the drop-off at activation raises that step from 25% to 25% + 0.1 x 75% = 32.5%. With every other step unchanged, final conversions rise by 250 x (32.5 / 25 - 1) = 75. The same effort at sign-up (40% to 46%) adds 250 x 0.15 = 37.5. At 40 per paying customer, the activation fix is worth 3,000 a period against 1,500.
Formulas and scoring rules
- Step conversion
step rate_i = count_i / count_(i-1)- Overall conversion
overall_i = count_i / count_0- Drop-off
drop_i = count_(i-1) - count_i; drop rate = 1 - step rate- Recovered drop-off
new rate = r + f x (1 - r)f is the share of drop-off recovered.- Opportunity
extra final conversions = final x (new rate / current rate - 1)Assumes later steps keep their rates.- Segment test
z = (p2 - p1) / sqrt(p(1-p)(1/n1 + 1/n2)), pooled p; two-sided p-valueRates shown to 1 decimal; evidence: strong p < 0.01, moderate p < 0.05.
Limitations: what the result does not prove
- It assumes the steps are nested - everyone at a step passed the one before. Funnels built from separate page-view counts often break this; the tool flags any step larger than the previous one.
- Opportunity sizing holds later steps constant. In practice, extra people pushed through a step are often less motivated and convert a little worse further down.
- Segment p-values are not corrected for the number of steps and segments compared; with many comparisons some will look significant by chance.
- Counts from analytics tools can be affected by consent banners, ad blockers and sampling, so compare trends in one tool rather than mixing sources.
Privacy: where your data goes
Everything you paste, type or drop is processed in this browser tab. It is not uploaded, logged, stored or sent to analytics. Session recording and tag-manager scripts are switched off on this page.
Standards and sources
Frequently asked questions
How do I calculate funnel conversion rate?
Divide the number who complete the final step by the number who entered the first. Step conversion divides each step by the one before it, which shows where people leave.
Which funnel step should I fix first?
Compare the extra final conversions each realistic improvement would produce, not just the lowest rate. Then weigh that against effort; a modest gain at a step you can change cheaply often beats a large gain that needs a rebuild.
What is a normal drop-off rate?
It depends entirely on the product, traffic source and step, so no single benchmark applies. Your own segments and past periods are the most reliable comparison, which is why the tool supports several columns.
Can I compare mobile and desktop funnels?
Yes. Put each segment in its own column. The comparison table tests each step rate against the first segment and labels the strength of evidence.
Why does my funnel show more people at a later step?
Usually because the steps are counted independently (page views rather than the same users), people can enter midway, or the date ranges differ. Fix the export so counts are of the same people moving through in order.
What does the Target % column do?
It sets the step rate you are aiming for at each step. With the target method selected, the opportunity table shows how many extra final conversions reaching each target would produce.
Last reviewed by the A2Z.Tools team against the sources listed above.
