🔎 Formulas

Conversion shortfall tester

Use an exact binomial test to investigate a conversion shortfall without overclaiming.

Run the calculator

Inputs stay in your browser

Events sentnConversions creditedkExpected rate%Your prior period, not the offer card.Significance%Payout per conversion$Optional — for the dollar shortfall.

Not explained by chance

A shortfall this large is unlikely if your true rate were 30.0%. Ask the network to break down the postbacks for this period.

Credited rate26.0%

p-value< 0.001

Wilson 95% CI24.1% – 28.0%

Missing conversions80

Events to be sure789

One-sided exact binomial test of whether the credited rate is below your expected rate. The p-value is P(X ≤ k | n, p0); the interval is a Wilson score band. The verdict never asserts intent — it reports whether the numbers can be told apart from chance and, if not, how much more data would settle it. No default expected rate ships with this tool. Use your own prior period, the same offer on a network you trust, or a per-vertical benchmark once one exists. A p0 pulled from the offer card is the offer card, not your history.

How the calculation works

Built-in figures are examples, not benchmarks

p-value = P(X ≤ credited conversions | sent events, expected conversion rate).

Reading the result

The tool compares credited conversions with your own expected rate and reports statistical significance, a Wilson confidence interval and the sample size needed for a clearer answer.