Choose the event you mean
In marketing reports, “conversion” may refer to an email sign-up, checkout completion, qualified inquiry or sale. A provider's percentage is not useful until the event and denominator are clear. For a paid product, separate initial purchases from retained revenue after refunds where those figures are available.
Recorded purchases ÷ eligible visitors × 100.
Recorded purchases attributed to leads ÷ eligible leads × 100.
Account for delay
Some people buy after reading a follow-up email or returning later. Record the attribution window and which system supplies sales data. A platform may not report every downstream event to an affiliate, so explain unknowns rather than treating missing data as zero or manufacturing a profit figure.
Do not mistake precision for certainty
One purchase in a small test can produce a percentage with many decimal places but little predictive power. Include sample sizes and compare several coherent tests before making large budget changes. Examine conversion alongside spend and refunds, not in isolation.
Our ROI calculator models a lead-to-customer assumption and makes the input adjustable.
Specify the event before displaying a percentage
Conversion rate is not one number. You can calculate sign-ups divided by observed visits, purchases divided by valid leads, or purchases divided by visits. Each answers a different question. A provider's claim that traffic “converts at 50%” is uninterpretable until it states which event counted as a conversion, the denominator, sample size and observation period.
An example of the same campaign told three ways
Imagine 200 observed visits, 20 confirmed subscribers and two attributed purchases. The visit-to-lead rate is 10%, lead-to-purchase rate is 10%, and visit-to-purchase rate is 1%. Reporting only “10% conversion” loses the distinction between a subscriber and a customer. If the two purchases are refunded, the gross purchase count and realized-revenue outcome must also be distinguished.
Account for time without inventing attribution
A visitor may subscribe during the send but buy days later. Decide how long you will watch for a downstream event and what evidence links it to the campaign. If you lack reliable attribution, say that instead of assigning all later sales to the solo ad. At small sample sizes, a single extra purchase can create a large percentage swing; avoid declaring stable rates from a handful of events.
Use the rate to diagnose rather than flatter
Strong opt-ins and weak purchases may suggest an offer or follow-up issue. Weak opt-ins with well-reconciled delivery may suggest audience mismatch or a page problem. A discrepancy between a seller's clicks and page visits points to an earlier measurement question. Preserve these stages in the report worksheet and explore plausible ranges in the calculator.
Use practical guardrails for small samples
When a campaign generates one purchase from 100 observed visits, the arithmetic is a 1% visit-to-purchase rate. A second purchase would double that displayed percentage even though the underlying count changes by only one person. Keep the raw numerator and denominator beside every rate and resist extrapolating a stable expected return. A percentage without its counts can make fragile evidence look more certain than it is.
If your business has repeat purchases, separate realized revenue in the defined window from modeled future value. A later purchase may be relevant, but it should not be silently folded into the original reporting period. Record returns and cancellations before using the rate to justify a larger send.
Useful next steps
Guide reviewed September 19, 2026. Provider recommendations and purchase terms can change; confirm current details before ordering.