The claim, taken apart
You were told Type the number somebody quoted at you.
Their numbers support — Fill in the sentence and this fills itself in.
Somebody quoted you a number. Take it apart.
Fill in the parts of a conversion-rate claim you were actually given; the blanks become the questions to send back. Browser-only — nothing is sent anywhere.
— supported
6 still unstated
Start anywhere. Every part you leave blank becomes a question to send back.
CONVERSION CLAIM CHECK raymondporrello.com/tools/conversion-claim-check Prepared 2026-08-07 This record reflects what you observed and entered. It is not a system audit. Nothing entered in the calculator was sent anywhere or saved by the tool. WHAT YOU ENTERED Claimed rate: Not stated Conversions: Not stated Conversion definition: Not stated Visits: Not stated Visit definition: Not stated Period: Not stated Segment: Not stated WHAT THE VALID FACTS SUPPORT Nothing entered yet. The useful result is already below: the definitions and questions a conversion-rate claim needs. You have answered none of the six things this can check. Six are still open. There is a seventh this page cannot check: whether both sides of a comparison use the same definition. WHAT IS STILL UNSPECIFIED [○] The number on top: Open — How many conversions the rate was built from. [○] What counts as a conversion: Open — The specific event or events on top are not stated. [○] The number on the bottom: Open — How many visits the rate was divided by. [○] What counts as a visit: Open — Sessions, users, and pageviews are different denominators. [○] What period it covers: Open — A rate with no period cannot be compared honestly. [○] Which traffic it describes: Open — All traffic and a filtered segment are different populations. [—] Whether both sides use the same definition: Not answerable here — Only the person who produced both reports can confirm that the numerator, denominator, period, and segment match. For a before-and-after, fill this page in once for each side and compare the two records line by line. QUESTIONS TO SEND BACK 1. What number were you dividing? How many conversions, exactly? 2. What counts as a conversion in that number — form submissions, phone taps, text taps, email taps, completed bookings, or something called a qualified lead? 3. What number were you dividing by? How many visits, exactly? 4. Does “visits” mean sessions, users, or pageviews? Those are three different numbers on the same report. 5. What period does this cover — a week, a month, a quarter, or a year? 6. Is this all traffic, or one segment — organic, paid, mobile, or desktop? 7. If this is a before-and-after or a lift, are both numbers using the same conversion definition, visit definition, period length, and segment? LIMITS - This does not tell you what a good conversion rate is. - This does not estimate leads, sales, revenue, or money lost. - This does not predict what a website change would do. - This does not audit analytics or verify the figures entered. - This does not accuse anyone of lying. - Arithmetic alone does not verify a definition; a clean division can still describe the wrong thing.
There is no useful answer to “is that a good conversion rate?” without the definitions behind it. One fictional dataset produces 0.2% and 11.0% — both arithmetically correct — because the numerator, denominator, period, and segment are different. So this page gives you the arithmetic and the questions instead of a number to compare against.
That refusal is not a dodge. A rate you can defend is worth more than a rate you can look up, and the only honest comparison is your own rate against itself over time, under one written-down definition.
The tool above does the arithmetic and writes the questions. The three sections below are the parts that live in the two source reports, where no calculator can reach them.
The method on paper
What the tool cannot do for you
The app does the arithmetic and writes the questions. These are the parts that live in the two source reports, where no calculator can reach them.
One fictional dataset, four honest rates
Open this to see how far one set of figures can legitimately move.
Northline Cabinetry is fictional. These figures are invented to demonstrate the method, not to describe a typical business.
| Figure | Count |
|---|---|
| Pageviews, full year, all traffic | 61,400 |
| Sessions, full year, all traffic | 18,900 |
| Users, full year, all traffic | 14,200 |
| Quote-form submissions, full year | 132 |
| Organic quote-page sessions, October | 127 |
| Quote-form submissions, October | 14 |
The arithmetic produces four different, accurately labelled statements:
- 0.2% — 132 quote forms ÷ 61,400 pageviews, full year, all traffic.
- 0.9% — 132 quote forms ÷ 14,200 users, full year, all traffic.
- 3.3% — 626 bundled events ÷ 18,900 sessions, full year, all traffic.
- 11.0% — 14 quote forms ÷ 127 organic quote-page sessions, October.
The figures can all be divided correctly and still be unfit for comparison.
Calling 0.9% the “before” and 11.0% the “after” would create the appearance of a dramatic change without establishing that business performance changed. The numerator definition, denominator, period, and segment changed together.
Checking a before-and-after, or a promised lift
Open this if the claim is a lift, a comparison, or a “we got them from X to Y”.
Most claims an owner is handed are comparisons: “you are at 2% and we get clients to 6%,” or “we lifted them 40%.” A comparison is only a comparison when both sides use the same four rules, and that is the one thing no calculator can check for you — it lives in the two source reports.
The procedure is the same as above, run twice:
- Fill the tool in for the “before” number and keep its record.
- Fill it in again for the “after” number and keep that record.
- Put the two records side by side and compare four lines: what counted as a conversion, what counted as a visit, the period length, and the segment.
- If any one of those differs, the two percentages are not a clean before-and-after. They are two different measurements, so the difference cannot be attributed to business change until the definition drift is resolved.
A 40% lift is a claim about two fractions. Ask which two.
What this cannot tell you
Open this before you use anything here in an argument.
- This does not tell you what a good conversion rate is.
- This does not estimate leads, sales, revenue, or money lost.
- This does not predict what a website change would do.
- This does not audit analytics or verify the figures entered.
- This does not accuse anyone of lying.
- Arithmetic alone does not verify a definition; a clean division can still describe the wrong thing.
Nothing typed into the tool is sent or saved. Copy, download, and print all produce a plain-text or paper record in the browser. No result is placed in the page address or a contact link.
For the longer explanation of why published rates disagree, read what a service-business conversion rate actually measures. For the separate problem of analytics totals disagreeing with business records, read why website analytics and lead records do not match.