Glossary
What is a conversion in marketing?
The action you count, how to choose it, and the counting rules that keep the number honest.
Definition: In marketing, a conversion is a visitor completing the action a page exists to produce: a purchase, a sign-up, a form submission, a download, a click on a key button. A macro conversion is the end goal, such as an order; a micro conversion is a step towards it, such as adding to the cart. An A/B test measures one conversion goal per test.
Macro and micro conversions
The macro conversion is the outcome the business is paid for: an order, a signed contract, a booked demo. Micro conversions are the steps on the way: viewing a product, adding to cart, starting checkout, opening the pricing page. Macro conversions are rarer and take longer to accumulate; micro conversions are frequent but further from the money.
Choosing the goal for a test
Pick the conversion closest to the money that still happens often enough to measure in a few weeks. On a page with 100 orders a month, the order is the right goal for a bold change but too rare to detect a modest one; add-to-cart, at 600 a month, may be the workable compromise. Whatever you choose, it must be caused by the change you are testing: a headline test on the home page should not be judged on checkout completions three pages later.
Counting rules
Count each visitor once per test, however many times they convert. Count the conversion only for visitors who saw a variant, and count it within a window that makes sense for the action. Define these rules before the test so the comparison is the same for every variant.
Conversion versus conversion rate
A conversion is the event; the conversion rate is conversions divided by visitors. Tests compare rates, not raw counts, because the variants may not receive exactly equal traffic.
A worked example
A software company's landing page has one macro conversion, the trial sign-up, and several micro conversions: scrolling to the feature list, opening the pricing modal, clicking "Start free trial". A headline test is measured on trial sign-ups, because the headline's job is to get people to sign up. A test of the pricing modal's layout could be measured on clicks of the trial button inside the modal, the nearest action the modal controls.
Common mistakes
- Measuring a test on a goal the tested element does not influence.
- Counting repeat conversions from the same visitor.
- Choosing a micro conversion because it is easy and then declaring a revenue win.
- Changing the goal mid-test, which mixes two different measurements.
In AB Test WP
AB Test WP Free goals: a page visit, an element click, a link click, a form submission (Contact Form 7, WPForms, Gravity Forms, Fluent Forms, Ninja Forms, Formidable, Forminator or a plain HTML form) and a WooCommerce add-to-cart. Pro adds scroll depth, time on page, engaged time and a WooCommerce purchase goal that records the order total. One goal per test; each visitor counts once.
Related: Conversion goals in AB Test WP · What to A/B test first · all glossary terms.
Common questions
What is the difference between a conversion and a lead?
A lead is one kind of conversion: a visitor who gave you a way to contact them. Conversions also include purchases, downloads, bookings and any other action you decide to count.
Should I test on clicks or purchases?
On purchases when they happen often enough to reach the sample size in a few weeks; otherwise on the nearest frequent action the tested element controls, and confirm the winner on purchases later.
Can one visitor convert twice?
They can act twice, but a test should count them once. Counting every repeat inflates the rate of whichever variant a few enthusiastic visitors landed in.
Run the numbers on your own test
Free calculators, no email wall: significance, sample size, conversion rate and SRM.