What to test first
Start with the parts most visitors read before deciding to act:
- Headline — the promise at the top of the page.
- Call-to-action label — for example “Book a call” vs “See available times”.
- The first section — headline, supporting line and button changed together as one version.
- Emphasis — highlighting or bolding a key phrase, or hiding a distracting element.
Test one idea at a time. If a version changes five things and wins, you won't know which change mattered unless you deliberately test them as a bundle.
Pick one goal that reflects real progress
A good goal is the action the page exists for: a form submission, a booking, or reaching the thank-you page. Clicks are useful when the next step happens elsewhere, but a click is not a lead or a sale — Flowvert reports it as a click.
How Flowvert does it
Install the Flowvert script once, select the element in the visual editor, write your alternatives and preview them on the real page. Flowvert splits new visitors between versions, keeps each visitor on the same version, and tracks your goal. See how it works for each step.
When can you trust the result?
Early numbers swing a lot. A version that is ahead after 50 visitors is often behind after 500. Flowvert keeps results labelled preliminary until a minimum sample, a minimum duration and a traffic-split check are all met.
As a rough guide, small pages with low conversion rates need much longer to detect small differences. If your page gets little traffic, test bigger, bolder changes.