Pick the elements Stellar can change and the goal to improve. Autopilot runs one A/B test after another, keeps every winner and builds the next test on top of it.
Illustrative example.
One element at a time, one winner at a time. Every test starts from the best version you have.
Enter your page URL. Stellar finds your headline, main button and hero copy. You choose which ones Autopilot can touch and which goal it should move.
A new version of one element, with a clear hypothesis, in your page's language and without invented claims.
The challenger runs against your champion until one of them clearly converts better. If it loses, nothing changes on your page.
Winners join the champion. The next test builds on top of it and learns from every previous result.
Most teams run a test, ship the winner and stop. Autopilot keeps going. Each test changes a single element on top of your best version, so every result is clear and every gain stays.
Autopilot never touches an element you did not select. Every test is a regular Stellar experiment you can open, edit and stop, with its full results.
We email you each new test. Tweak the copy in the visual editor and launch it when you are happy with it.
Each test starts as soon as the previous one ends. No gaps and no backlog waiting on a meeting.
One click stops Autopilot from deciding or launching anything new.
See every challenger, its hypothesis and its result, and how your champion changed over time.
Nobody is watching when Autopilot calls a winner, so the rules are strict. Bayesian statistics, minimum sample sizes and a full weekly cycle before anything changes on your page.
Set up Autopilot in a few minutes. Your first challenger is ready before you finish your coffee.