Duplicate an experiment

Re-running a test on another template, another market or another season means rebuilding a setup that already exists. Duplicating copies all of it: the variants, every change on each variant, the screenshots, and the URL, device and cookie targeting rules.

Copying targeting is the part that matters most, because it is what people forget when they rebuild by hand, and a test that quietly runs on the wrong pages is worse than one that does not run at all.

The copy starts clean where it should. It is a draft, so nothing is live until you start it. It carries none of the original traffic or results, and it gets its own activity log rather than inheriting the history of what was done to the original.

The same mechanism backs test experiments, which are duplicates marked as tests so a rehearsal never contaminates the reporting of the real one.