LFW Experiments

Decided at the edge, not in the browser

A cached page can’t decide anything client-side without a flash of the wrong version first. So the variant is chosen and applied at the Cloudflare edge, before the HTML reaches the visitor: no flicker, no JavaScript deciding after the fact, and no fight with your host’s page cache.

Assignment is sticky: the same visitor gets the same version every time. Your site stays on Kinsta, Pressable, or WP Engine; only nameservers move.

Judged honestly: “too early to call” is a real answer

Most A/B dashboards always seem to show a winner. This one refuses to: until each variant has been seen by enough people, and enough have done the thing the test is about, the verdict reads too early to call, in words, on the page. When a winner is declared, the gap cleared a standard 95% test you could check by hand.

Measured on your own domain

Exposures and goals travel the same first-party path as LFW Analytics: a beacon on your own domain, relayed server-to-server. No third-party script, no ad-blocker holes in the data, and the collector refuses events for any test that isn’t actually yours and running.

Used on LFW.com

Analytics includes Replay: page-level friction signals, optional masked page-visit replay, and click-density maps. Save a finding and hypothesis there to create an experiment draft here.

Drafts do not change your website. LFW reviews the variants, deployment, and completed goal before starting a test.

Our homepage uses the same experiment registry to compare calls to action. It records which variant was shown and whether a visitor completed the defined goal.

The portal example above uses sample data. A real test needs enough traffic and completed goals before its results can support a decision.

Pricing