Every visit, cached and fresh
Signal source
Audience, stage, and arrangement
Attribution scope
Spectare treats the arrangement of components on a page as a separate variable from the content inside them. Your atoms stay exactly as you wrote them, but the order and combination in which they appear shifts based on what has converted best for each audience and funnel stage. Every time a visitor converts, Spectare records which arrangement they saw, and that signal feeds directly into the next assembly decision.
The learning loop now extends to cached responses as well. Previously, only fresh AI assemblies generated conversion attribution data, which meant the learning signal came from a small and unrepresentative slice of traffic. Now every page load, whether served fresh or from cache, registers an impression against the specific arrangement the visitor saw. Conversions can always be attributed to the exact layout that earned them.
Spectare deliberately skips the winning-arrangement signal when a visitor is assigned to the exploration pool. Exploratory visitors see arrangements the data has not yet evaluated, which lets Spectare discover layouts that would otherwise never be tested. Once those arrangements accumulate enough conversion data to cross your confidence threshold, they can become the new default for future visitors.
You set the exploration rate and the confidence threshold in Settings. Everything else, discovering winners, retiring underperformers, and promoting new layouts, happens without any manual intervention from your team.
This is one atom from the Spectare content library. Spectare assembles the right atoms for each visitor in real time, based on who they are and how they arrived.