Retail storefront / Mobile navigation experiment

Knowing when to keep the original experience

A contextual Back to Results link for mobile shoppers, tested against Control.

The problem

A shopper opening a product from a collection or search result can lose the context they were browsing. The proposed mobile Back to Results link aimed to give them a clear route back, retaining useful collection or search state where possible.

Direct product visits needed a different treatment. A return link with no valid browsing origin would be misleading, so the feature had to handle absent or expired context as well as the happy path.

My contribution

I prepared the design-stage handover, built the mobile variant, addressed timing and expiry issues, documented the staging journeys and handed the work through QA. I also corrected an experiment activation mistake after a colleague found that the test was not running.

I worked with the CRO strategist on the proposed journey and with colleagues on QA. The strategist analysed the test, while a colleague handled a later correction to the mobile targeting.

How I approached it

I preserved the original experience as Control and built the return link for the intended mobile journey. The code checked for a valid collection or search origin, kept the relevant query state and hid the link when it had no useful destination.

I worked through refresh and expiry behaviour and the timing of experiment assignment. I checked collection and search returns, keyboard and touch operation, desktop hiding and the delayed reveal of the selected experience.

The solution

The implemented variant provided a contextual return path for qualifying mobile product-page visits. It remained hidden for direct visits without a valid origin, and the original desktop experience remained outside the intended variant scope.

After a colleague found that the experiment wasn’t running, I corrected its activation. A later review also led to a mobile-targeting adjustment. Those setup issues are part of the context for interpreting the test.

The outcome

The strategist recommended keeping the original experience after reviewing the test. The team paused the experiment and retained Control.