I ship experiences that move conversion, not just features.

Designed a one-tap quick-filter surface placed directly in the listing flow and shipped it across 12 product listing pages as a controlled test. I scoped the surface against the highest-traffic categories first so the read would be fast and decisive.

Clear conversion and revenue-per-session lift on the primary brand and on mobile for the second brand. The one exception was a desktop regression on one brand — which I treated as a signal to segment, not as a reason to hold the wins.

Built the business case to graduate this from the personalization-tool implementation onto native search infrastructure, so the surface is owned in-platform rather than maintained as a campaign — durable, not a one-off.

Ran a conversational shopping agent at a 50/50 traffic split to 5M+ sessions to prove value before further investment. I synthesized the experience data for both internal leadership and an external partner audience, and analyzed over 1,200 individual conversational sessions to understand intent.

Engagement was excellent where it happened — strong conversion and add-to-bag lift among users who opened it. But banner visibility sat at ~28% and modal open rate near 1%. The experience wasn't the constraint; discoverability of the entry point was.

Reframed the roadmap around the front door — entry-point placement, visibility, and trigger logic — rather than further investment in the agent itself. The lesson I carry: a great feature behind a closed door is a distribution problem, not a product one.

A/B tested an always-visible search field in the mobile nav across both brands, with two variants. I read results per-brand and per-variant rather than in aggregate, because aggregate would have masked a real divergence.

One brand showed a statistically significant unique-conversion lift — clear ship. The second brand showed an add-to-bag decline at high significance, which I would not ignore for the sake of a clean launch. Variant 2 showed no lift on either brand.

Shipped the winner, killed variant 2, and held the second brand pending a session-replay investigation into why the add-to-bag metric moved. Holding to diagnose is a decision, not indecision.

Designed a model where engineering owns the actual rendering on critical surfaces and the personalization platform operates only as a server-side allocation and measurement relay, passing results to analytics. Documented it as a PRD with a bow-tie process diagram so engineering, marketing, and analytics shared one source of truth.

Removes the third-party tool from the transaction-critical render path while preserving the ability to allocate and measure tests — marketers keep their experimentation surface, engineering keeps control of checkout stability.

Rolling the model out MVP-first, proving it on a contained surface before generalizing. The principle: experimentation infrastructure should never be a single point of failure on the path to revenue.