A proprietary app built from scratch, and a whole base moved onto it

Context

A fashion retailer ran on a white-label app: little personalization, and an e-commerce team with no flexibility to launch campaigns or tune the funnel.

Problem

Every change went through the white-label vendor's roadmap, so the product could not answer to the business. Unable to ship changes, the team had no way to work on retention.

Evidence

• Conversion sat under 1%.
• The team did not use the data it had and events were not mapped correctly, so product arguments were settled by opinion and seniority.
• Behavioral analysis and research, co-led with the product designer, showed that customers habitually reached products through search and filters, and that most used the app one-handed, often on the move.

Hypotheses

• Campaigns and collection pages were not clear, nor segmented in a way that was easy to browse, so searching and filtering was the fastest route to a product.
• The previous white-label app had barely personalized storefronts, and customers learned there to go straight to search. The habit came along to the new app.
• One-handed and on the move, search and filters reach a product with less scrolling than long storefronts.

Metrics

Decision

I settled the hardest question first: what earned a place in version one and what could wait. Then I planned the migration of the entire user base, with the definition of done that nobody was left on the legacy product. Two connected apps, one for the customer and one for the sales team, on a unified data base.

Execution

• Owned the product decisions on a full replacement with a proprietary iOS and Android platform.
• Coordinated delivery across several engineering teams and third-party integrations, plus custom BFF and microservices.
• Instrumented the full funnel with GA4, Firebase, Microsoft Clarity and Power BI, and corrected the event mapping so the numbers meant what they claimed.
• Rebuilt the collection and product pages around observed browsing and comparison behavior, and redesigned the navigation for one-handed use.
• Ran the CRO program, with A/B tests and a checkout rebuilt for returning and first-time buyers.

Outcome

• 100% of the base migrated, with nobody left on the legacy product.
• Conversion from under 1% to 5.0%.
• Customer lifetime value doubled, new customer acquisition +50%, and 42% retention.
• A record Black Friday at 3× its sales target, and H1 2026 alone reaching 107% of the previous full year's revenue.

Learning

The new app only counted once the last customer had left the old one, and it only converted because it kept the path they already knew, search and filters. Today I plan the base migration together with the v1 scope and keep the shortcuts customers already use before changing anything else.