Rachel Taber

Students spend money in the real world. Food, drinks, groceries, nights out. A diary study confirmed it. So we built for where they actually are.

I oversaw the launch of the in-store map, a feature that lets students find discounts near them anywhere in the UK or US. It's been live over a year. No competitor has shipped one.

Post-launch, we found two problems. Students were using a generic in-app ID to redeem in-store, which meant clicks weren't tracked and we weren't getting paid. And at the register, when time actually matters, getting to the right discount took too long. We built the ID Wallet to fix both: a branded, Apple Wallet-style feature that puts the right discount one tap away. Then Account Builder nudged new users to discover both features from day one.

Six weeks after the ID Wallet launched, four major brands left the platform. Performance held anyway. Adjusted for that lost volume, underlying clicks were 18% above the pre-launch peak.

Every user saw the same homepage. Changing it meant a designer and an app release, so it rarely changed. And our brand recommendation algorithm had never been tested against anything better.

I worked on both. The multi-tile rail gave the marketing team a CMS-driven homepage surface they control directly. Content changes in real time, no release required.

For recommendations, I noticed another team was already running AWS Personalize for CRM and pushed to test it on the app. We kept the UI identical. The only variable was the algorithm. It won by a lot: 66% more brand follows, 88% more clicks through to brand pages.

Brand follows correlate directly with retention and revenue per user. That one change moved a metric that's hard to move.

The algorithm is now live for all users. We've expanded the same recipe to brand pages across the app, and it opened up appetite for deeper personalization work, including further AWS recipes and an internal tagging system that will let us match offers to users based on attributes like category, values, and brand type. One test unlocked a roadmap.

Brands partnering with us had no visibility into how their discounts were performing. Any change meant an email to our team. Onboarding new clients was slow and headcount-dependent. It wasn't scalable.

I led the 0-1 build of Pion Portal, our B2B client platform, from discovery through to launch and post-launch growth. We ran workshops with customer success, account management, B2B marketing, and RevOps, then built toward an MVP in 6 months. I led discovery interviews, worked iteratively with design and engineering, built product-led upsell surfaces inside the portal, and ran sales training to drive adoption at launch.

Zapp is a quick commerce app delivering groceries in minutes. Users could find what they needed, but they didn't know what they wanted. Discovery interviews kept surfacing the same thing: browsing a grocery list isn't inspiring. People wanted ideas, not just items.

We built Collections, a curated showcase of product groups that rotated throughout the day. Themes tied to seasons, holidays, current events, and influencer partnerships. Think “Pasta Party” on a Sunday, or a meal plan curated by a creator our users actually followed. Content marketing built into the product. I led the discovery interviews, worked iteratively with design and engineering on prototypes, and ran and analyzed the A/B test.