Championing Rive at Fetch, then building the systems it unlocked

Company: Fetch

Role: Product Design Leadership, Experience Strategy, Interactive Systems, Cross-functional Advocacy

Fetch had looked at Rive before and set it aside. I restarted that conversation through a contact of my own, then spent the next year turning a stalled idea into the standard for how Fetch builds its most personal product moments: experiences that respond to a user's own data instead of playing the same way for everyone. Along the way I became the person other designers and engineers turn to for it, built the data and documentation standards that made it safe to depend on, and shipped features now tied directly to company priorities, including a referral experience that drove a confirmed 4.4% lift in converted referrals during a company-wide growth push, and a location prompt that lifted click-through by 64%.

The Challenge

Fetch's animation ran entirely on Lottie, a format that plays a fixed sequence and can't respond to a user's own data or state. For a rewards app whose best moments are personal, points earned, progress made, an offer that's actually relevant, that's a real ceiling: the animation could never show someone their own numbers.

Rive could solve that, and Fetch already knew it in theory. It just hadn't moved. Early attempts to convert existing Lottie files into Rive often performed worse than the originals, which made Rive look like a downgrade rather than an upgrade. Nobody owned it end to end, so every project restarted the same learning curve, and with no defined handoff between the animation file and the mobile code, building anything truly data-driven was risky.

The opportunity was real, and so was the reason it kept stalling.

The Approach

I restarted the conversation myself, using a contact of mine at Rive to get their team in a room with Fetch's engineering, product, and creative leaders, and staying close to engineering leadership afterward to work through the technical questions a bet like this raises. From there I made myself the answer to the problem that had stalled it before: I became the person who owns Rive end to end, the one other designers learn it from and the one engineers loop in before they build.

I also solved the trap that had made Rive look like a downgrade. Rather than converting existing Lottie files, which is what had performed badly before, I rebuilt experiences natively in Rive, matching the timing and detail of the original by hand. That distinction, native versus converted, is what turned Rive from a risk into something worth building the company's future on.

To make it safe to depend on, I created a data-contract standard for every handoff: what each property means, how it should be tested, and what "done" looks like, so integrating a Rive experience became a checklist instead of a guessing game. I wrote that standard, along with everything else designers and engineers need to build with Rive, into a 23-page documentation space that's now the company's source of truth for it.

With the foundation in place, I moved from building most of it myself to scaling it through other people, mentoring designers and engineers so the practice could outlast any one project, including mine.

The Outcome

Every result maps back to a goal a team was already being measured on. A data-driven location prompt lifted click-through 64%, feeding a nearby-stores experience meant to drive opt-in. A referral experience I designed and built in Rive drove a confirmed 4.4% lift in converted referrals, a 34% lift in signups, and a 255% lift in activation, during a company-wide push on growth. And an earnings-reveal experience built to make users feel the value they'd earned lifted engagement and click-through across the surfaces it touched, the same story retention is built on.

That track record is why Fetch chose to move its celebration moments to Rive company-wide, with the legacy Lottie celebration service now being phased out, and why the documentation I built is the standard other designers and engineers build from today.

“A tool only earns its place in the stack once it's tied to the metrics the business already cares about.”