All work

Case Study

ReferMe

Incentivized app-referral platform — share, track, and get paid

Frontend owner + product-to-technical translation2024Mobile (iOS & Android)
React NativeNode.jsMySQLStripeTwilio
Private, client-owned build — source and live app are confidential.

01Problem

Apps that grow by word of mouth rarely reward the people doing the referring, and when they do, tracking who drove what is messy. ReferMe gives users a dashboard to share, track referrals, and get paid — with attribution handled behind the scenes.

02Constraints

  • Referral attribution and payouts had to be trustworthy from day one — real money is involved.
  • The UI had to represent campaigns, stats, and earnings without becoming cluttered.
  • Alpha-first: it needed to reach a small group of real users before any wider release.

03Approach

  • Owned the dashboards, onboarding, and referral-tracking UI.
  • Designed a modular, reusable UI system for campaign cards, stat displays, and earnings — so a new campaign type never meant a new one-off screen.
  • Translated product requirements into technical design and partnered on the database models and REST endpoints powering the referral engine.
Product Thinking

Structure and real feedback drove the layout, not guesswork.

  • Ran a limited alpha with early users and used structured feedback loops to shape layout and CTA decisions before wider release.
  • Chose a modular UI system early because the product’s whole point is running many campaigns — reusability was a product call, not just an engineering one.

04Architecture

Architecture

A React Native client over a Node.js / MySQL backend, with a referral engine at the core and Stripe / Twilio handling payouts and notifications.

  • React Native — iOS and Android
  • Node.js REST API — the referral engine and attribution
  • MySQL — users, campaigns, referrals, earnings
  • Stripe — payout requests
  • Twilio — real-time SMS referral updates

05Decisions & Tradeoffs

Built a reusable component system for campaign / stat / earnings UI up front.

More design effort early, but new campaign types and stats slotted in without rebuilding screens.

Implemented secure Stripe payout requests rather than manual payouts.

More upfront integration and testing, but payouts scale without manual work and stay auditable.

Shipped to a limited alpha before wider release.

Delayed the public launch, but caught layout and CTA problems while they were still cheap to change.

06Outcome

Shipped

  • User-facing dashboards, onboarding, and referral tracking.
  • A modular UI system for campaigns, stats, and earnings.
  • Secure Stripe payout requests and real-time Twilio SMS referral updates, tested end-to-end.