Custom apps · Furniture Brand
Product Configurator App with Real-Time Pricing
A UK furniture brand was stitching together three separate SaaS tools to handle product configuration. The combined cost was £1,200/month — and the tools didn't talk to each other. We replaced all three with a single embedded Shopify app.
- Role
- Architecture and full-stack development
- Stack
- Remix, Admin API, Storefront API, Node.js, PostgreSQL, React
- Conversion uplift
- 28%
- Monthly active users
- 15k
- Tools consolidated
- 3 → 1
The problem
The brand sold modular furniture with hundreds of configuration options — fabric, frame, leg finish, dimensions. Three apps were involved: a configurator that couldn't update prices in real time, a pricing engine that required manual exports, and a quote generator that broke weekly. Customer service was handling 50+ pricing queries per week that should have been self-serve.
The technical challenge
Building a performant configurator inside the Shopify admin iframe meant working within the App Bridge and Polaris constraints while maintaining sub-200ms pricing updates. The Storefront API cart mutations needed to be atomic — a partial configuration should never create a cart line. PostgreSQL stored the configuration rules, with a typed API layer between the React frontend and the database.
How it was built
- 01
Remix embedded app
Built with Shopify's Remix app template, deployed on Fly.io. App Bridge 4 handles authentication and host navigation. Session management via PostgreSQL (not cookies) for reliability at scale.
- 02
Configuration engine
A typed rule engine in PostgreSQL defines valid combinations, pricing deltas, and inventory constraints. The React frontend queries this on every selection change, returning a new price within 80–120ms.
- 03
Storefront API cart
Confirmed configurations are added to cart via a single Storefront API mutation with all variant attributes and custom properties. The mutation is idempotent — duplicate submissions are safely ignored.
- 04
Admin product sync
A webhook listener watches product/variant update events from the Admin API and triggers a re-validation of any saved configurations that reference those products.
Results
Conversion rate on configured products increased 28% within 90 days of launch. Customer service pricing queries dropped to near zero. The £1,200/month in SaaS tools was eliminated. The app reached 15,000 monthly active users within 6 months without any infrastructure scaling events.
Related projects
Have a project like this?
Tell me what you're building and I'll tell you exactly how I'd approach it.