Letting Customers Design the Product, Then Sending the Factory What It Needs

The challenge

The concept had been proven, but a proof of concept and a commercial engine are different things. Customers needed to configure a layout themselves and see it; the resulting order needed to reach the storefront correctly priced; and the factory needed specifications precise enough to manufacture from, not a picture of what the customer wanted.

Where we came in

The risk in a configurator project is an ambiguous handoff — the builder produces something beautiful and manufacturing cannot act on it. We set the division of responsibility explicitly at the start: our work ends when validated production data and assets land on the order, and the client's fulfilment picks up from there. Naming that line in the proposal is what stops it becoming a dispute later.

What we did

• Opened with an integration blueprint phase to confirm the storefront embedding method and product mapping before building
• Built the layout engine and the customer-facing builder application as the core deliverable
• Implemented pricing, asset handling and visual generation as its own phase
• Automated the paid-order path so a completed purchase produces validated production data and metadata on the order
• Closed with a launch hardening phase rather than treating go-live as the finish line

Where it landed

• Customers configure their own layout and buy it through the existing storefront
• A paid order carries production-ready specifications and assets, not just a customer's intent
• Responsibility for the engine and for fulfilment was defined in writing before development started