-
A mattress catalog specified across five dependent dimensions produces combinations in the millions, with the overwhelming majority not manufacturable.
-
The build evaluates conditions at each selection and closes invalid paths before a buyer can reach them.
-
BigCommerce remained the commerce system while the product logic moved to a rules layer outside the storefront.
-
Configured orders reach the production floor as build instructions carrying the resolved component list and the approved specification.
Custom product configurator development becomes necessary when valid combinations can no longer be stored as variants, and Optimum7 has documented a build at that threshold for a mattress manufacturer whose catalog could not be expressed inside native e-commerce product options. The build evaluates each selection against the manufacturer’s own rules, shows the item in three dimensions as it is configured, and holds a preview and approval step before the order reaches production.
Native product options on Shopify and BigCommerce work well when valid combinations can be represented as stored records. At greater configuration depth, the product logic has to move into a custom rules layer that evaluates them, reconciles inventory and pricing, and sends the resolved configuration downstream to enterprise resource planning or production systems.
Five dependent dimensions define every finished item in the catalog
Nothing in the manufacturer’s catalog exists until an order specifies it. A finished item is defined across five dependent dimensions: core construction, layer composition, cover material, dimensions and firmness. Each selection narrows what can be chosen at the next step.
Modeled as stored variants, the catalog produces combinations in the millions, and the overwhelming majority are not manufacturable. The requirement was therefore a rules layer that decides, at the moment of each selection, which options remain valid.
The build placed the rules layer outside the storefront and kept BigCommerce as the commerce system, in a headless architecture.
Read More: SalesTechStar Interview with Matt Alexander, VP of Channel and Alliances at Synthflow AI
The rules layer evaluates each selection live and closes invalid paths
The build models each dimension and the conditions that govern it, then evaluates those conditions live as the buyer configures. Invalid paths are closed before they can be selected.
The configurator presents a three-dimensional representation of the item as it is specified, so the buyer sees the item taking shape while configuring it. Preview, approval and release steps were built into the ordering process so the specification can be confirmed before the order moves to production.
“The difficulty was never the volume of combinations,” said Duran Inci, CEO and founder of Optimum7. “It was that the rules deciding which combinations were real had never been written down. They lived with the people taking the orders.”
A completed order arrives at production as a build instruction
Each completed order carries the component list the configuration resolved to, the sequence, and the approved specification the customer confirmed. A configurator that stops at a valid specification and hands operations an order number leaves someone on the floor to work out the components by hand.
A separate print build resolves configured orders into press-ready files with the emboss on its own layer
In a separate build for a print manufacturer, a configured order has to resolve into press-ready files carrying the uploaded artwork, a bleed and trim specification, and a die position. When the order includes an embossed panel, the emboss is produced as its own file so the press operator receives an accurate plate. Optimum7 built that file output into the manufacturer’s ordering process.
The same pattern appears in made-to-order furniture, cut-to-size materials, embroidered apparel and configured equipment. Optimum7 has also documented a BigCommerce product configurator for component-based inventory, where the configured unit resolves to its component products and each option is checked against available stock.
Read More: Intent Mesh Salestech: Connecting Buyer Signals Across Every Digital Touchpoint
A custom rules layer is warranted by dependent options, component-level stock and generated production output
Optimum7 builds custom configurators and rules layers when option dimensions depend on one another, configured items draw stock from multiple component records, or production requires generated output. These builds often include an ERP integration to systems such as SAP Business One, NetSuite or Epicor P21, so component availability can be reconciled against the source system.
Catalogs without those conditions are well served by native product options, and Optimum7 does not recommend a custom rules layer for them. Shopify and BigCommerce both support large option and variant counts, and the firm scopes many configurable catalogs to run inside them without custom development.
“Buyers ask us which platform handles this, and that is the wrong first question,” added Inci. “Every catalog at this depth ends up with a rules layer that someone has to own. The decision is who builds it and who maintains it after launch.”
Optimum7 maintains the rules layers it builds after launch, revising the conditions as products, component sources and production requirements change.
