Parts That Can't Be Wrong. A Platform Built to Match.
Since 2015, a US boiler parts supplier has run its entire online sales operation through a platform built around one constraint: a wrong part cannot reach a customer's system. Their catalog of 10,000+ components, each with distinct technical specifications, was too complex for any standard e-commerce platform. PlushWire built the compatibility search algorithm together with the client's engineering team. It has been running without a rebuild ever since.
The Client
Based in McDonough, GA, the client is a leading US supplier of boiler industry components: valves, heat exchangers, tube bundles, boiler controls, and steam traps. Their products serve industrial heating and process systems where each component is part of a larger, precision-dependent assembly.
Their customers, maintenance engineers and procurement teams, needed to source exact replacement parts fast, without calling a specialist for every order.
The Problem
Before the platform was built, online sales were not possible.
- xProduct listings could not handle variable specs: different items required completely different technical fields, and no standard platform supported dynamic configuration at the catalog level
- xFinding a compatible part required specialist knowledge - a wrong component could fail to meet load, throughput, or dimensional requirements, with failures appearing only months after installation
- xWhen a required part was out of stock, buyers had no structured way to identify alternatives that would still meet system specifications
- xThe catalog had no online presence: buyers could not search or self-serve without contacting staff directly
- xManual quoting and order processing could not keep pace with a growing catalog and customer base
What We Built
A custom platform built around the technical reality of industrial parts.
The platform addressed every layer of the problem: how to represent technically complex products accurately, how to help buyers find the right part without specialist intervention, and how to connect the store to the client's existing business systems.
Product management built for variable specs
Standard product templates do not accommodate parts where a heat exchanger and a boiler valve share no common attributes. We built a product management system where each item carries its own specification fields - keeping the catalog structured without forcing every product into a fixed template.
- Flexible product cards: specification fields defined per product type, not shared across all categories
- Product variants generated from configurable parameters, not created manually one by one
- Admin tools to manage a catalog of 10,000+ items without engineering involvement
The compatibility search algorithm
The core of the platform: a search and filtering system built in direct collaboration with the client's technical team. It validates compatibility across every relevant parameter simultaneously - not just keywords.
- Parameters checked include tolerances, load ratings, throughput limits, material compatibility, dimensional fit, and expected service life
- When an exact match is unavailable, the algorithm surfaces substitute components whose combined parameters satisfy the original system requirements
- Compatible substitutes surface in search results alongside exact matches, increasing average order value and reducing dead-end searches
Integrated sales operations
The public store connects to the client's existing business systems so orders flow without manual re-entry and shipping costs are quoted automatically at checkout.
- NetSuite integration for order and inventory synchronization across systems
- ShipperHQ for real-time shipping cost calculation at checkout
- Stripe and PayPal for payment processing, with Apple Pay and Google Pay supported at checkout
- JotForms for quote requests on complex or out-of-spec inquiries
The Outcome
An online sales channel for a catalog that could not have one before.
The platform opened an online sales channel for a product category that had resisted e-commerce. A compatibility algorithm built on the client's own technical knowledge reduced ordering errors and average order value grew as buyers found compatible substitutes they would not have located without the algorithm. Manual quote volume dropped as buyers became able to self-serve.
Ongoing partnership
The platform launched in 2015 and is still running. The same team that built the original compatibility algorithm continues to maintain and develop it. Over eleven years, the catalog has grown, integrations have been updated, and the underlying logic has been refined - all without a rebuild.
Selling products too complex for a standard store?
When a wrong order has real consequences, buyers cannot afford to guess. Book a Free Gap Analysis and we will map what it would take to bring your catalog online with the precision your products require.
Book a Free Gap AnalysisSee more case studies