Most e-commerce design advice assumes a fairly simple product: a t-shirt in three sizes, a chair in two colours. The patterns are well established, filter by price, sort by popularity, add to cart. But that pattern breaks down fast the moment your product has real technical depth and your buyer isn't a casual shopper.
I ran into this designing a B2B catalog for a diamond and jewellery wholesaler. The buyers were not browsing for fun. They were jewellers and traders sourcing stock against very specific criteria, often for a client waiting on the other end. That single fact changed almost every design decision.
The buyer isn't browsing, they're evaluating
A B2C shopper can be nudged, inspired, upsold. A B2B buyer sourcing diamonds already knows exactly what they need before they open the site: a certain carat range, a specific cut, a colour and clarity grade, a certification body. Discovery isn't the goal. Confirmation is.
That reframes the whole homepage and navigation problem. Instead of designing to inspire browsing, I designed to get an expert user to their filtered result set as fast as possible, with as few assumptions made on their behalf as possible.
Filters aren't a feature, they're the product
In a consumer store, filters are a convenience layered on top of the catalog. In a technical B2B catalog, filters are effectively the interface. If the filtering logic doesn't match how the domain actually works, the product fails no matter how polished the visuals are.
Filter design for technical catalogs
- Use the domain's own vocabulary and units, not a generic e-commerce filter pattern
- Support ranges, not just single values, wherever the underlying attribute is continuous (carat, price, dimensions)
- Let filters combine freely rather than forcing a rigid category tree first
- Show result counts per filter option before the user commits, so they don't filter into an empty result set
- Persist filter state across sessions, professional buyers often return to refine the same search over days
None of this is exotic UX theory. It's just what happens when you take the domain seriously instead of reaching for a template built for browsing sneakers.
Product data has to be a first-class citizen
For a physical, highly variable product like a diamond, the specification is the product, in a way that's rarely true for consumer goods. Certification, grading, provenance, these aren't secondary details tucked into a tab. They needed to sit on the product card itself, visible before a single click.
I ended up treating the product detail page less like a marketing page and more like a technical data sheet with strong visual hierarchy: image first, core specs immediately beside it, everything else below the fold. Buyers were scanning for disqualifying details as much as they were looking for a reason to say yes.
"In a technical B2B catalog, the spec sheet isn't supporting content. It's the product page."
Search needs to speak the buyer's language
Generic e-commerce search assumes fuzzy, forgiving, natural-language queries. Professional buyers search differently, they often type exact codes, grading shorthand, or industry abbreviations they already know cold. Search that "helpfully" corrects or broadens those queries actively gets in the way.
The fix wasn't clever autocomplete. It was making sure the underlying data model and search index respected the domain's own shorthand, and getting out of the expert user's way.
Checkout is rarely just checkout
Consumer checkout optimises for speed and minimal friction: fewer fields, guest checkout, one-click pay. In this project, the equivalent flow was closer to a formal request or inquiry process. Trust, account history, and sometimes negotiation mattered more than shaving off form fields.
- Account-first, not guest-first — repeat professional buyers expect to be recognised, with saved preferences and order history
- Request over instant purchase — for high-value, variable items, a "request this item" flow can be more appropriate than a hard add-to-cart
- Clear next steps — B2B buyers want to know exactly what happens after they submit, who follows up, and when
What I'd tell any designer starting a B2B catalog
Don't start from a consumer e-commerce template and strip things down. Start from how the domain expert actually works today, on a spreadsheet, over email, on the phone, and design the digital version of that expertise, not a simplified version of a t-shirt store.
The visuals still matter. A clean, professional interface builds trust with buyers who move real money through it. But the visuals are the last layer, not the first decision. Get the filtering logic, the data hierarchy, and the request flow right, and the rest of the design gets a lot easier to get right too.
