Case Study
Four Hands is a furniture designer and wholesaler based in Austin, Texas. Their pieces are defined by careful material choices and construction, but the digital experience had fallen behind the work itself. Years of organic growth had left categories and attributes inconsistent and hard to navigate, even as the catalog expanded past 6,000 SKUs, with new products launching every two weeks.
Handsome was brought in to elevate the platform to match the quality of the products. My role was to restructure the content and architecture underneath, so trade customers could find the right piece faster, and Four Hands could merchandise and scale without breaking the system.
The visible design of a catalog is only as good as the structure beneath it. Without a coherent taxonomy, no amount of polish on the UI would make the site easier to use: filters would miss, search would misfire, merchandising would feel arbitrary. The work had to start with the data.
Inconsistent attributes. Years of ad-hoc additions had produced overlapping definitions and uneven naming.
Multiple mental models. Designers shop by type, by room, by material, and by dimension.
Functional and inspirational content. Room-based editorial had to coexist with precise, attribute-driven search.
Two audiences, one system. Customers and consumers needed different feature sets from a single architecture.
I started by documenting the existing site end-to-end, annotating every node with open questions. The messy first draft was a thinking tool, but more importantly an alignment tool. It became the artifact the team used to build consensus across merchandising, marketing, engineering, and customer support, and to break the experience into manageable chunks for sprint planning.
Sitemap v8: alignment, not just architecture
The central decision was to build the catalog around two parallel taxonomies: Products (by furniture type) and Rooms (by context of use). Designers routinely find uses for pieces beyond the room they'd traditionally belong in, like a bench as a coffee table, or a garden stool as a bedside. A dual taxonomy let the same product live in multiple contexts without being duplicated, opening up real room-based merchandising while still supporting the specific, attribute-driven searches trade customers perform every day.
The classification was tested with trade professionals through card-sorting exercises, which validated where designers expected each piece to live and surfaced the edge cases early.
The taxonomy also reached into the visual design. It determined where products appeared and how they were treated: which categories surfaced in the mega menu, which needed an environmental shot versus a silo, which warranted editorial attention on a category page. That turned the taxonomy into a content-planning tool as well, making clear which assets the team needed to create, and in what priority, to support the launch.
Chairs: one record, linked to types, filters, synonyms, and image needs
The taxonomy's job was to disappear into the experience. It became the spine of the global navigation, enabling three ways into the catalog: Products, Rooms, and Inspiration. The mega menu revealed the structure progressively: hovering over a super category surfaced its product categories; selecting one revealed its types.
Progressive disclosure: hovering over categories reveals sub-categories without overwhelming the user
Filters are where taxonomy becomes interactive. Each input type had to match the data beneath it and the way designers actually search: multi-select checkboxes for additive attributes, a dedicated pattern for Color and Material tied to predefined spectrums, single-select radio buttons for sort order.
We split filters into two layers. Global filters (Availability, Color, Price) worked consistently across the catalog, so designers could apply the same constraints no matter where they were browsing. Category-specific filters (Format for Tables, Swivel for Chairs) surfaced only where they were relevant, letting each category expose the attributes its buyers actually care about. The combination improved the experience exponentially: general filters for breadth, specific filters for depth.
Range sliders did real work here. A designer looking for a credenza to fit a 62-inch nook can drag the Width slider and surface only the pieces that will fit, without opening a single product detail page. With more than 20 measurements per piece in the new spec data, filters let customers narrow the catalog to the exact right piece in seconds.
On cards and filter states, my role was to identify every permutation and annotate the logic for the design and engineering teams: what shows logged out versus logged in, how availability interacts with other selections, how voided options grey out rather than disappear.
Filter patterns: input matched to data and intent
Card states: where the logic gets resolved
The Chairs page opens with a row of category types, Dining, Lounge, Office, Side, Recliners, Gliders & Rockers, treated as hero filters at the top of the page. The imagery gives each type a visual identity and adds interest to what would otherwise be a dense grid. Editorial modules break up the grid at intervals, giving merchandising a place to live without disrupting browse flow.
Chairs category page: taxonomy in its final form
The most satisfying part of the engagement wasn't the labels themselves: it was the conversations the labels forced. Every category decision surfaced questions across merchandising, marketing, sales, and customer support about how Four Hands thinks about its own products. Watching a shared vocabulary emerge, and become the backbone of a platform the team could use, was the work.
This project was a true collaboration. Grateful to Charles Law and Alex Poole, and to the team at Handsome, for trusting me to hold the thread across so many stakeholder conversations, and to Four Hands for the openness they brought to questioning their own assumptions.