IC PARTS HUB

Buyers have searched an exact part number because Digi-Key or Mouser came up empty, and the alternative is a costly redesign of their own product. They show up expecting what authorized distributors trained them to expect: live pricing, live stock, instant answers.

The work was to make the experience read as the fastest, most credible path to the part, not as a site that's missing half its information.

The decisions downstream are only as good as the understanding behind them, so I started with the business, not the screens.

• Discovery. A structured intake with the owner and sales lead pinned down the real buyer (procurement engineers at OEMs and contract manufacturers, on desktop), the RFQ process, and the company's genuine edge: reliable sourcing, short lead times, and named human experts.
• Competitive audit. Five competitors reviewed across positioning, RFQ flow, trust, and craft. The gaps were the opportunity: weak trust signals across the board, almost no related-parts or comparison tooling, and US-based identity as a live advantage under 2026 tariff pressure.
• Scope of work. That fed a sitemap, URL scheme, and page-level requirements tied to three business outcomes: discoverability across the catalog, fast quote conversion, and trust through transparency.

Every interface decision that followed traces back to one of those three outcomes.

• Reframe the model as a feature. A 3-step Search → Request Quote → We Source It explainer answers "why is there no price?" before it becomes a reason to leave.
• Designed for paid acquisition. This business runs on Google Ads, so the product page is a paid-traffic landing page first. I designed the PDP as a conversion surface, quote action above the fold, a pre-filled RFQ, and the trust and availability cues a buyer needs before committing, so spend doesn't leak on a slow or ambiguous page.
• Trust on the surface. Named experts, explicit US-based identity, and industry credentials, placed where a procurement engineer checks before submitting to an unfamiliar vendor.
• Desktop-first by choice. The buyer is at a workstation, not a phone. I put the effort into the dense desktop experience first rather than splitting it thin across breakpoints.

A 59,000-part catalog where every part is its own page, and its own potential entry point, is fundamentally a data problem. The job was to make that volume resolve into a single clear path to a quote, and the catalog is still growing, with the team actively adding parts beyond the current count.

• Information architecture built around the exact manufacturer part number as the page's H1, matching how buyers actually search, and how the SEO and paid strategy both land them.
• Dense, filterable category tables for engineers comparing dozens of parts at a glance, not consumer-style cards.
• Pre-filled RFQ at peak intent, part number and manufacturer carried straight from the product page into the form, dropping the buyer's effort to near zero.
• Related parts and a multi-select RFQ tray so buyers can bundle several parts into one quote, closing a gap most competitors left open.

I designed for states, not just screens:

• Availability states treated consistently, so the buyer always knows where a part stands.
• A resilient RFQ form, required versus optional fields tuned to capture leads without deterring them, with email and phone validation and an integrated country-code selector.
• Full interaction states, hover, pressed, focus, and disabled defined across the system, with an accessible keyboard focus treatment.
• The unglamorous cases, empty and truncated content, click-to-expand for long category text, confirmation flows, and routing so a dead part number never strands a buyer.

I hardened all of it through structured QA on the live build, resolving issues as developer-ready checklists and lifting recurring problems into single global fixes rather than patching screen by screen.

A token-based system with a library of reusable components, each defined across its full set of states and built to scale across a growing catalog, so 59,000 pages stay consistent and a non-technical client team can extend the site without breaking it.

I made the front-end and architecture decisions and delivered implementation-ready design, then partnered closely with the development team that built it.

• The platform decision, server-side rendering so every product page is indexable, was central to the design strategy, since organic and paid traffic both depend on it. I designed for that constraint and worked with the developers through implementation.
• A non-technical CMS so the client's team can add parts and edit content without a developer.
• Designed and QA'd against the deployed product across multiple rounds, translating real data models and routing into the UX rather than designing in a vacuum.