Shift4Shop

The problem

3dcart, now Shift4Shop, was an ecommerce platform with roughly 20,000 merchants running their stores on it. At that scale, platform-level constraints multiply: every limitation in the templating system was a limitation on twenty thousand storefronts.

And the templating system had real limitations. Design constraints baked in years earlier capped what themes could express, which meant designers kept hitting walls the platform couldn’t justify. Meanwhile the admin panel, the tool every merchant used every day to run their business, was aging badly: built on legacy code, not responsive, and impossible to take offline for a rewrite. You can’t tell 20,000 businesses their back office is closed for renovation.

The approach

I joined as a web developer building themes, and spent six years working up to leading the front-end team. Two projects anchor the story.

The templating engine. I architected and built an enhanced common core templating engine that removed the design constraints inhibiting 3dcart templates. It became the foundation the theme ecosystem was built on: themes that had been impossible to express became normal, for the platform’s own theme catalog and for every merchant customizing a store.

The admin panel. Rather than a big-bang rewrite, I pioneered a dual-version strategy: two versions of the admin panel running concurrently, with the differences kept nearly invisible to merchants. That made a phased modernization possible. Sections moved to the responsive rebuild incrementally, legacy behavior stayed supported throughout, and no merchant’s daily workflow broke along the way.

Around those anchors, I shipped the merchant-facing tools that expanded what the platform could do (a page builder, theme editor, mega menu, and delivery calendar among them) and fixed the ecosystem’s craft problems: base templates plus a custom SASS framework that gave theme developers a structured starting point (cutting QA revision cycles dramatically), and version control across the free and premium theme catalog.

The other half of the job was the team: sprint planning, task delegation, interviewing and hiring front-end developers, and giving directors straight strategic input on stabilizing and improving profitability across the company’s multi-vertical sites. Earlier in my tenure I also led the responsive overhaul of the company’s own 1,000+ page website, a project with a history of failed attempts, by segmenting it into phases that could actually ship.

The outcome

The templating engine lifted the design ceiling for every store on the platform and became the base for years of theme development after it. The admin panel got modernized without the rewrite cliff: phased, reversible, invisible to the merchants who depended on it. And the front-end team came out larger, faster, and more consistent than I found it: structured starting points instead of blank files, version control instead of folder copies, and a hiring bar I’d helped set.