Why We Still Hand-Build a CMS Instead of Reaching for Shopify or Webflow
Shopify, Webflow and their equivalents are the right answer for most client briefs, and we say so more often than clients probably expect from a studio that also builds custom platforms. The interesting cases are the ones where they are not the right answer, and two of our own projects, Auto Barn and the Gun Trader rebuild, explain why better than a general rule would.
Auto Barn is a car storage service with a listings-driven site: vehicles, availability, pricing tiers, all managed through a CMS built specifically around how that business actually operates day to day. The default instinct for a listings site is to reach for an off-the-shelf commerce or content platform and bend it into shape. We tried that path first, on paper, and the bending was the problem. Off-the-shelf platforms model a generic version of product and listing that fits most retail businesses well and fits a niche storage service's actual workflow, availability by bay, pricing that varies by vehicle size and storage duration, awkwardly. A hand-built CMS let the data model match the business instead of the other way round.
Gun Trader was a different flavour of the same problem, at a much bigger scale. Rebuilding the UK's biggest gun marketplace meant an ecommerce platform carrying real transaction volume, real regulatory requirements around what can and cannot be listed, and search and filtering behaviour specific to a category no general platform is built to handle well out of the box. At that scale, the cost of fighting a platform's assumptions every sprint outweighs the cost of building the right foundation once.
The test we actually use is simple: does the platform's data model match the business, or does the business have to be described in the platform's terms instead. Most businesses fit comfortably inside Shopify's idea of a product or Webflow's idea of a page, and forcing a custom build on a brief like that is expensive vanity, not craft. A niche storage service's availability rules or a regulated marketplace's listing requirements do not fit, and pretending otherwise just moves the cost from build time to every sprint afterwards, chasing a workaround for a decision the platform made for you.
Production-grade custom work is more expensive up front than configuring an off-the-shelf platform, every time, no exceptions. It only earns that cost back when a business's actual workflow does not fit anywhere close to what the platform assumes. Knowing which side of that line a brief sits on, before a single screen gets designed, is worth more than any framework choice that follows it.