Download websites that use Centra

About Centra

Headless commerce is designed to be invisible from the front. The shop a customer sees is a custom application built by an agency or an in-house team, and the platform underneath delivers products, prices, stock and orders over an interface instead of rendering pages. Anyone trying to identify the platform from outside is looking at somebody else's frontend code, which is rather the point of buying it.

Centra sells into that model in the Ecommerce space, aimed at brands running direct sales and wholesale from the same catalogue across several markets and currencies. Apparel and lifestyle labels are the audience the vendor addresses, which shows in what the platform treats as a serious problem: size and colour variants, seasonal drops, returns, and the same item priced differently by territory. The commercial profile that implies is specific. Multi-market pricing and wholesale ordering are problems that arrive with scale, so the merchants involved are past their opening storefront and have a development budget attached.

The name is shared with unrelated products and organisations, so it is worth stating what our figure counts. A record here is a fingerprint match against the commerce platform, made from checkout endpoints, asset hosts and script signatures, not against anything that happens to use the same word. As of October 2026 that match appears on 14 sites.

The count should be read as conservative, for the same reason the architecture is attractive in the first place. A decoupled storefront can hide its backend almost completely, and the traces we rely on may be proxied behind the brand's own domain or bundled into a build where the vendor's fingerprint disappears. Sold under a Proprietary licence since 2012, the platform has had time to accumulate customers who are simply not visible from the public web. The honest position is that a detection is reliable evidence of presence while an absence proves very little.

For a researcher the value sits in the joins rather than in any single row. Checking one brand's stack takes a browser. Assembling the population of brands on the same headless platform, then sorting by country suffix, by the payment and delivery services detected beside it, and by when each record was last confirmed, is what an export makes possible. Integration vendors use that to find where their connector already has a home. Competing platforms use it to find brands whose architecture makes a migration realistic. Both questions are about a set of domains, and a set is not something a per-domain lookup can hand anyone.