BigCommerce sells itself as open SaaS, and the catalog backs the claim: 65 distinct APIs — the largest pure-commerce surface in the index. What makes it worth profiling isn’t the count, though. It’s the deliberate split between the management surface and the storefront surface, which is the cleanest example of headless commerce architecture in the catalog.
What’s actually in the surface
The 65 APIs sort into a few clear groups:
- Catalog —
Products,Brands,Categories,Category Trees,Product Variants,Variant Options,Product Modifiers. The merchandising core, decomposed to the attribute level. - Cart and checkout —
Carts,Checkouts,Orders,Abandoned Carts,Abandoned Cart Emails. The conversion funnel as discrete, addressable resources. - Customers and pricing —
Customers,Customer Login (SSO),Price Lists,Pricing,Currencies,Subscribers. - Payments, shipping, tax —
Payment Processing,Payment Access Token,Shipping,Shipping Providers,Tax Classes,Tax Rates & Zones,Tax Provider. Each commerce-operations concern is its own contract. - Storefront APIs —
Storefront Carts,Storefront Checkouts,Storefront Customers,Storefront Orders,Storefront Subscriptions,Storefront Token,Cookie Consent. A parallel, browser-safe surface. - Extensibility —
Webhooks V3,Scripts,Widgets,Themes,Channels,Content,Pages.
What’s interesting about the shape
- Two planes, by trust boundary. The management APIs (server-side, full-access tokens) and the Storefront APIs (browser-side, scoped tokens) are separate surfaces for the same nouns. A
Cartexists in both — but the storefront version is built to be called from untrusted client code. That’s the headless split made explicit in the contract, not just the docs. - Channels make it multi-surface. The
ChannelsAPI treats a web storefront, a marketplace listing, and a POS as peers. Commerce isn’t assumed to happen on one front end. - Tax is over-decomposed on purpose. Seven separate tax APIs look like fragmentation until you remember tax is jurisdictional and pluggable —
Tax Provider Connectionlets a third-party engine slot in cleanly.
The takeaway
BigCommerce is the catalog’s best argument that “headless” is an API-design decision before it’s a marketing word. The management/storefront split by trust boundary is the pattern any platform with both server and browser consumers should copy. Walk all 65 on the BigCommerce provider page — and notice how rarely a single resource is forced to serve two very different callers.