Headless commerce
A store split into a front end and an engine joined by an API: what it gives, what it adds and when it pays off.
E-commerceUsing since 2018
- Category
- E-commerce
- Using since
- 2018
In practice
Headless commerce pays off when the storefront has to behave differently than a theme allows, or when one engine has to serve several sales channels. In a simple store it is one more system to maintain, with nothing to offset it, so I advise against it there.
What is Headless commerce?
In headless commerce the storefront and the commerce engine are two separate systems joined by an API. The engine holds the catalogue and orders. The front end, what the customer sees, is a separate app in any stack, Next.js for example. The opposite is a platform where front and back end run together and you change things with themes and apps.
You gain freedom in the front end and in which systems the store connects to, a CMS or an ERP for instance. You take on a second system to maintain, separate deployments and developers after launch. A headless front takes more work to build and run than a theme, and Shopify does not hide it.
How I decide on headless
I don't begin by asking whether to go headless. I begin with what a ready-made platform cannot do. A second system needs a concrete list behind it.
Needs beyond a theme
I list what a theme and apps will not cover, such as a custom cart or several sales channels. If the list is empty, I advise against headless.
Who runs two systems
I ask who will look after the front end and who the engine once the store is live. Without an answer headless turns into debt: every change touches two places.
Choosing the engine
Only then do I pick the engine. I check whether Shopify is enough as the back end or whether an open engine such as Medusa.js is needed. Content goes into Payload CMS.
When it fits, and when it doesn't
Headless adds work, so I count how much of it you win back.
A good fit when
- A front end a theme cannot deliver, with a custom cart for example.
- Selling across several channels from one engine: a site, a mobile app or a buy button.
- CMS content assembled on one page together with the catalogue.
- A development team that stays with the store after launch.
Not a fit when
- A small store with a standard buying process.
- Nobody will maintain the front end after launch.
- A launch in a few weeks on a small budget.
In the remaining cases a ready-made platform is enough, usually Shopify with a theme and apps. For custom store rules rather than a custom front end, see the Medusa.js entry.
Frequently asked questions
Let us check whether headless makes sense for you
List what your current store cannot do and who will look after it once it is live. From that I will say whether to separate the front end from the engine, and show what my [[service:ecommerce|e-commerce]] offer looks like.