Every headless quote you receive prices the build. Almost none of them price the three years after it, and that is where brands get caught — not by a bad build, but by a good one whose running costs nobody mentioned.
This post puts real numbers against the whole thing. Ours are on the headless service page too; the point here is the shape of the total, which is broadly the same wherever you buy it.
Why headless costs more than a theme at all
On a Liquid theme, Shopify runs the storefront. You rent an enormous amount of engineering — rendering, caching, CDN, checkout, the lot — for your platform fee.
Go headless and you take the storefront back. Your Next.js application renders the pages, you decide what is cached and for how long, and you own everything that used to be Shopify's problem.
That is the trade. You are buying control, and control has a maintenance cost. Anyone who tells you headless is faster because Next.js rather than because of how it is cached is selling you a framework, not an architecture.
The build
Most headless builds with us land between $40,000 and $80,000. Complex programmes — multi-region or B2B storefronts, deep CMS work, heavy custom functionality — run $100,000 and up.
What actually moves the number:
| What moves it | Weight | Why |
|---|---|---|
| Content complexity | Large | A standard catalogue versus a rich editorial site with localised content across markets |
| Custom functionality | Large | Configurators, quizzes and bundle builders are application features, not theme settings |
| Integrations | Medium | ERP, PIM, subscriptions and search each add a workstream |
| Markets | Medium | Every region adds currency, content and routing work |
| App replacements | Varies | Every theme-injected app needs an API integration or a substitute |
Ranges depend on scope, which is why we quote fixed after a discovery call rather than publishing a price for work we have not seen.
That last row is the one most commonly missing from a proposal, and it is covered properly in what breaks when you go headless.
The part nobody quotes
Hosting
A headless storefront runs somewhere. At real traffic that is typically a few hundred dollars a month — more if you are serving many markets or rendering a lot on demand rather than caching.
It is not a large number. It is a number that did not exist before, and it recurs.
API version migrations
Shopify versions the Storefront API and retires old versions on a schedule. Your application has to keep up. This is not optional and it is not rare — it is a recurring, unglamorous maintenance task that somebody is either doing deliberately or discovering when something breaks.
Ask any agency quoting you headless how they handle Storefront API deprecations, and who pays for it. If the question lands as a surprise, the plan does not exist.
Framework upgrades
Next.js moves. Major versions bring changes that are usually manageable and occasionally are not. A codebase left on an old version for three years becomes expensive to touch precisely when you need to touch it.
The integrations you now own
Every app you replaced with a custom integration is now code you maintain. Reviews, loyalty, search, subscriptions — each was somebody else's maintenance problem and is now yours.
Talk it through
Want the three-year number for your store, not a range?
Send your store URL and your installed app list. We will come back with what a headless build would actually involve — including whether your traffic and catalogue justify it, which is frequently the honest answer.
Get a scoped readA real read within one business day — not a sales call.
So when does it pay back?
Headless earns its cost when the storefront is genuinely the constraint. Concretely, when several of these are true:
- Revenue is comfortably past $2M/year, so a five-figure build and its maintenance are a sensible fraction of the business
- Core Web Vitals are red after a theme has been properly optimised — not before
- The experience you need cannot be expressed in Liquid: configurators, complex merchandising logic, deeply custom PDPs
- Content and market complexity have outgrown the theme — multiple regions, languages, editorial depth
If your Core Web Vitals are already green, your catalogue is a few hundred SKUs with standard variants, and you sell in one market, a well-built theme will serve you better for a fraction of the money. We say the same thing on our own service page, under When NOT to go headless, and we mean it — most brands considering headless should not buy it.
A worked comparison
Three years, mid-sized DTC brand, deliberately rough:
| Well-built theme | Headless | |
|---|---|---|
| Initial build | $8k–$25k | $40k–$80k |
| Hosting | Included in Shopify | A few hundred a month |
| App ecosystem | Install and go | Integrate or replace |
| Ongoing maintenance | Low | Real and continuous |
| Ceiling on experience | Liquid | Effectively none |
| Time to a new landing page | Minutes, by marketing | Depends entirely on the CMS work |
That last row decides more projects than performance does. A headless build with a properly modelled CMS is faster for marketing than a theme. One without is slower than the theme you replaced — and that difference is a choice made during the build, not a property of headless.
The honest summary
Headless is not expensive because agencies overcharge for it. It is expensive because you are taking ownership of a layer Shopify was running for you, and ownership is ongoing.
If the storefront is what is holding your business back, that is a good trade and the payback is real. If it is not, you are buying a maintenance obligation to solve a problem you do not have.
Send us your store and we will tell you which one it is.
About the author
Manpreet Singh
Manpreet Singh is the founder of Proscube, an ecommerce growth studio. He leads the studio's Shopify and Shopify Plus engineering, headless builds, CRO, and its work on AI engine optimization, and writes its guidance on how to grow a DTC brand without wasting money. He works directly with founders — no account-manager layers between you and the people doing the work — and would rather tell a client not to build something than sell them work they don't need.
More about Manpreet
