How to pick a headless commerce development company

Three different engagements hide behind the same request. An audit answers whether headless is the right move at all and produces evidence rather than a proposal. A frontend build puts a new storefront on the commerce backend you already run. A replatform changes the backend too and is a substantially larger project. Companies that quote the third when you asked about the first are not necessarily dishonest — but you should be the one deciding which you are buying, and the scoping conversation should make the distinction explicit.

• Audit: evidence about whether to do the project at all
• Frontend build: new storefront, existing backend untouched
• Replatform: new backend, migration, operations rebuilt
• Insist on knowing which one a quote covers

Ask what happens to organic traffic during the migration, and listen for a redirect map, metadata parity and post-launch coverage monitoring rather than reassurance. Ask which of your plugins survive and which do not, and expect a two-column list rather than "we will handle it". Ask how checkout is approached, and be cautious of anyone whose default is a fully custom checkout without asking about your gateways. Ask what they measure after launch, and whether it is field data or a lab score. Vague answers to any of these predict the invoice you will receive later.

A written scope naming the templates being built, the integrations being connected, the data being touched and what is explicitly excluded. A price that is fixed where the work is knowable and an audit priced separately where it is not — padding an estimate to cover unknowns costs you money whether or not the unknowns materialise. And a change process: what happens when you want something mid-project, quoted before it is built rather than absorbed and billed afterwards.

Establish before signing who owns the repository, whether the storefront depends on any proprietary layer, whether another developer could reasonably take it over, and what the handover contains. A codebase only its author can change is a commercial position rather than a technical necessity, and it is the mechanism behind most unhappy long-term client relationships in this field. Ask for the handover documentation to be part of the deliverable, in writing.

• Repository ownership stated in the contract
• No proprietary layer required to run or deploy
• Documentation and environment setup in the handover
• Another competent developer could take it over

Fixed price against a defined scope is normal and appropriate for frontend builds. Time and materials suits genuinely exploratory work and requires more trust in both directions. Retainers make sense after launch, when what you need is maintenance and iteration rather than a project. Our own engagements start at $1,999 for a straightforward setup and scale with catalogue, templates, integrations and how much search traffic must be protected — the cost estimator produces a planning range and pricing shows the package structure.

Guaranteed rankings or conversion percentages — nobody can promise these, and offering them is either inexperience or dishonesty. A quote with no audit when your store has meaningful organic traffic. No staging environment in the plan. A proposal that never mentions redirects. Enthusiasm for a custom checkout before anyone has asked which payment gateways you use. And a refusal to start with a small piece of work, which usually means the commercial model depends on the large one.

A larger agency brings process, redundancy and the ability to absorb someone leaving mid-project, at higher rates and often with the senior person who won the work replaced by a junior who delivers it. A small senior team gives you direct access to the people writing the code and lower overhead, with real key-person risk. Neither is universally right. What matters is asking who specifically will do the work, and whether the person answering your questions in the sales call is the same person writing the code.

The best first engagement is a scoped audit or a single template rebuild — enough to see how a partner communicates, estimates and delivers before anything large is at stake. A good partner will suggest this themselves. If a company resists starting small, that reluctance is information about how they handle risk, and it is cheaper to learn it now than after a six-week build. The hiring guide covers the same discipline for individual developers.

We build Next.js storefronts on WooCommerce backends, run the audits that decide whether that is warranted, and handle the migration risk that comes with it. We do not manage advertising accounts, and we do not take projects where the honest answer is that hosting, images or a plugin audit would deliver most of the win — you will hear that from the free audit before there is a proposal. Turning down work that will not pay back is the only credible version of this pitch.