Headless storefronts
Hyvä, PWA Studio or ScandiPWA, chosen for your case. A frontend decision for Adobe Commerce based on your team, budget and roadmap.
Who this is for
- Merchants on Luma who need a faster storefront and are weighing Hyvä against a headless rebuild
- Teams running a PWA Studio or ScandiPWA storefront that has become hard to change
- Agencies that want a second opinion before committing a client to a frontend stack
What you get
A frontend decision with reasons
Hyvä, PWA Studio and ScandiPWA compared on your catalog, extensions, team skills and budget, in writing.
A cost you can plan with
Build, extension compatibility and ongoing maintenance estimated in hours, not only the launch price.
A GraphQL layer that holds up
Missing queries, custom attributes and caching gaps found and fixed before the frontend depends on them.
A storefront your team owns
Build, rescue or code-review support until launch, so your developers can carry it afterwards.
How it works
Week 1
Assess
Week 2
Decide
From week 3
Build
In detail
Headless is not a goal in itself. A separate React storefront means two codebases, a GraphQL layer that must cover every feature your store uses, and extensions that often ship without frontend support. For many merchants a Hyvä theme delivers most of the speed at a lower cost and risk. For others, such as a content-heavy brand or a team already strong in React, a headless storefront is the right call.
I have worked with all three options. At Guidance I maintained a ScandiPWA headless storefront in production, including the GraphQL work behind it, for 8-figure-GMV merchants in e-learning and digital content. I also work with Hyvä and PWA Studio, and I design GraphQL and REST APIs on Adobe Commerce.
The engagement starts with a two-week assessment of your theme, extensions, integrations, team skills and traffic. You get a written recommendation with the cost of building and of maintaining each option, and the risks I see. If you go ahead, I either lead the build or review the work of your team or agency until launch.
If you already run a headless storefront that has become slow or hard to change, I start from the code and the GraphQL layer and fix what blocks you first.
Related case study
−50 %Typical page load
Halving page-load times for an 8-figure e-learning store
Read the case study