Shopify Headless is a Mistake for 90% of DTC Brands
I have been a CTO. I have sat in rooms where agencies pitched headless commerce as the future of DTC. I have watched brands spend six figures chasing a technical architecture that solved problems they did not have. This post is the one I wish had existed before those meetings happened.
The Promise
The pitch is seductive: decouple your frontend from Shopify's backend, use the Storefront API, build with Next.js or Remix, host on Vercel or Cloudflare, and unlock complete creative freedom, blazing page speeds, and an omnichannel-ready architecture. You get full control over your UX, rendering pipeline, and design system. No more Liquid templates holding you back. It sounds like the adult version of e-commerce engineering.
You Are Now Responsible for Everything Shopify Used to Handle
This is the part that the agency conveniently skips. The moment you go headless, Shopify's battle-tested frontend becomes your problem. Cart behavior, discount logic, checkout redirects, metafield rendering, localization, currency formatting, product variant selection, every single one of these used to just work. Now you own the implementation, the edge cases, the bugs, and the regression testing. Shopify's Liquid theme ecosystem has years of production hardening behind it. Your custom Next.js storefront has whatever your agency shipped over the past eight weeks.
The Checkout Wall That Nobody Warns You About
Here is the dirty secret of headless Shopify: You cannot go headless on the checkout. Shopify locks checkout behind its own hosted pages. Unless you are on Shopify Plus and willing to invest in Checkout Extensibility. Your beautiful custom frontend funnels directly into a Shopify-controlled checkout that may not match your brand aesthetic at all. You spent $150K on a custom storefront, and your conversion moment still lives inside a template you do not own. The seam is visible and jarring, and your customers will feel it.
Your Engineering Team Does Not Exist or Cannot Sustain This
Headless is not a one-time build. It is an ongoing engineering commitment. Every Shopify API update, every new product feature, every change to the metaobject schema requires frontend work your in-house team may not be equipped to handle. The brands I have seen go headless without a dedicated frontend engineer on staff invariably end up in the same place: frozen in time on an outdated implementation, paying an agency retainer to keep the lights on, unable to react quickly to merchandising changes because someone needs to touch the codebase. Vercel or Sanity are not free. Preview environments, build pipelines, caching invalidation strategies, these are real operational costs that accumulate quietly until they are not quiet anymore.
Performance Gains Are Real but Massively Overstated
Yes, a well-built headless storefront with incremental static regeneration and edge caching can be fast. Very fast. But Shopify's own Online Store 2.0 themes, particularly Dawn, are not slow. They score well on Core Web Vitals out of the box. The performance delta between a properly optimized Dawn theme and a well-built headless frontend is measurable in milliseconds for most brands. The conversion lift from that delta is statistically negligible unless you are operating at a serious scale, think tens of millions of monthly sessions. For brands doing $1M to $10M in annual revenue, the performance argument is almost never the real bottleneck. The bottleneck is offering quality, traffic acquisition, and retention.
The App Ecosystem Breaks in Subtle and Painful Ways
The Shopify app ecosystem is one of the platform's greatest assets. Klaviyo, Rebuy, Okendo, Attentive, Yotpo, these tools plug into Shopify's Liquid themes via script tags and theme extensions. Go headless, and you immediately break most of them. Their embeds no longer work automatically. You need to wire them manually through APIs if those APIs even exist. Some vendors do not support headless at all, or their headless support is a second-class citizen maintained by one engineer on a Friday afternoon. Every app you depend on becomes a custom integration project. Your "faster time to market" promise evaporates the moment you need to add post-purchase upsells or review collection widgets.
The SEO Argument Has Flipped
In 2020, the headless SEO argument had merit. Server-side rendering was the obvious path to crawlable, indexable content when client-side React was still opaque to Googlebot. That world no longer exists. Google has significantly improved JavaScript rendering, and Shopify's native themes render server-side HTML by default. The technical SEO gap that made headless compelling has narrowed to near zero for most use cases. Going headless today for SEO reasons is solving a problem that has largely solved itself.
The Hidden Cost Nobody Puts in the Spreadsheet
I have seen the pitch decks. They include agency fees, Vercel hosting, and maybe a CDN line item. What they never include: the developer hours spent debugging hydration errors, the QA cycles on cross-browser cart behavior, the time your marketing team loses while waiting for engineering to make what would have been a two-minute theme edit, the cost of re-implementing every Shopify-native feature that breaks during an API version migration. Add those up over 24 months, and the ROI case for headless collapses for any brand not running genuinely differentiated, engineering-intensive commerce experiences.
The Conclusion: Below $50M ARR, Think Very Hard Before Going Headless
I will be blunt. If your DTC brand is doing less than $50M in annual revenue, I am convinced headless Shopify is almost certainly the wrong call. Below that threshold, your conversion and revenue problems are not caused by your frontend architecture. They are caused by your acquisition strategy, retention mechanics, offer construction, or supply chain. None of those gets fixed by switching from Liquid to Next.js.
Between $50M and $150M, going headless can make sense if you have a dedicated frontend engineering team, genuine product complexity that the Liquid ecosystem cannot express, and a clear omnichannel or composable commerce roadmap. Otherwise, you are burning engineering bandwidth on plumbing that Shopify would have maintained for you for free.
Above $150M, the calculus changes. At that scale, performance margins matter commercially, engineering investment is amortized across genuine volume, and the control and flexibility of a headless architecture become real strategic assets rather than premature abstractions.
The uncomfortable question I keep asking: if Shopify, one of the most engineer-dense product companies on the planet, keeps investing in Liquid, Dawn, and Online Store 2.0 as first-class citizens, who exactly is telling you those tools are not good enough for your $5M brand?
