What Became of Headless Commerce? The Rise, Reversal, and Redefinition in the AI Era

A few years ago, "headless commerce" was a cutting-edge keyword in the e-commerce industry. More recently, however, Shopify itself, which was at the heart of promoting it, has adopted a tone suggesting it's "not for everyone," leading to questions like "What ever happened to headless?"

This article will chronologically explain why headless commerce was praised (rise), why interest cooled (reversal), and what its current status is (redefined by AI), with as little exaggeration as possible. Finally, we will consider "when headless is effective now."

It should be noted that much of the information in this field comes from stakeholders (those selling headless solutions/those promoting their own platforms). This article will strive for neutrality by clearly stating "who positioned it that way."

What Was "Headless" in the First Place?

Headless commerce is an architecture that separates the "look and feel (frontend)" of a site from "commerce functionalities (backend)" and connects them via APIs. The idea is that by decoupling the frontend and backend, each can be updated independently, and a single backend can deliver content to multiple channels (web, apps, in-store terminals, etc.).

This concept was propelled as a technological trend by the slogan "MACH architecture." MACH stands for Microservices, API-first, Cloud-native, and Headless, coined around 2020 by the MACH Alliance (established by commercetools, Contentstack, EPAM, Valtech, and others). Taking this further, the idea of "modularizing the entire commerce stack and freely combining best-of-breed tools in each area" is called composable commerce.

Why Was It Praised? (The Rise: 2018-2023)

The buzzword "headless" gained traction due to a powerful narrative from its proponents.

• The term "headless commerce" itself is attributed to commercetools CEO Dirk Hoerig, and in 2023, the company declared that "after 10 years, headless has finally become mainstream."
• From the hosting/framework side, Vercel (Next.js) highlighted "the industry-wide shift from monolithic to headless commerce" and "dynamic at the speed of static" storefronts, emphasizing performance advantages and modernity.
• And in November 2021, Shopify also announced "Hydrogen," a React-based framework for headless commerce. Its positioning at the time was "achieving performance and personalization without compromise," truly embodying the expectations of the era.

The touted benefits generally included: (1) expressive power to create custom frontends, (2) fast performance, (3) omnichannel delivery from a single backend to multiple channels, (4) flexibility to freely swap functions with best-of-breed tools, and (5) avoidance of vendor lock-in.

However, it's crucial to note that many of these claims came from "those selling headless solutions." While the assertions were accurate in terms of "vendors positioning them that way," they shouldn't have been taken as objectively verified truths. And that verification came later.

Why Did Interest Cool? (The Reversal)

After the initial fervor, the gap between reality and expectations became apparent. Issues such as "headless entails high costs, high complexity, heavy operational load, and loss of convenient native features (theme editor, app integrations)" became clear through actual implementations (this is covered in more detail in our separate column, "Pros and Cons of Hydrogen + Oxygen").

Then, a decisive event symbolizing this reversal occurred: Shopify itself, a proponent of headless, contextualized the "performance advantage" narrative with its own real data.

In 2023, Shopify's performance team published an analysis based on real user data (CrUX). According to this analysis, the percentage of sites that passed all Core Web Vitals (Web performance metrics) was 59.5% for native Liquid storefronts, while Hydrogen was only 35%, and Next.js/Remix hovered around 28-29%. Shopify explicitly stated: "By default, Hydrogen is not as fast as Liquid." and "The decision to go headless should be based on reasons other than web performance. If you choose headless for reasons other than performance, Hydrogen is currently the best option."

This was definitive. The biggest selling point, "headless = fast," was acknowledged by its proponents to "not automatically be true." From this point, a sober reversal, what could be called "headless fatigue," began in earnest.

(To be fair, this performance gap is not fixed. Data from 2026 suggests that Hydrogen is closing the gap (around 78% CWV pass rate). It's also worth noting that "Liquid is fast by default" did not happen on its own: it has been achieved and maintained through Shopify's own remarkable optimization efforts, such as rewriting the storefront rendering engine from scratch to make average server response times 4-6x faster. While "Liquid is fast by default" remains true, the gap is narrowing.)

Has Shopify Really "Stopped Recommending Headless"?

It's important to understand this accurately. In conclusion, saying it has "stopped recommending it" is an overstatement. More precisely, the reality is that Shopify has adjusted its position from "recommending it to everyone" to "an option for specific requirements."

On one hand, Shopify continues to enhance its native Online Store (themes) (Online Store 2.0, full section functionality, structured data via metaobjects, etc.). The rationale is that "native is sufficient for most stores; headless is for specific reasons."

On the other hand, it hasn't abandoned headless. In fact, in June 2026, Shopify announced a major overhaul of Hydrogen. In partnership with Vercel (Next.js team), a developer preview redefines Hydrogen from a "framework" to a "framework-agnostic, runtime-agnostic toolkit." The direction is to decouple core commerce logic from specific frameworks, making it usable with any JavaScript framework like Next.js, SvelteKit, Astro, or Nuxt, and on any runtime (however, this is explicitly a "preview phase," and the current Hydrogen remains the canonical choice for production).

Therefore, Shopify's stance is not "headless denial," but rather a "transition to a mature and realistic approach" of "stopping its treatment as an excessive panacea, intelligently enhancing native solutions, and broadening options for headless based on requirements."

What Has Changed with AI? (Redefinition: 2025-2026)

Today, the situation surrounding headless is being reshaped by AI from two directions.

1. AI Has Lowered the "Cost of Creation"

The biggest bottleneck for headless was the "high cost and specialized personnel required for development and maintenance." AI coding agents (tools like Claude Code) address this. The aforementioned 2026 version of Hydrogen is precisely "re-designed for agents" and boasts of including "skills" for coding agents to automatically scaffold storefronts.

This embodies the trend that AI lowers the barrier to custom development, changing the economics of headless construction. However, soberly, this is not "complete autonomy," but rather the stage of agent assistance and human-led processes, where humans define the framework and AI accelerates practical tasks. The responsibility for headless maintenance itself has not disappeared.

2. AI Has Changed the "Meaning of the Storefront"

A more fundamental change is agentic commerce. A world is emerging where generative AI, such as ChatGPT, Gemini, and Copilot, discovers products and makes purchases on behalf of users.

Shopify is actively involved in this, developing the "Universal Commerce Protocol (UCP)" in collaboration with Google to provide a system (including MCP servers) that allows AI agents to safely discover products, add to cart, and complete purchases. Furthermore, with "Agentic Storefronts," it claims that once merchants set up their data, products will be displayed across multiple AIs like ChatGPT, Microsoft Copilot, Google AI Mode, and Gemini. This expands the premise of "storefront = a single company site" to "distribution to AI conversation channels."

However, here we must not get carried away by the hype and consider both sides. There is strong skepticism regarding agentic commerce. Independent analyst Andrew Lipsman describes it as a "collective hallucination," arguing that there is little evidence of actual consumers entrusting purchases to AI. Indeed, even "Instant Checkout," launched by OpenAI and Stripe in September 2025, still requires users to explicitly approve each purchase, meaning it's not yet "fully autonomous agent buying." The technology (protocols and mechanisms) is rapidly maturing. So what about actual demand? Brand discovery through AI conversations and referrals to product detail pages are already a clearly established trend. The real question lies beyond that: how far will experiences spread in which a purchase is completed without a human ever laying eyes on the storefront? Shopping has an element of enjoyment, and people tend to check prices carefully. On the other hand, "labor-like" shopping, such as restocking cleaning supplies, will be the first to be automated. Shopping as enjoyment will remain in human hands, while automation advances first in shopping as a chore — that is our honest read of where things stand today.

So, How Should You Choose "Headless" Now?

Based on the analysis so far, here is our view. The key is to separate the decision of "what to use for the commerce backend" from "whether to go headless."

Many of the original grand claims of headless—free combination of best-of-breed tools, avoidance of vendor lock-in, microservice-like scalability (i.e., the MACH/composable world view)—lose most of their meaning the moment you decide to "fix the backend to Shopify." This is natural since you're committing to a single commerce engine.

In that case, the remaining reason to go headless boils down to one point: "whether you need to design and manage how your site looks, and how it connects to other systems, yourself — beyond the confines of Shopify themes." Specifically, there are two main valid patterns:

(a) Sites where content/experience is the main focus (suited for Hydrogen)

These are sites where structured information and content presentation are primary, rather than just product display and sales. Like corporate or brand sites, the structure is "information and brand world first, with product sales as an extension." Native themes are designed with a "product and collection-centric" philosophy, which can feel restrictive for sites that prioritize information design. In such cases, Hydrogen (+ Oxygen), which allows for custom storefront creation, shines. In fact, the very site you are reading right now is built with Hydrogen: a corporate site with information design at its core, custom-built with full freedom, using Shopify as the backend (this column, too, is delivered through Shopify's blog feature) — a real-world example of this exact pattern.

(b) Integration of product data into existing systems (often, Hydrogen is not needed)

This applies to cases where Shopify product data needs to be "mounted and displayed" on an existing company website, media, business system, or app. In this scenario, Shopify functions purely as the commerce backend (data container), and the frontend resides in the existing system. This is the most "pure" form of headless, but since a full storefront is not being built, Hydrogen is not necessary. In most cases, simply calling the Storefront API to retrieve and display product data is sufficient.

This distinction is important. "Headless = Hydrogen" is incorrect. The key to avoiding unnecessary costs and complexity is to differentiate: "If creating a new primary storefront, choose (a) = Hydrogen; if only integrating products into an existing system, choose (b) = plain Storefront API."

Conclusion

Headless commerce was neither "dead" nor a "silver bullet." After a period of being overhyped as a buzzword, it is now settling into its rightful place as "one realistic option to choose when it meets specific objectives."

• The touted performance advantage was contextualized by Shopify itself, stating that it's "not automatically fast."
• Shopify has not abandoned headless but has matured its positioning from "for everyone" to "based on requirements," and is remaking it into a framework-agnostic toolkit for the AI era.
• AI is lowering the "cost of creation" (reducing barriers to entry) and changing the "meaning of the storefront" (distribution to AI channels). Brand discovery and referrals are already a clearly established trend, but automation of the full purchase is expected to advance gradually, starting with "shopping as a chore."

The practical conclusion is simple: If you've decided on Shopify for your backend, the reason to choose headless boils down to whether you need to design and manage how your site looks and connects to other systems, beyond the confines of themes. If the answer is yes, and content and experience are primary, then Hydrogen is a shining option. If the goal is integration into an existing system, then a lightweight implementation using the plain Storefront API is wise. Choosing based on your own objectives—rather than "because it's trendy" or "because it's outdated"—is the healthy way to approach headless after navigating waves of enthusiasm and disillusionment.

References and Sources

• Liquid vs Headless: A look at real user web performance (Shopify Performance)
• How Shopify Reduced Storefront Response Times with a Rewrite (Shopify Engineering)
• Hydrogen developer preview - 2021 announcement (Shopify Partners)
• Hydrogen developer preview - 2026 toolkitization (Shopify Changelog)
• Hydrogen Updates (Hydrogen official)
• The differences between Composable, Headless, and MACH (commercetools)
• Modern headless commerce is now mainstream (commercetools)
• Next.js Commerce 2.0 (Vercel)
• AI commerce at scale - UCP / Agentic Storefronts (Shopify)
• Skepticism Grows Over AI Shopping Agents as ChatGPT Checkout Launches (PPC Land)

This column is based on public information as of July 2026. Major sources for performance comparisons and standard promotions include statements from stakeholders themselves (Shopify, Vercel, commercetools, etc.), so efforts have been made to maintain neutrality by noting "who positioned it that way." The trends in agentic commerce are fluid, and the evaluation may change depending on future developments.