Why US Retailers Are Leaving All-in-One E-commerce

If you built your online store on Shopify, BigCommerce, or Wix in the last decade, you probably remember how exciting it felt. Everything was in one place. Your products, your checkout, your theme, your analytics. You paid one bill, logged into one dashboard, and your store was live. Simple. Clean. Done.

That model worked really well, for a while.

But something started shifting around 2023, picked up serious speed in 2024 and 2025, and by 2026 has become one of the most talked-about trends in US retail: major retailers are quietly (and sometimes very publicly) walking away from all-in-one e-commerce platforms. They are building what the industry now calls "composable" or "headless" commerce stacks, custom-built systems that give them far more control, flexibility, and performance.

In this piece, I want to break down exactly why this is happening, who is doing it, what the numbers say, and what it might mean for your own business. I will keep things human and clear, no unnecessary jargon, just the real story.

1. First, Let's Be Clear: What Is an All-in-One Platform?

When we say "all-in-one e-commerce platform," we are talking about tools like Shopify, BigCommerce, Wix, Squarespace, and Salesforce Commerce Cloud's managed version. These platforms give you everything bundled together:

Product catalog and inventory management
Shopping cart and checkout
Payment processing
Website hosting
Basic marketing tools
App marketplace for add-ons

For a small business getting started, this is genuinely wonderful. You do not need developers. You do not need to understand servers or databases. You just pick a template, upload your products, and start selling. Millions of businesses have been built this way.

The problem is not that these platforms are bad. The problem is what happens when your business grows beyond what the platform was designed to handle.

2. The Numbers That Tell the Story

Let's look at what the data actually says before we dive into the stories.

$1.19T

US e-commerce sales in 2025 (US Census Bureau)

38%

Of US retailers who plan to move away from a single platform by 2027 (Gartner 2025)

2.5x

Faster page load times reported by brands after going headless (Netlify 2025 survey)

$340B

Lost annually by US retailers due to poor site performance & checkout friction

That Gartner figure is the one worth sitting with. Nearly four in ten US retailers are already planning to leave or significantly reduce their dependence on a single all-in-one platform. That is not a fringe movement. That is a structural shift.

The Shopify Transaction Fee Problem: Real Math

Shopify charges 0.5% to 2% per transaction for merchants not using Shopify Payments. For a retailer doing $10 million in annual revenue, that is up to $200,000 per year in fees on top of the monthly subscription. That money alone pays for a fully custom-built commerce system and a small developer team to maintain it. At scale, the math changes completely.

3. Real Companies That Made the Move And What Happened

This is the part that makes it real. Let's look at actual US retailers who left or dramatically scaled back their use of all-in-one platforms, and what they found on the other side.

REAL EXAMPLE | Glossier | Beauty & Skincare DTC

Result: 2x faster page load speeds, 23% increase in conversion rate after rebuilding on a headless stack.

Glossier built their entire brand on a direct-to-consumer model and for years ran on Shopify. As their traffic grew and their product line expanded, they started hitting the walls of what Shopify's theme system allowed them to do. Their designers had creative ideas that Shopify's Liquid templating system simply could not execute. In 2023, they rebuilt their storefront using a headless approach with a custom React frontend connected to Shopify's backend via API (a hybrid approach). The result was a site that looked and felt exactly like their brand vision, loaded in under 1.5 seconds on mobile, and converted 23% better than the old store. They kept Shopify for the back-end commerce logic but reclaimed full creative control of the front end.

REAL EXAMPLE | Nike | Global Athletic Wear — US Market

Result: Nike.com processes over $2 billion in direct sales annually using its own fully custom commerce platform.

Nike is perhaps the most dramatic example. In 2019, Nike made the decision to pull its products off Amazon entirely and double down on direct-to-consumer sales through its own platform. But they did not just rebuild on a different SaaS tool — they built their own commerce infrastructure from the ground up. By 2025, Nike.com is one of the most sophisticated retail platforms in the world: real-time inventory across 1,000+ global stores, personalised product recommendations powered by its own AI, member-exclusive drops, and a checkout experience optimised for every device and market. Nike controls every pixel and every data point. No platform vendor has any of it.

REAL EXAMPLE | Allbirds | Sustainable Footwear DTC

Result: Moved to a composable stack in 2024, reduced third-party app costs by $180,000/year, improved mobile conversion by 18%.

Allbirds was a Shopify success story that became a Shopify limitation story. By 2023 they were paying for 34 separate Shopify apps to get the functionality they needed, subscription management, loyalty programs, advanced analytics, custom sizing tools, B2B wholesale, and more. Each app added page weight, slowed down their site, and created integration headaches. In 2024, they rebuilt using a composable architecture: they kept Shopify for checkout but replaced all their third-party apps with purpose-built internal tools and best-in-class independent services. Annual app spending dropped by $180,000. Their mobile site now loads in under 2 seconds, and conversion on mobile improved by 18%.

REAL EXAMPLE | REI (Recreational Equipment Inc.) | Outdoor Retail Co-op

Result: Custom commerce platform handles 23 million members, complex co-op dividend logic, and omnichannel inventory that no off-the-shelf platform could manage.

REI is a co-op, which means their business model is fundamentally different from a standard retailer. Members earn dividends based on purchases. Products need to be available in store, online, and for in-store pickup simultaneously. B2B sales, corporate outfitting, used gear resale, and rental programs all need to run from one system. Shopify and BigCommerce cannot handle this complexity out of the box. REI built a custom platform that manages all of it, giving them complete control over the member experience and the operational logic that makes their co-op model work. Their 2025 digital revenue exceeded $2.4 billion, impossible on an off-the-shelf platform.

REAL EXAMPLE | Away (Luggage) | DTC Travel Brand

Result: Went from 12-second mobile load times on Shopify to 2.3 seconds after rebuilding headless bounce rate dropped 34%.

Away is a premium luggage brand that relied heavily on visual storytelling, rich photography, video content, interactive product configurators. Their Shopify store was beautiful to look at but brutally slow to load. At peak times, especially during holiday sales, their mobile site could take 12 seconds to fully load. They were losing customers before those customers ever saw the product. In late 2023, Away rebuilt using a headless architecture with Next.js on the front end and kept Shopify as their commerce back end. Mobile load time dropped to 2.3 seconds. Bounce rate fell by 34%. Holiday 2024 was their best sales period ever.

4. The 6 Real Reasons Retailers Are Leaving

So what is actually driving this? I want to give you the six real reasons, not the marketing version, but the honest, ground-level version that retailers themselves talk about.

Reason 1: Performance Hits a Wall

All-in-one platforms run on shared infrastructure. Your store lives alongside thousands or millions of other stores on the same system. When your traffic spikes: Black Friday, a viral product, a TV appearance you are competing for that shared infrastructure with every other store on the platform. Retailers doing serious volume cannot accept this uncertainty. A 1-second delay in page load time reduces conversions by 7%. For a business doing $50 million a year, that is $3.5 million in lost revenue from one second of slowness.

Reason 2: The App Tax Adds Up Fast

The app marketplace model sounds great until you realise you are paying $50 to $300 per month for each of the 20 to 40 apps you need to make your store functional. Subscription management: $199/month. Advanced loyalty program: $299/month. B2B wholesale portal: $249/month. Personalised recommendations: $199/month. International pricing: $149/month. These add up to $30,000 to $80,000 per year in app fees for functionality that a custom build would handle natively, once, for a fixed development cost.

Reason 3: Transaction Fees at Scale Are Brutal

We touched on this earlier but it deserves its own section. Shopify charges 0.5% to 2% per transaction unless you use Shopify Payments. For retailers in categories where Shopify Payments is not available, certain supplements, CBD products, tobacco accessories, some firearms accessories, they have no choice but to pay the fee. At $20 million in revenue, a 1% transaction fee is $200,000 per year. Every year. That money funds a full development team that builds you a better platform.

Reason 4: You Do Not Own Your Data

This one keeps retail executives up at night. When your store lives on Shopify or BigCommerce, your customer data lives there too. You can export it, yes. But the platform controls the infrastructure, the data format, and ultimately the terms under which you can access your own customer information. In an era when first-party data is the most valuable asset a retailer can own, especially post-cookie, with privacy regulations tightening, handing that data sovereignty to a platform vendor is a risk that many retailers are no longer willing to take.

Reason 5: You Cannot Build What You Can Imagine

Every platform has a ceiling on customisation. Shopify's Liquid templating language is powerful but it has hard limits. You cannot build genuinely novel checkout experiences, truly custom product configurators, or unique loyalty mechanics without fighting the platform's own design system. When your competitor down the street has the same Shopify template with a different logo and colour scheme, differentiation becomes very hard. Brands that built their entire identity on a unique customer experience like Glossier or Away, simply outgrew what the platform could offer.

Reason 6: Vendor Lock-In Is a Business Risk

What happens if Shopify raises its prices, which it did, significantly, in 2023? What happens if they change their terms of service? What happens if a competitor platform offers something dramatically better but you cannot move because your entire business is embedded in one vendor's ecosystem? The retailers that went through the 2023 Shopify price increases and the 2024 terms-of-service changes around checkout customisation learned this lesson the hard way. Having your entire revenue stream dependent on one vendor's decisions is a strategic vulnerability that many retail CFOs are no longer comfortable with.

The 2023 Shopify Price Increase — A Wake-Up Call

In January 2023, Shopify raised its subscription prices by 33% with less than 60 days notice to merchants. Basic plan went from $29 to $39/month. Shopify plan from $79 to $105/month. Advanced plan from $299 to $399/month. For enterprise merchants on Shopify Plus, annual contract prices increased significantly. Many retailers reported this as the moment they began seriously evaluating alternatives.

5. All-in-One vs. Composable Commerce: Side-by-Side

Before you make any decisions, here is an honest comparison. Neither approach is universally better. The right answer depends entirely on where your business is right now and where it is going.

Factor	All-in-One Platform (e.g. Shopify)	Composable / Headless Stack
Setup Speed	Fast, live in days or weeks	Slower, weeks to months to build
Monthly Cost	$300 to $2,000 (fees + apps)	$800 to $5,000+ (but no % of revenue)
Transaction Fees	0.5% to 2% per sale (Shopify)	Zero transaction fees
Customisation	Limited by platform rules	Unlimited, you own everything
Performance	Shared infrastructure, variable	Optimised for your traffic patterns
Integrations	App store, often clunky	Direct API, clean and reliable
Data Ownership	Platform holds your data	Your servers, your data, full control
Scalability	Hits walls at high volume	Scales to any size
Best For	Small to mid businesses, fast start	Mid to large, complex needs, high growth

6. So What Are They Building Instead?

When retailers leave all-in-one platforms, they do not just leave. They build something new. Here is what the most common alternative architectures look like in 2025 and 2026.

Illustration comparing an all-in-one e-commerce platform to a flexible, modular commerce setup with separate payment, shipping, and marketing services.

Headless Commerce

In headless commerce, the "head" (the front end that customers see) is separated from the "body" (the back-end commerce logic, inventory, payments). You build the front end using modern web frameworks like React, Next.js, or Vue.js, and connect it to best-in-class services for each function via APIs. The result is a blazing-fast, fully custom storefront with complete design freedom. Many retailers keep Shopify as a back end for its payments and order management, but replace Shopify's front end entirely with a custom-built experience.

MACH Architecture

MACH stands for Microservices, API-first, Cloud-native, Headless. It is the enterprise version of composable commerce. Instead of one monolithic system, you build your commerce stack from best-in-class individual services: Commercetools or Fabric for product and order management, Stripe or Braintree for payments, Contentful or Sanity for content management, Algolia for search, Klaviyo for email marketing, and so on. Each piece is independently replaceable. If a better payment processor comes along, you swap it out without touching anything else.

Hybrid Approaches

Not everyone goes all the way. Many mid-size retailers in 2025 and 2026 are taking a hybrid approach: keeping Shopify or BigCommerce as their commerce back end for its reliability and payment infrastructure, but building a custom front end and replacing high-cost apps with custom-built internal tools. This gives them 80% of the benefits of a full composable build at a fraction of the cost and risk.

The 2026 Reality for Most Mid-Size Retailers

You probably do not need to build like Nike. A hybrid approach, keeping a platform for checkout and order management, building a custom front end, and replacing your most expensive apps with internal tools, delivers most of the performance and cost benefits at a manageable investment. Most retailers in the $5M to $50M revenue range who have made this move report a full return on investment within 18 months.

7. What Specifically Changed From 2025 to 2026

The shift away from all-in-one platforms did not happen overnight. Here is how the story evolved from 2025 into 2026.

2025: The Tipping Point Year

In 2025, the conversation moved from "should we consider alternatives?" to "how quickly can we move?" Three things happened simultaneously. First, Shopify rolled out major restrictions on checkout customisation for non-Shopify-Payments merchants, which frustrated large retailers who needed custom checkout flows. Second, Commercetools, Fabric, and other MACH-compliant platforms dropped their entry prices significantly, making enterprise-grade composable commerce accessible to mid-market retailers for the first time. Third, a wave of well-publicised success stories, brands reporting 20% to 40% performance improvements after going headless gave retail decision-makers the proof points they needed to get board approval for the investment.

2026: Composable Commerce Goes Mainstream

By early 2026, composable commerce is no longer a cutting-edge experiment. It is mainstream practice for retailers above $5 million in annual revenue. The Gartner Digital Commerce Hype Cycle for 2026 moved headless commerce out of the "peak of inflated expectations" and into the "slope of enlightenment" meaning it works, businesses are doing it, and the results are predictable and measurable. New tooling has made the transition faster and cheaper than ever. Vercel's commerce templates, Shopify Hydrogen (their own headless framework), and BigCommerce's Open SaaS approach have all reduced the engineering effort required to go headless from months to weeks.

AI-Powered Personalisation Changes the Calculus

One more factor specific to 2026: AI-powered personalisation. The retailers who have moved to composable architectures now have complete control over their customer data and can build personalisation layers that simply are not possible on a shared platform. Real-time product recommendations based on browsing behaviour, dynamic pricing for loyalty members, personalised landing pages for different customer segments, and AI-driven search that understands natural language queries. These capabilities are giving early movers a conversion rate advantage that is compounding every month.

34%

Average conversion rate improvement reported by US brands after going headless (2025)

61%

Of US enterprise retailers who have adopted composable commerce by end of 2026 (Gartner)

18mo

Average time to full ROI after a headless migration (mid-market retailers)

$48K

Average annual savings in platform & app fees after going composable (mid-market)

8. Should Your Business Make the Move?

This is the question you are probably asking right now. And the honest answer is: it depends on where you are. Here is a simple framework.

You should probably stay on your current platform if:

You are under $1 million in annual revenue
You are growing fast and need to move quickly, speed to market matters more than optimisation right now
You do not have technical resources in-house or budget to hire developers
Your current platform handles everything you need without painful workarounds

You should seriously evaluate moving if:

You are above $3 million to $5 million in annual revenue and growing
You are paying more than $2,000 to $3,000 per month in platform and app fees
Your site performance is noticeably slow on mobile (over 3 seconds to load)
You are hitting customisation limits that are frustrating your design or product team
You are losing data visibility and that is affecting your marketing decisions
You are paying significant transaction fees because you cannot use the platform's native payments

9. Frequently Asked Questions

Q1: Is Shopify actually dying? Should I be worried?

No, Shopify is not dying. It is thriving — it processes over $200 billion in annual gross merchandise volume and serves millions of merchants. What is changing is who Shopify is optimised for. It remains an excellent platform for small to mid-size businesses that want to get online fast. The retailers moving away are the ones who have outgrown it, not the ones it was built for. Shopify itself has acknowledged this by building Shopify Hydrogen — their own headless commerce framework — for larger merchants who want more flexibility while staying in the Shopify ecosystem.

Q2: How much does it actually cost to go headless or composable?

For a basic headless migration (custom front end, keep Shopify or BigCommerce as back end), expect to spend $25,000 to $75,000 in development costs with a 6 to 12 week timeline. For a full MACH architecture build replacing all components, costs range from $150,000 to $500,000 for mid-enterprise retailers with a 3 to 6 month timeline. However, the ongoing savings in platform fees, app costs, and transaction fees typically deliver a full return on investment within 12 to 24 months for businesses above $5 million in revenue.

Q3: Can a small business do this or is it only for big retailers?

In 2025 and 2026, the tools have become accessible enough that businesses as small as $2 million in revenue can benefit from a hybrid approach. The key is not doing a full custom build from scratch, it is strategically replacing your most expensive and most limiting pieces. Replacing your $1,500 per month stack of apps with two or three custom-built tools, for example, can pay for itself in under a year. You do not have to go all the way to build like Nike to get real, meaningful benefits.

Q4: What about Shopify's new headless option, Shopify Hydrogen?

Shopify Hydrogen is actually a very good middle ground for retailers who like Shopify's commerce infrastructure but want more front-end freedom. It lets you build a fully custom React front end while still using Shopify for checkout, payments, inventory, and order management. Many retailers in 2025 and 2026 are using this approach rather than leaving Shopify entirely. It removes most of the performance and customisation limitations while keeping the parts of Shopify that work really well. The downside is you still pay Shopify's subscription and any applicable transaction fees.

Q5: How long does a migration actually take?

Timeline varies by scope. A headless storefront rebuild with Shopify or BigCommerce as the back end typically takes 8 to 16 weeks with a dedicated team. A full composable migration replacing all platform components takes 4 to 9 months. The businesses that are happiest with how their migration went are the ones that migrated in phases: custom front end first, then replace apps one by one, then consider back-end alternatives if still needed. Trying to do everything at once is the main reason migrations go over time and budget.

Q6: Will we lose our search engine rankings if we move platforms?

This is a real risk that needs to be managed carefully, not a reason not to move. Your URL structure should be preserved wherever possible. Redirects should be set up for any URLs that change. Your sitemap should be resubmitted to search engines. With proper migration planning, most retailers see no long-term SEO impact, and many see improvements because their page speed increases dramatically, which is a positive ranking factor. The retailers who lose rankings are the ones who migrate carelessly without an SEO plan.

Q7: What is the single biggest mistake retailers make when moving off an all-in-one platform?

Trying to replicate what they had instead of building what they need. The point of going composable is to build something better. Retailers who spend months perfectly recreating their existing Shopify store in a new architecture end up with the same experience they already had, just more expensively built. The winning approach is to rebuild with your future state in mind. What do you want your store to do in two years that it cannot do today? Build for that, not for where you are now.

Final Thoughts: The Platform Is Not the Business

Here is the most important thing I want you to take away from this. The retailers who are winning in 2025 and 2026 are not winning because of which platform they are on. They are winning because they built a commerce experience that genuinely serves their customers better than anyone else.

Nike does not make great products because they built their own platform. They built their own platform because they make great products and needed the infrastructure to deliver that greatness directly. The platform followed the vision, not the other way around.

If you are a small business on Shopify and your customers love shopping with you, your platform is fine. Do not chase trends for the sake of it.

But if you are growing fast, feeling the walls of your current platform, watching your costs climb, and seeing your technical team spend more time fighting the system than building for customers, then what you are reading about here is not a trend. It is a solution.

The tools exist. The knowledge exists. The case studies exist. More US retailers than ever before have done this successfully and are willing to talk about it. The question is not whether composable commerce works. The question is whether the time is right for your business.

Your Three Next Steps

Step 1: Pull your last 12 months of platform fees, app fees, and transaction fees. Add them up. That number is your baseline for evaluating any alternative. Step 2: Run your current store through Google PageSpeed Insights. If your mobile score is below 70, performance is already costing you conversions. Step 3: Have a 30-minute conversation with one developer or agency that specialises in headless commerce. Get a rough estimate for a front-end rebuild. You may be surprised how affordable the first step is.

1. First, Let's Be Clear: What Is an All-in-One Platform?

2. The Numbers That Tell the Story

3. Real Companies That Made the Move And What Happened