Banner: Build the difference, buy the rest. Build versus buy for ecommerce. A Playbook from IQ4hire.

Insights

Build the difference, buy the rest: build versus buy for ecommerce

A growing retailer can match custom-built competitors without their engineering budgets. Three paths, and one rule: build where customers choose you.

Your largest competitors built their own commerce platforms, so you may feel you should build yours. The better question is which parts of the platform your customers choose you for, and who will maintain the rest. Build those parts, buy the rest, and you can match the leaders without matching their engineering payroll.

Why the category leaders build

In many specialty categories, the leaders run custom commerce platforms. Marketplaces and premium retailers write their own checkout, search, apps and pricing tools, and call them an edge. For a growing retailer the conclusion seems obvious: they built it, so should we.

Look at who pays for those platforms. They sit behind venture capital, growth equity, family wealth or public markets, and they are run by standing engineering teams. Chrono24, the luxury-watch marketplace, is a public example. It cut about 110 of its roughly 460 jobs in January 2025, according to press reports. I do not know how many of those were engineers. The point is that a large standing team is a cost you carry through slow markets.

A custom platform is not a project. It is a permanent payroll. Most growing retailers cannot carry it, and they do not need to. Their edge is rarely the code. It is their inventory, their expertise, their pricing and the trust customers give them. None of that depends on writing your own checkout.

What a custom build has to own forever

Customers never choose a retailer for its sales-tax engine. But every commerce site depends on dozens of capabilities like it, and each has to work every day: payments, tax, search, accounts, accessibility, fraud checks, and security on checkout pages. A mature platform provides these out of the box or through maintained extensions. Built custom, each one is yours to write, test, secure and update, and the list never shrinks.

A custom commerce build has to own twelve capabilities forever, from checkout and tax to security patches. (opens the full-size image in a new tab)

What I see in mid-sized retailers that built custom

Many have already run this experiment. A small, capable team built a site years ago, and the same symptoms show up as it outgrows the team:

  • Two sites to maintain. Separate desktop and mobile experiences double the work for every change.
  • Security debt. Patching and access control fall on the people who are supposed to be building features.
  • One person holds it all. Every departure becomes a business risk.
  • A wish list that never shrinks. Better filtering, service tools, an app or a new channel wait year after year, because upkeep always comes first.

This says nothing bad about the developers. It is arithmetic: a small team carrying a platform that, at the leaders, dozens or hundreds of engineers carry.

Three ways to get a strong site

The choice is more than build or buy. There are three practical paths:

  • Build custom. Total control and unlimited room to differ, paid for with a standing engineering team, the longest time to launch and full responsibility for security.
  • Rent a closed SaaS platform. Fast to launch, with a small team and vendor-managed security. In return, checkout and core behavior follow the vendor's rules, costs often grow with sales through transaction fees and paid apps, and leaving later is hard.
  • Own the core with an open, headless platform. Buy a maintained engine whose core is open source, and build your own storefront on top through its interfaces. You keep the code, the hosting account and the data. Your developers work on the experience instead of the plumbing.
TopicBuild customRent closed SaaSOwn the core (open, headless)
Control of checkout and cartTotalLimited; follows the vendor's rulesFull; open-source code
Room to differUnlimitedWithin the platform's rulesYour own headless storefront
Team to run itLarge standing teamSmallYour developers plus a partner
Time to a new siteLongestFastFast
Security patchingAll on youVendorVendor core; partner operations
Lock-inNoneHigh; fees and appsLow; you own the code and hosting
AI and agent readinessAll on youVendor roadmap, on vendor termsVendor roadmap, on open interfaces
Run-cost profileSalaries, indefinitelyFees that often grow with salesLicense plus hosting

IQ4hire assessment. Terms vary by vendor and contract.

Each path is right for someone. A retailer with a standard catalog and a standard checkout, and a premium on speed, may do best on closed SaaS. A company whose platform is the product, such as a marketplace with its own transaction model, may need to build. The middle path fits retailers that must stand apart on experience but cannot fund a platform engineering team, which describes most growing specialty retailers.

The middle path has costs too. A headless storefront still needs a front-end team, usually an in-house developer working with an implementation partner. And open source changes the license, not the operating model: someone still has to host, secure, monitor and upgrade it. Put those costs in the comparison, next to the salaries a custom build would need.

Build the difference, buy the rest

In this model you still build, in the right places. Your team writes the storefront, the part customers see, designed around your products instead of a template. On top go the features that are yours alone: visual product search, instant buy-back or trade-in offers, consignment, and links to the systems that run your back office.

The platform carries the commodity core: catalog, pricing rules, cart and checkout, orders and accounts, content, business-to-business accounts, and the interfaces that serve both your website and a mobile app. The vendor and its community maintain that core, so your developers move from maintaining plumbing to building what sets you apart.

In a headless model the retailer builds the storefront and the distinctive features, and the platform carries catalog, cart, checkout, orders and content. (opens the full-size image in a new tab)

The AI layer is moving into the platform

Until recently, AI shopping was something a retailer had to build. Leading open commerce platforms are now putting agentic AI into the core in three forms:

  • Agentic customer experience. Storefronts that AI assistants and shopping agents can read and act on: checking availability, building a cart and moving toward checkout, with the customer confirming the purchase.
  • Agentic business-to-business buying. Reorders and quote-to-order flows that an agent carries most of the way, with a buyer approving the decisions that matter.
  • Agents for the merchant. Assistants that draft changes to catalog, pricing and promotions for a person to review and approve.

An agentic shopping experience is exactly what an IT team will propose to build from scratch. It is becoming part of what the platform maintains. Some of these features are generally available, and some are in beta or on roadmaps, so confirm what is live, in which edition and on what terms before you count on it.

Match what competitors offer, without their budgets

The test of a platform is whether it can deliver what the leaders offer. On a capable open, headless platform, most of it maps to configuration or a focused build:

Competitor featureHow you get itEffort
Native mobile appAn app on the same commerce interfaces as the websiteBuild
Instant buy and trade-in pricesYour pricing data plus the platform's rules engineBuild and configure
ConsignmentA flow built on standard orders and payoutsBuild
Rich filtering and AI searchKeyword plus vector searchConfigure and build
Storytelling and editorial contentThe platform's page and content builderConfigure
Dealer and partner salesBusiness accounts, quotes and price listsConfigure
Virtual appointmentsDigital sales rooms for guided sellingConfigure
AI shopping assistant and agent-ready checkoutThe platform's agentic features and agent-ready interfacesConfigure
Agentic business reorders and quotesPlatform agents with buyer approval, where availableConfigure

IQ4hire assessment of capabilities on leading open, headless commerce platforms.

The "build" entries fall on the features customers notice. That is where your scarce development time should go.

Common concerns, answered straight

"Off the shelf will make us look like everyone else." The storefront is headless and designed by your team. Customers see your experience, not a template.

"We will be locked in to a vendor." With an open-source core you own the code, the hosting and the data, and you can change implementation partners.

"Our developer could build this." Yes. But not maintain checkout, tax, search, security and an app alone while also delivering the wish list.

"Our products do not fit a standard catalog." Custom fields and rules can model attributes such as reference numbers, condition, provenance or documentation without changing the core.

"Competitors built their own, so we should too." They fund engineering teams of dozens to hundreds. You win on inventory, expertise and trust, not engineering spend.

"We need our own AI agent to stay competitive." Maybe not. Agentic shopping and business buying are moving into the platforms. Build agents only for what is unique to you, such as pricing your own inventory.

"Licenses cost money every year." So do developers. Negotiate a fixed license in writing and compare it with the salaries a custom build requires.

Three questions for leadership

  1. Where do customers choose you? Build there, and only there.
  2. What can your team responsibly maintain? Count the people who will patch, secure and support the platform in year three, not only the people who will launch it.
  3. How much control do you need over checkout, data and hosting? If the answer is "a lot", renting may cost more than it saves. If it is "not much", renting may be the fastest path.

What this means

For most growing specialty retailers, those answers point to one place: own the core, build the difference, and let a maintained platform carry the rest. Compete on experience, not on engineering headcount. And when someone says "we should build our own platform", ask which part of it your customers would miss.

How we are paid: Suppliers don't pay to be listed. If you buy through us, the supplier you choose pays IQ4hire, and we tell you before work begins. That includes commerce platforms of the kind discussed here. We still tell clients which layers not to buy, and when building is the better answer.

Related: "I can build that": who owns it after launch? Digital Sales Rooms Digital commerce solutions

Sources

← All Insights