Contact us
There's a question that sits at the heart of almost every major technology solution decision: do we build it ourselves, or do we buy something that already exists?
It sounds simple, but in practice it's one of the most consequential calls a business can make and getting it wrong is costly in ways that can take years to fully understand. This is part of the challenge. With a keen eye on the next quarter or even next year’s results, an obligatory near-term bias drives what are consequential long-term decisions.
Archie Norman, chairman of Marks & Spencer (M&S), gave one of the more candid answers on this topic at the Retail Technology Show in 2025. He raised it again at the same event a year later, such is the importance of build vs buy decisions to M&S’s successful turnaround. Speaking about the culture he inherited when he joined M&S in 2017, he said: "Every bit of technology had to be redone for them to fit their systems in the past. No, I said, that ends up with a very overcomplicated company."
That's a chairman describing a business that had spent years building bespoke solutions where off-the-shelf ones would have done the job without the complexity and overhead that comes with building and maintaining inhouse solutions.
M&S isn't unusual. The instinct to build is embedded in large businesses, and retailers are no exception. The logic goes: we're different, our processes are complex, and if we buy something generic, we'll compromise on what makes us unique and successful. There's some truth in that. But there's also ego and tradition wrapped up in it.
When a retailer builds its own technology, it is designed to fit the operations and priorities of the business. There's no workaround for a feature the vendor didn't build, no waiting for an update that solves a problem the vendor hasn't prioritised, and no dependency on a third party's roadmap.
For genuinely differentiating capabilities, building can be the right answer. If your supply chain logic or your loyalty mechanics are part of what makes you better than competitors, and loved by your customers, handing that over to a third-party platform means others can have the same thing and the competitive advantage erodes.
For retailers operating at scale with complex, distinctive operations, there are real moments where building makes sense. Amazon is a good example. Its fulfilment and recommendation systems aren't off-the-shelf, and the advantage they create is material. But Amazon also has engineering resources that most retailers will never have.
The risk is applying this logic too broadly, building things that don’t need to be built to maintain differentiation in a competitive market and paying a high price as a result.
Buying technology that already exists is faster, cheaper upfront, and comes with a team of people who've spent years making it work. The vendor has already found the bugs, handled the edge cases, and built the integrations. You're getting a product that other businesses have already tested in production and proven it works.
There's also the question of ongoing cost. Building an enterprise grade solution isn't a one-time investment. It needs maintenance, updates, security patches, and people who understand every aspect of it. That’s just to keep it functioning. Then there’s new features and enhancements required to meet the evolving needs and expectations of the teams using it.
This can all be acknowledged, and even manageable, for a time. When the individuals that built a solution move on, the intricate knowledge, the ‘IP’, often goes with them. The same can be said for users. New starters need to learn a custom solution, and can’t hit the ground running with software skills learned at previous employers. What looked like control starts to become a liability.
Norman made this point. IT costs at M&S were heading well north of £200 million the following year, driven by the explosion in cloud data storage costs, and the compounding bill of years of bespoke development. The more custom the system, the more expensive it is to run and enhance.
There's a broader point here too. The pace of development has accelerated so much that trying to build advanced solutions across diverse areas of the business puts retailers in a race they're unlikely to win against dedicated technology companies. Buying the best solution available and integrating it well is almost always faster and delivers anticipated benefits sooner than building from scratch. It also avoids an ever-increasing burden of technical debt and outdated solutions.
A common mistake is treating build vs buy as a philosophical stance rather than a practical question. A retailer can accumulate a patchwork of vendors that create integration headaches along with ballooning IT expenses. Others opt to build everything, running up internal costs and complexity that become too demanding to manage.
The culture Norman inherited at M&S was more like the latter. Vendors were asked to rework their products to fit M&S's systems. Every solution and integration was custom. The result was an organisation that was expensive to run and slow to change.
Bespoke systems can become single points of failure in ways that enterprise grade platforms are not. With cyber attacks a very real threat, this is an increasingly important aspect to be managed. Vendors of widely used software have strong commercial incentives to maintain security and resilience.
With all this in mind, the right starting point isn't 'should we build or buy?' It's 'where do we actually need to be different?'
Most retailers don't need a bespoke point-of-sale system. They don't need a custom-built supply chain platform, a homegrown warehouse management system, or their own customer marketing infrastructure. There are excellent, proven solutions for all of these. Buying them frees up resource, budget and management attention for the things that genuinely matter to creating competitive advantage in the market.
What a retailer might legitimately want to build, or heavily customise, are the capabilities that are unique to its model and hard for a vendor to replicate. The way it sources product. The logic behind its merchandising. The experience it creates for customers that no platform can fully deliver out of the box.
Even here, the answer is increasingly buy the platform and configure it well, rather than write it from scratch. The quality, flexibility and affordability of modern technology platforms have improved dramatically. The perceived customisation and affordability gaps between a bought solution and a built one have narrowed in many categories.
Norman's transformation of M&S's technology culture is a useful case study because he came in from the outside, saw the costs in all its forms, and was blunt about what had to change. He didn't approach it as a technology question. For him, it was a business question: why are we making this so complicated, and what does it cost us?
Technology decisions in retail are business decisions. The choice between build and buy should come down to where you genuinely create competitive advantage, what it will actually cost over time, and how quickly you want to start gaining benefits from the solution.
You may be thinking this all sounds good in theory, but is the new direction delivering results for M&S? After rebuilding profitability M&S returned to the FTSE100 in 2023. From 2023 to 2025 productivity across their store network continued to grow with sales per sq ft up 19% and operating profit up an incredible 65%. The subsequent market enthusiasm drove the share price up a further 260%.
Buy broadly, configure smartly and build where you're genuinely different. For most retailers, it's the better strategy. Just ask Archie Norman.
If you'd like to talk through a build vs buy challenge in your business, get in touch and we can help you work it through.