Retec Group CEO and Co-founder Chris Griffiths: retail technology consulting has a legacy system too
Retec Group CEO and Co-founder Chris Griffiths explores how AI is reshaping the consulting space, why legacy operating models are holding firms back, and what the future of transformation looks like.
Here’s a parallel that every retail technology leader will recognise immediately. The single biggest challenge in delivering transformation is not setting the technology strategy or choosing the right platforms. Instead, it’s the complexity of change inside an organisation.
Organisations where decades of accumulated systems, integrations, and decisions that made sense at the time, are now load-bearing walls you cannot remove without bringing something else down. The legacy estate is not just a technical problem: it’s an organisational one. The business processes, the informal relationships, the institutional knowledge baked into how an organisation behaves: these are harder to replace than the technology itself.
The large consultancies face an identical problem with their own operating model. The governance layers, the approval hierarchies, the utilisation targets, the account structures. These are their legacy estate. They know it needs to change and many are genuinely trying. But structural problems of this scale cannot be resolved by good intentions alone.
Modernising from the inside - while keeping the lights on for existing clients and managing the politics of change at scale - is the hardest thing to do. The retail industry knows this better than most.
The number of large scale transformation programmes that have delivered their original promise, on time and on budget, does not reflect well on the industry. And the organisations attempting those transformations are, very often, being guided by the same firms now trying to transform themselves.
Why change is harder than it looks
The consulting industry’s response has been predictable: AI centres of excellence, innovation labs, and rebranded service lines with ‘AI’ somewhere in the name. The intent may be genuine. But structural problems are not solved by renaming them.
Large organisations accumulate what could be considered ‘organisational scar tissue’. In effect, every failed project produces a new governance process, every risk event generates a new approval layer, every growth phase adds a management tier that, when growth reverses, proves almost impossible to remove.
Over time, the organisation becomes less a vehicle for delivering outcomes and more a system for managing its own complexity. One where processes designed to protect quality end up preventing it. As a result, approval chains outlast the problems they were built to solve, and the most talented people quietly start to leave.
AI has accelerated all of this. Analysis, documentation, reporting, coordination - this is precisely what AI does well. Not perfectly, and not without oversight, but well enough and fast enough to fundamentally change how many people you need to deliver transformation and change.
A billing model built on hundreds of heads becomes structurally indefensible when that headcount can be augmented by a well-configured set of tools. Tens of thousands of roles have been cut across the major consultancies in the past 18 months. This is often at the same firms simultaneously winning AI transformation contracts from the clients whose programmes those people were delivering. The arithmetic is telling.
Greenfield is always easier. That is not a criticism of anyone. It’s simply true.
A billing model built on hundreds of heads becomes structurally indefensible when it can be augmented by a well configured set of tools.
What replaces the pyramid
When you build from scratch, you make different choices. Not because you’re smarter or more virtuous, but because you’re not constrained by what already exists.
You don’t inherit governance structures designed for a business ten times your size. You don’t have utilisation targets that incentivise billing over outcomes. You don’t have account hierarchies that put the wrong person in the room. You build what you actually need for the environment you’re operating in.
In the AI era, that means small, senior teams with deep domain expertise and genuine proximity to the client’s business. Teams who can move at the speed the problem demands, not the speed the governance structure permits. Augmented by AI tools that compress analytical and coordination work which once required large teams, senior practitioners are freed up to focus on judgement, relationships, and outcomes.
This model only works with the right culture underneath it. A high calibre team operating under a command-and-control hierarchy isn’t a high-calibre team, it’s a small bureaucracy.
Culture is the delivery model
Building this kind of firm requires deliberate choices at the start; about who you hire, how you structure authority, and what you are willing to say no to.
But what does this actually look like? For starters, slow hiring for alignment; not just on capability, but on the values that make high-performing teams function. Teams built on trust, accountability, the willingness to tell a client something they don’t want to hear, and the commitment to stay until the work is done, not until the statement of work expires.
Senior practitioners who have lived the problem from both sides, as consultants and as the client side leaders accountable for outcomes when those consultancies fell short.
That combination: genuine expertise, genuine proximity, and a culture that doesn’t have to be rescued from its own legacy, is what the next generation of consulting firms will be built on. The evidence, where it exists, is not found in credential decks. It shows up in staff retention, in NPS, in whether the person who sold the engagement is the person doing the work.
It’s also, it turns out, what makes people want to come to work. When Retec was named a Sunday Times Best Places to Work this year in our first year of operation, it wasn’t a surprise to us. It was confirmation that the model we’ve built is working in the way we intended.
What comes next is smaller, faster, with a culture that didn’t have to be rescued from it’s own legacy.
The questions worth asking
The retailers who navigate the next five years well won’t simply be those who adopted AI earliest. They’ll be those who were clear-eyed about what kind of partner they needed.
Before you sign the next statement of work, ask your systems integrator what their staff retention rate looks like on client programmes. Ask how many of the people who started your programme are still on it. Ask whether the person presenting the proposal is the person who will actually do the work.
The answers will tell you more about your likely outcomes than anything else in the proposal.
The consulting pyramid had a good run. AI will end it. What comes next is smaller, faster, and built on genuine expertise, genuine proximity, and a culture that didn’t have to be rescued from its own legacy.
About the author: Chris Griffiths is CEO and Co-founder at Retec Group, a boutique retail technology consultancy working with some of the UK’s leading retailers and hospitality operators. Retec was named a Sunday Times Best Places to Work 2026.
