The Future of service companies in an AI first world

I'm going to skip the preamble.

The IT services industry as you know it is in serious trouble. Not in ten years. Not when the next wave hits. Now. And if you don't fundamentally redesign how your service company operates, not optimize it, not digitally transform it, not add an AI layer to it, you will be on the wrong side of the most brutal industry shakeout since the dot-com collapse.

I want you to understand exactly why I believe this, and what you need to do about it.

THE MODEL IS BROKEN

The services industry was built on a simple₹ idea: clients have technology problems, you have skilled people, you sell those people's time at a margin. For decades, that worked. Offshore labor arbitrage made it work even better.

AI has structurally destroyed that logic.

The work that took ten engineers now takes two, with the right AI tooling and one engineer who knows how to use it. Clients know this. They are already asking why their invoices look the same when the technology has changed this much. They are building internal AI capability. They are expecting us to do more for less, and they are right to expect it. The vendors who can't deliver that are already losing contracts. Are you one of them?

Meanwhile, AI-native competitors are entering your space without your cost base, without your legacy delivery model, and without your hesitation. They are not trying to fit AI into a headcount model. They built around AI from day one. They will undercut you on price and outperform you on speed, and they will do it at margins you cannot match if you stay the way you are.

The brutal truth: you are currently structured to sell something the market is rapidly devaluing.

WHY INCREMENTAL CHANGE WILL KILL YOU FASTER

I want to address something directly, because I know it will come up: the instinct to adapt gradually. To pilot AI in a few projects. To retrain some teams. To add an AI feature to your portfolio and call it transformation.

That is not a strategy. That is a slower version of the same outcome.

Optimizing a horse-drawn carriage doesn't make it a car. If you apply AI to the existing model, you will compress your own margins, reduce your own headcount justification, and undermine your existing revenue streams , without ever building something new enough to replace them. you will disrupt yourselves into a smaller, youaker version of what you are today.

The companies that survive this era will not be the ones that adapted their old model the most elegantly. They will be the ones that had the courage to build a new one.

WHAT REBUILDING ACTUALLY MEANS

I am not speaking in abstractions. Here is what ground-up redesign looks like for you:

FROM SERVICES TO PRODUCTS

You have spent years selling expertise. The future is selling outcomes, embodied in repeatable, scalable products and platforms that you can own.

Productizing your services means taking what you know how to do and building it into tools, accelerators, and platforms that deliver value without requiring you to rebuild from scratch for every client. It means recurring revenue, defensible IP, and margins that do not depend entirely on how many people you can deploy. Pure service businesses in your space are being commoditized. Companies that own something, that have built something, are not.

You need to become a company that owns something.

FROM TIME-AND-MATERIALS TO OUTCOME-BASED CONTRACTS

Stop selling hours. Sell results. This is a fundamental shift in how you price, how you contract, and how you measure success. It will be uncomfortable with some clients and will require you to be genuinely confident in yur ability to deliver, which is exactly why you are also rebuilding your delivery model.

FROM LARGE TEAMS TO AI-HUMAN PODS

The era of large delivery headcounts as a proxy for capability is over. You need to restructure around small, high-leverage teams that use AI to do what large teams used to do, faster, and at higher quality. Utilisation rates as a measure of performance become largely obsolete. Value delivered per person does not.

FROM EXECUTING TASKS TO ORCHESTRATING INTELLIGENCE

Your talent model changes fundamentally. You need people who design systems, orchestrate AI, validate outputs, and manage outcomes, not people who execute repeatable tasks that AI now handles. This changes your hiring profiles, your training investment, and your definition of what a high performer looks like. This means hard decisions about roles, retraining, and hiring profiles. I will not pretend otherwise.

NOW, THE PART I NEED YOU TO HEAR MOST CLEARLY

We have been living in the AI era for three years. Three years in which every one of your employee has had access to AI tools, has seen what they can do, and has had the opportunity to engage with them seriously.

Some employees have built with AI, experimented with it, integrated it into how they work, and developed genuine instincts about what it can and cannot do. Most have not. And I have to be honest with you about what that means.

If someone has had three years to engage with the most significant technological shift in our industry's history, and has chosen not to, that is not a skills gap anyone can easily close with a training programme. It is a signal about adaptability, curiosity, and appetite for change. These are not things you can afford to bet the company's future on.

But here is the more uncomfortable truth: not all roles are equally exposed, and knowing the difference matters. There is a useful distinction between intelligence work and judgement work. Intelligence work follows rules, it is complex, but it is learnable and repeatable. Writing code to a spec. Drafting a contract. Processing a billing cycle. Triaging an IT alert. AI has already crossed the threshold where it can do most of this autonomously. Judgement work is different, it is built on years of experience, domain instinct, and taste. Knowing what to build, not just how to build it. Knowing when a client relationship is at risk before the data shows it. That kind of judgement is not automated yet. It is, for now, the moat.

If your value to this organisation sits primarily in the intelligence layer, executing well-defined tasks reliably, you are in the most exposed position. If your value sits in judgement, in the instincts, relationships, and strategic thinking that AI cannot yet replicate, you are in the strongest position. The honest question every person on this team should be asking themselves is: which one am I?

The window for sitting on the fence is closed. Every leader and every team member is expected to be actively building, using, and thinking with AI. Not planning to. Not curious about it in theory. Actually doing it.

If you are not there yet, start today. That is not a suggestion.

And if you are honest with yourself and know that this is not the direction you want to go, I respect that, and I would rather have that honest conversation now than in twelve months when the gap has widened further.

I KNOW WHAT I AM ASKING

This is hard. It will disrupt teams, roles, client relationships, and internal power structures. There will be people in this company, good people, people who have contributed enormously, whose current roles do not exist in the model you are building. That is a painful reality and I am not going to dress it up.

But I want you to weigh that difficulty against the alternative: a company that declines gradually, loses contracts quietly, cuts costs defensively, and eventually faces a far more chaotic and uncontrolled reckoning with far fewer options. That is not a more humane outcome. It is just a slower one.

Chosen transformation, on your terms, with your people leading it, that is the better path. It is the only path I am willing to lead.

WHERE you GO FROM HERE

I have already had multiple conversations with the leadership teams about how to redesign the roadmap , the commercial model, the delivery model, the talent model, and the financial targets that go with each.

The AI era is not coming. It is here. The only question is whether you are building for it or being buried by it.

I choose building. I expect you to, as well.