What does your post-AI operating model look like?

Every business leader I speak to agrees that AI will change their business. The conversations I have in boardrooms and at conferences are remarkably consistent on this point. Yes, AI will change how we deliver our product or service, yes, it will change what our clients expect, and yes, it will change how we compete. The consensus is almost universal.

What isn't universal, and what I find striking every time I ask the follow-up question, is whether anyone has actually redesigned their business model to account for it. When I ask that question to a room full of leaders, it's rare that anybody raises their hand.

That gap, between knowing change is coming and doing the structural thinking to prepare for it, is where most businesses are sitting right now. And the longer they sit there, the more disruptive the eventual reckoning will be.

The billing model problem

Let me share a concrete example, because abstract conversations about AI and business models tend to stay abstract unless you put numbers and real professions into them.

Take a law firm. For decades, legal billing has been built around time. Partners and associates charge in six-minute units, and a significant portion of those units has historically been reading time. Reading contracts, reading correspondence, reading case law, reading the documents the other side has produced. That reading time is billable. It is, in many firms, a substantial part of revenue.

AI can now read a 300-page contract in seconds, summarise the key clauses, flag the risks, and produce a briefing note that would have taken a junior associate several hours to prepare. The AI doesn't charge by the six-minute unit. It charges by token, a unit of processing that costs a fraction of a penny.

So, the question for that law firm is not whether AI will affect their business. It clearly will. The question is what their billing model looks like when the reading time disappears, when the first-draft time compresses, when the research time that used to justify three hours of associate billing now takes four minutes and costs less than a cup of coffee. Do they pass that efficiency on to clients, absorbing the margin reduction? Do they charge the same and pocket the difference, at least until a competitor doesn't? Do they redesign their pricing around outcomes rather than time? Do they use the freed capacity to take on more clients, deliver more value, or develop new services?

These are not hypothetical questions. They are live commercial decisions that law firms, and every other profession that charges for time and expertise, need to be making now.

It isn't just law

The legal profession is a useful illustration because the billing model is so explicit, but the same dynamic applies across almost every knowledge-based business.

Accountancy firms charge for the hours their people spend preparing returns, reviewing figures, and advising on tax. AI compresses all of that. Consulting firms charge for the research, analysis and synthesis that their teams produce. AI compresses all of that too. Marketing agencies charge for the thinking, writing and creative production that their people deliver. Architecture practices charge for design time. Recruitment firms charge for sourcing and screening. Financial advisers charge for research and recommendation.

In every case, the question is the same. If AI reduces the time required to deliver what you currently sell, what happens to your pricing? What happens to your margins? What happens to your headcount? What happens to your value proposition when the thing you've been charging for becomes faster, cheaper, and available to your clients directly if they choose to use it themselves?

These aren't reasons to be pessimistic about AI. They're reasons to think carefully and act deliberately, because the businesses that work through these questions now will be in a fundamentally stronger position than those that wait until the market forces the conversation.

The token cost question

There's a further layer to this that most business leaders haven't yet reached, and I think we need to talk about it because it will become increasingly significant.

As AI tools become embedded in professional workflows, they introduce a new cost line that didn't previously exist: token consumption. Every document you feed into an AI tool, every prompt you run, every output you generate, has a cost measured in tokens. At the moment those costs are relatively small, but they scale with usage, and as AI becomes more deeply integrated into how businesses operate, the aggregate cost becomes meaningful.

For businesses that are considering passing AI-related efficiency gains on to clients, or building AI usage into their service delivery, the token cost needs to sit alongside the people cost in any honest assessment of margin. A firm that replaces three hours of associate time with an AI workflow needs to account for what that workflow actually costs to run, not just assume it's free because the subscription is paid.

This is new commercial territory, and most finance directors haven't mapped it yet. The businesses that build accurate models of their AI cost base now will make better pricing decisions than those that discover the numbers later.

The boardroom conversation you need to have

None of this requires a crystal ball. It requires a willingness to ask uncomfortable questions in the right room with the right people present.

What proportion of our current revenue is directly tied to time that AI will compress or eliminate? How quickly is that likely to happen in our sector? What would our margins look like if we delivered the same output in half the time? What would they look like if our competitors did, and we didn't? Are there new services we could offer, or new clients we could serve, if our capacity weren't consumed by work that AI can now handle?

These are strategic questions, not operational ones. They belong in the boardroom, not delegated to whoever is running the AI pilot. And they need to be answered before the market answers them on your behalf.

The opportunity in the disruption

It would be easy to read this as a warning, and in some respects it is. But the other way to look at it is that AI is offering most businesses a genuine opportunity to redesign themselves around the things that actually create value, rather than the things that have historically been easiest to charge for.

A law firm that builds its model around judgment, relationships, and outcomes rather than reading time is a more valuable law firm. A consulting firm that uses AI to deliver better analysis faster, and prices accordingly, can serve more clients with the same team. A financial adviser who spends less time on research and more time on the human conversation about what a client actually needs is a better financial adviser.

The post-AI operating model isn't just about protecting margin. It's about asking what your business is really for, what your people are really good at, and how AI can free them to do more of it.

That conversation is overdue in most boardrooms. The good news is that it's not too late to have it. The less good news is that it won't stay that way for long.

Frequently asked questions

Why do knowledge based businesses need to redesign their operating models for AI?

Many knowledge based businesses charge for time spent on tasks like reading, research, and analysis. AI compresses these tasks from hours into minutes or seconds. If firms do not redesign their business models, pricing, and operating structures, they risk losing margins or falling behind competitors who adapt to these efficiency gains faster.

How does AI affect traditional billable hour models, such as in law firms?

Traditional billable models rely heavily on charging clients for time spent reading, researching, and drafting. AI reduces hours of work to minutes at a tiny token cost. Law firms must now decide whether to pass savings to clients, shift to outcome based pricing, or use freed capacity to serve more clients.

What is token consumption cost, and why does it matter for AI strategy?

Token consumption is the cost incurred every time an AI tool processes data, prompts, or outputs. While individual tokens cost fractions of a penny, costs scale with usage. Finance directors must account for token expenses alongside workforce costs to understand true profit margins as AI becomes embedded in daily workflows.

Which strategic questions should business leaders ask in the boardroom about AI?

Leaders should ask what proportion of revenue relies on time that AI compresses, how margins will change if outputs take half the time, and how competitors might react. They should also explore new services or clients they could support once staff are freed from repetitive research and drafting tasks.

How can businesses turn AI disruption into a commercial opportunity?

Rather than just protecting margins, businesses can use AI to redesign their operations around value creation. By letting AI handle time consuming tasks, professionals can focus on human judgment, client relationships, and outcomes. This allows firms to deliver better work faster and serve more clients with the same team.