Time Is a Weak Proxy for Value

An hour was never the best measure of value, but it was a perfectly good measure of effort, and for most of my career the industry agreed that effort was close enough. So did clients, which is the part that tends to get forgotten.

Both sides had a use for it. On the client side you would look at the number of people involved and the time it was going to take, and form a rough judgement about whether that represented decent value for money. Inside the firm it was doing something else entirely, because it was the engine of the business. You planned against the hours you expected to bill, the utilisation of each person and the rate they carried, and you checked that the overall economic model still worked. The hour was not just how we charged clients. It was how we ran the place.

AI has made that metric either unworkable or, at the very least, far more disputed than it has ever been. The arithmetic is simple enough. If AI lets you deliver the same piece of work faster, you bill fewer hours, and fewer hours means less revenue for a client result that has not changed at all. So you either find other work to fill the time you have freed up, or you move the rate to reflect the fact that you will be billing fewer of them. If you do not fix that quickly, your revenue falls while the client gets exactly what they were getting before.

None of this is new

The industry has been circling this for years, and it is worth remembering how the last attempt went. Back in 2011 in the UK there were projects being delivered where the firm put its fees at risk, and I saw enough of them to know what that tended to mean in practice. The margin was secured first, and the at-risk element sat on top of it, topping up profit against outcomes that were loosely defined. Most of those outcomes were outputs wearing an outcome's clothing: something being created, rather than the value the thing created once it existed. It looked like accountability, and it was mostly presentation.

The easy question is already settled

Whether firms will move off the billable hour is not really in dispute. I went through what the 2026 results season did to consulting valuations in AI Isn't Killing Consulting. The Market Is Repricing It., and those numbers make the direction obvious enough. The harder question is the one sitting underneath them: if everyone can now see that the hour has stopped working, why are so few firms actually able to leave it?

Deloitte gave its own people a stark version of the first half. The Wall Street Journal reported on a town hall where a leader in its US government consulting practice presented a chart showing traditional labour-based, hourly-rate work shrinking to a sliver of the market by 2035, with AI agents taking the majority of a larger overall pie. One consultant's summary of the session was that the model is "toast" and they are basically being replaced by robots. Business Insider, separately, found clients actively moving their advisers off hourly and fixed fees towards arrangements where the firm carries real risk.

I have a lot of sympathy for that consultant and I would not file the reaction under pessimism. It is exactly what I wrote about in Fear Is an Operating-Model Issue: when the people doing the work can see the model changing and nobody has drawn them a place in the new one, you get fear, and the fear is information.

The hard part, which has been a problem for far longer than AI

Here is the thing about hourly rates and utilisation. I remember being in a partner presentation, talking about utilisation, when a very senior partner cut in: who cares about the utilisation, what were the hours actually spent doing? The penny dropped. More important than anything else was whether the hours being spent were going into the right thing.

Looking back, for years firms have treated utilisation, the share of a person's time that is chargeable, as a headline measure of health, and it is a badly flawed one. If those chargeable hours are going into the wrong work, the outcome does not arrive and no value is created, but the metric still reads beautifully. You can run a fully utilised team all the way to a disappointed client and never see it in the numbers until the relationship is already in trouble. AI has not created that problem. It has just made it impossible to keep ignoring.

Moving off hourly models brings its own risks, though, and they are real rather than theoretical and almost everyone is already here. We've all agreed a fixed-price deal, the client makes some difficult decisions along the way, the project overruns, your change request gets rejected, and the overrun is now something you're trying to manage during the monthly reviews, knowing there is really no way to recover it within that project. But it gets worse, the cash flow gets harder too, because you are waiting for deliverables to be signed off or outcomes to be proven before you can invoice anything at all. That is a long way from the smooth monthly invoice that time and materials gives you.

Clients also have not all caught up. Baker Tilly's chief executive Eric Miles points out that buyers still compare competing bids on an hours times day rate basis even when hours form no part of the pricing you offered them. So you can do the hard work of redesigning your commercial model and still be marked against the rate card you are forced to put in the appendix to meet the procurement team's spreadsheet jockey doing the final reviews. Changing what you sell is not enough on its own, because the buying side has to change with you.

Pricing is downstream of the operating model

All of which points at the thing I think most firms are getting wrong. They are treating this as a pricing change when it is an operating-model change.

Pricing is not an independent lever. It falls out of how you designed the firm in the first place, and out of the measures the leadership actually watches to decide whether the business is on track and whether the future looks any good. Work backwards through that chain and you get sales, then revenue, then predicted utilisation, then rate per hour (built up from your staff costs, overheads etc.), and every one of those is a choice made in the operating model rather than in the pricing conversation. So if you think AI means you need to change how you price, you have understood about a fifth of it. It changes how you value the people you employ, what you are willing to pay them, how you staff and deliver the work, what you deliver vs. out-source to partners and only then how you price an engagement and how much risk you are prepared to carry inside it.

You can see the same thing from the other end, in the shape of the working week. If you are paid for hours, you are paid to arrive, diagnose, present and leave, and the whole firm is built around that arc: how you staff an engagement, when you rotate people off it, what a partner is measured on, and where the work is considered finished. If you are paid for an outcome, almost none of that is relevant. You stay well past the point where the old model would have sent the invoice, you need people who can operate rather than only analyse, you have to measure the thing you said you would move even in the months when it has not moved, and someone has to stay accountable long after go-live, which is precisely the stretch of time the old model was built to avoid.

The best evidence for that is the firm that has actually done it. Newton, the UK consultancy formerly known as Newton Europe, has been built around this since it formed in 2001, and its commitment is unusually specific: it guarantees 100% of its implementation fees against delivering outcomes the client can measure. The outcomes it signs up to are refreshingly unglamorous, things like reducing manufacturing waste, taking cost out of warehousing and trucking, cutting the volume of fresh produce thrown away in retail and lifting store conversion. That is what an outcome looks like when it is real. It is a number inside somebody's operation, not a milestone in your workplan. Even Newton still bills a small share of clients by the hour, usually where a procurement team insists on it, which tells you how strong the pull of the old yardstick remains.

What Newton's managing partner Steven Phillips says about everyone else is the line I would pin to the wall. He doubts the legacy firms can pivot to this at all, because they lack the culture for it and were never built that way. James O'Dowd, who runs the talent advisory firm Patrick Morgan, makes the structural version of the same point: if your business is built around utilisation, moving off it means redefining what you actually do, and that is an expensive transition rather than a pricing tweak.

I also came across another example that stood out. Elevate, an AI-native accounting firm, pays its people to stop billing time, rewarding junior staff for getting into the weeds of integrating AI and mapping new processes instead of hitting billable-hour targets. That is somebody redesigning the incentive that actually drives behaviour, while most of the industry is redesigning the invoice and hoping the behaviour follows.

Every attempt at outcome pricing I have watched fail has failed because of the way people behave. The commercial team writes an outcome-based contract, the delivery model underneath it goes untouched, and the firm ends up carrying an operator's risk while still working like an adviser. That is not a new business model, it is the old one with worse economics. It is the argument I made in The New Consulting Premium Is Deployment Muscle approached from the other side: there I was talking about what a client should be buying, here about what a firm has to become before it can credibly sell it. I made that move myself, and the part that genuinely caught me out was not the work, it was the accountability, which I wrote about honestly in From Consulting to Building.

Your takeaway

The uncomfortable question is not whether you should move off time & materials deals, because that time already came, but it's whether you could actually deliver an outcome if a client agreed to pay you for one tomorrow, and more importantly, how would you decide?

So take a live engagement, the real one you are delivering, and ask three things about it.

1. What is the outcome this client would genuinely pay for? It needs to be a number inside their operations rather than a deliverable in your plan.

2. What would have to change about how you staff and run the project in order to own that number with confidence?

3. Who is still standing behind it six months after the final presentation, and how are you taking their time into account in your calculations for the deal?

If you can answer all three, the pricing conversation becomes straightforward. If you cannot, no amount of contract redesign is going to save you, because the problem was never the price. Time was only ever a proxy for value, and it has stopped being a good one.