AI Killed the Billable Hour
The firms that figure out what they're actually selling will charge more, not less.
Hi friends,
This week we welcomed Shawn Yeager from Upshift to The Good Stuff. Shawn has spent years helping professional services firms rethink how they price and package what they do, and the conversation went somewhere deeper than we expected. It’s definitely one to check out if you’re thinking about how AI might impact your business model.
Three news items this week, including data showing how quickly the shift away from hourly billing is already registering on the client side of professional services engagements.
The Free AI Audit we launched last week is still live at otherstuff.ai/ai-audit. If you're trying to work out what AI means for your business, it's a useful place to start.
Now to this week’s Good Stuff.
AI Killed the Billable Hour.
If AI cuts the client delivery costs of a professional services firm by 30%, who gets to capture that 30% - the professional services firm or the client?
For any firm with a billable hour, the answer is generally the client.
This week we welcomed Shawn Yeager from Upshift onto The Good Stuff and this was one of the questions we got into around how AI is killing the billable hour.
Shawn explained that clients aren’t waiting for cheaper versions of what you already sell. They’re waiting for something better. The firms that deliver that, and price it on outcomes, will likely charge more, not less.
This is why Wachtell, Lipton, Rosen & Katz is consistently among the most profitable law firms in the world, measured by revenue per lawyer, and they’ve achieved this without the associate pyramid most BigLaw firms depend on.
They take on a small number of high-stakes matters, deploy lean teams, and charge fees tied to the size and outcome of the transaction rather than the time it took. On a multi-billion dollar acquisition, a Wachtell fee can run to tens of millions of dollars.
Importantly, that value equation doesn’t rely on the billable hour.
Instead, Wachtell asks what it’s worth to a client to have the firm the other side doesn't want to see across the table?
For many firms the billable hour defined what they were selling.
Because a client can't always evaluate whether the legal advice they received was optimal, whether the tax position was sound, or a strategy was well-formed.
But they know what an hour is.
The problem with selling judgment, trust, and risk reduction is that none of these have obvious units of measure.
It’s hard to invoice for three units of certainty or put a line item on the value of someone who picks up the phone at midnight when everything's falling apart. So professional services reached for the nearest available proxy - time.
Pricing in hours solved an information asymmetry problem by re-pricing away from something abstract into something that everyone could easily measure.
That worked for a long time too, and then Goodhart's Law found it.
Goodhart states that when a measure becomes a target it ceases to be a useful measure, and the billable hour is one of the clearest examples of that.
Associates are evaluated on billable hours logged. Partners carried billing minimums. Whole performance management systems optimised for time, and value delivered to clients became a byproduct of that model, rather than its purpose.
Firms that resolved a client's problem cleanly and quickly were effectively penalised by a model that paid for time.
The same pattern is showing up in SaaS right now, as companies try to move from per-seat licensing to token consumption and discover they don't have good language for what they're actually charging for.
Nobody ever actually wanted to buy a seat. They wanted the outcome the software produced. The seat was just a proxy you could measure.
In that sense, the billable hour isn't uniquely a professional services problem. It's a measurement problem that shows up wherever the thing being sold is genuinely hard to quantify.
Thanks to AI, we're finally being forced to answer a question we've been avoiding - what is professional judgment and taste actually worth, and how do you price something that can't be easily measured?
Why Now is Different From the Last 15 Years of ‘Selling Outcomes’
Firms have tried to shift to outcomes pricing over the last fifteen years but it’s not been without its challenges. So why is it different this time?
Two things:
1. The compression is happening whether you price it or not.
Previously, a firm could say they sold outcomes but still staff five people on the engagement, and the client would pay for five people because that's what the invoice said. Now the work that took five people takes one person and AI. That time compression is now much more visible and the client can see they're paying for five and getting one.
2. The client is often showing up with 80% of the work done.
The client has already used AI to do a first draft of that contract, or a strategy document, and now they just want a judgment call. In this scenario, the billable hour collapses because the client stops generating a raw request large enough to substantiate it.
So the old "we sell outcomes" pitch failed because it was voluntary and theoretical. This time it's being forced, practically, from the demand side.
This is something any professional services provider will have to grapple with as AI continues to compress time to delivery.
Shawn highlights that you essentially have two options:
You can pass the cost savings to your client. As AI makes the work faster, your invoice gets smaller, and so the firm must shift to volume to make up for the shortfall. This feels like a losing game. Once every competitor has the same AI tools, and they will, the only differentiator is price.
You can repackage the savings as a new capability and charge for outcomes. The time AI frees up goes toward higher-value work. Things like analysis, judgment, and taste. Like Wachtell, the firm prices for the result - what the client is getting thats uniquely valuable - not just the time.
What AI is actually dismantling
It's not just that AI only made the work faster too, in some sense it’s also the case that AI made the work more visible.
The client can now see, albeit roughly, what the professional is actually doing. And what they're seeing may not match the invoice in the way it used to.
When Canva launched it didn't kill design agencies - but it did make an element of design more accessible.
Canva separated the commodity layer - things like layout, templates, basic visual coherence - from the craft layer - taste, narrative, brand judgment.
Agencies that had been selling the whole stack as an undifferentiated service suddenly had to explain why the craft layer was worth paying for separately.
The best agencies have always tended to exist in that craft, judgement and taste layer and charged more for it.
That's the same separation that's coming for legal, accounting, and consulting services. AI is going to make the commodity layer visible, and everything that survives has to clearly be the craft layer.
The firms that win won't necessarily be the ones with the best AI stack either. They'll be the ones that can finally answer the question the billable hour let them avoid - what are we actually selling, and what is it worth?
Wachtell has the answer to this question.
So does Pentagram.
Clients hire Pentagram because it’s a partner-owned, flat-hierarchy structure, where the actual owners of the business are the creators doing the work and each one has a public body of work that carries its own cultural weight. Unlike traditional design agencies, it has no CEO or central creative director. Clients work directly with the world-renowned partners themselves.
Clients buy the right to say "Pentagram crafted our identity" and everything that signals about their organisation. The product is the association and that’s not being commoditised.
The a16z model in venture is a version of this too.
Marc Andreessen's thesis when founding the firm was that capital was becoming a commodity so the product had to be everything around the capital - the network, the talent pipeline, the operational support, the brand signal of being an a16z company.
They defined what they were actually selling - a platform, not just a fund - and built a pricing model that captured the value of that.
It's not outcome pricing in the simple sense. It's something more like access-to-platform pricing.
This is what ‘better’ value may look like for clients. The firms that identify what that is and price accordingly will charge more, not less.
Shawn's point is that professional services firms are now in the business of business models, which means the firms that thrive will be the ones that approach what they sell the way a product company thinks about its product, working back from who the specific client is, what they actually need, and what a successful outcome looks like clearly enough to price it, with the confidence to stand behind the result.
AI makes that possible in ways it never was before.
Firms can deliver in hours what used to take weeks, which means the gap between the cost of producing the outcome and the value of it to the client is wider than it's ever been.
Judgment and taste were always what clients were paying for. The firms that define that clearly and price accordingly will find that AI made them considerably more valuable.
We got into this and more this week in the Big Episode 58 of The Good Stuff with Shawn Yeager.
Three Things in AI for SMEs.
1. Nearly two-thirds of professional services clients now prefer fixed-fee pricing over hourly billing — Margin Up
A Deloitte study on AI in professional services found 67% of consulting buyers now prefer fixed-fee arrangements over time-and-materials contracts, up from 41% three years ago. The shift is being attributed directly to AI making it easier for buyers to anticipate outcomes in advance and question why hours should be the variable. Separately, Wolters Kluwer's 2026 Future Ready Lawyer Survey of 810 legal professionals found more than half expect AI to push firms toward either higher volume or more competitive pricing structures within the next two years.
What it means for you: The client preference is moving before most firm conversations have caught up with it. Firms that get their pricing model right before being asked to will be better positioned than the ones scrambling to respond. This week's essay goes into why.
Consulting Success →
2. 74% of AI's economic value is going to just 20% of organisations — and the gap is widening — Capital Up
PwC's 2026 AI Performance Study, covering more than 1,200 senior executives across 25 sectors globally, found that the businesses capturing most of AI's economic value aren't simply deploying more tools than everyone else. They're using AI to pursue new revenue streams and go after markets that weren't accessible before. The remaining 80% are largely stuck in what PwC describes as "pilot mode," running experiments that haven't converted into financial returns.
What it means for you: Most of the AI conversation for SMEs still focuses on where to start. The PwC data suggests the more pressing question is whether what you've started is actually compounding into something. The businesses pulling ahead aren't running more experiments. They're running fewer, better ones that change how the business generates revenue.
PwC 2026 AI Performance Study →
3. The freelance platforms built on task-based work are losing ground to AI — and the freelancers holding their rates know why — Risk Down
Downloads of major freelance platforms have fallen significantly over the past two years as clients turn directly to AI tools for work they used to outsource. The freelancers losing ground are the ones whose pitch is essentially "I have a skill, hire me to use it." The ones holding their rates have stopped selling what they do and started selling what it produces, using AI to handle the parts that should have been faster anyway while keeping the strategy, the judgment, and the client relationships for themselves.
What it means for you: This is the freelance version of the same argument this week's essay makes about professional services firms. If what you're charging for is time and tasks, AI is already competing with you and the gap is widening. If what you're charging for is judgment and outcomes, AI is your leverage.
Self Employed →
That's all for this week.
If this resonated, we’d love for you to forward this newsletter to one person who might enjoy exploring these ideas too. See you next week!
Cheers,
Pete & Andy