Managing partners aren’t confused about AI anymore. Two years ago the quarterly meeting was all vendor noise and guesswork. Now they’ve read the reports, and clients have started asking whether AI should lower the fee. Nobody disputes where this is going.
Then the meeting ends, and the firm goes back to selling what it sold before.
It goes the same way in firm after firm. The firm puts AI on the quarterly agenda, everyone agrees something has to change, and nobody agrees on what. Someone proposes a flat-fee pilot, someone else asks who absorbs the write-downs, and they table the question until next quarter. In between, they go back to the work that pays this month’s draw.
Partners at one law firm described the vote that never gets called: “Partners want their PPP at the end of the day … that’s a difficult conversation that requires a lot of matrix political conversations.”
Before AI goes back on the quarterly agenda, write down the offering, scope, price and pilot client it must produce.
Six months of notes and the same service list
A decision has an owner, a scope, and a date. Agreement has a follow-up meeting. The difference is easy to miss from the inside, because everyone walks out with notes, next steps, and a volunteer to look into it. Six months of that and the firm owns a folder of concept documents and the same service list it started with.
For two years the working theory was that firms froze because they couldn’t see clearly. The technology was new, and the pricing pressure was easy to dismiss as anecdote. The reports landed and the pilots ran. They can now describe the path ahead in detail, and describing it is where they stop. In a Thomson Reuters Institute study of standout law-firm lawyers (August 2026), 36% strongly agreed their firm’s AI strategy is clear, and 25% strongly agreed the firm has a plan to make money from AI.
Why firms don’t execute strategies they believe in
In Agency Core 2026, White Label IQ and Audience Audit’s survey of marketing-agency leaders, roughly one in seven said they fully execute their own top-three strategies. Nobody handed those leaders their priorities. They named the strategies themselves, believe in them, and still aren’t running them.
Marketing agencies have the fewest structural excuses. Many are founder-owned, so one person can decide on a Tuesday and the firm moves Wednesday. Partnerships freeze harder, because the votes belong to the people the change costs first. An agency founder has to overrule one person’s caution. A partnership needs a dozen partners to vote against their own draw in the same meeting. In law firms and larger accounting firms, a change in what the firm sells runs through compensation formulas, origination credit, and equity tiers, and every partner who votes on that change is paid from the work it would cut.
Each partner, looking at their own year, makes a rational choice: protect the billable work that pays this year’s compensation. Add those choices up and you get a firm that struggles to act on what it already believes. They’re doing what the compensation system pays them to do. That’s why they get the same result when they run the conversation again, and why another agenda item rarely fixes it, even for a managing partner who already knows all of this.
Most managing partners have already tried to fix it from inside. A mandate comes out of a retreat or a rough quarter, a committee gets a name, and a respected partner agrees to lead it on top of a full client load. The group meets three times and produces a memo with real thinking in it. Then busy season arrives. By the time anyone looks up, the memo is eight months old, the committee hasn’t met since March, and raising it again would mean admitting it stalled. The second attempt starts from a weaker position than the first, because now the firm has evidence that these pushes go nowhere.
Why the decision is hard to make from inside
Firms in this position mostly get marketing advice: commit to a niche, publish consistently, market it harder. The advice is premature. You can’t market an offer you haven’t defined, and defining the offer is the step these firms keep circling. The sharper diagnosis in that same survey came from an agency leader, not the report: the billable hour is the problem, and margin expansion on productized outcomes is the way out. Deciding what those outcomes are, what they cost, and who sells them first isn’t a marketing task.
Getting there in another internal meeting is hard, because everyone in it is paid by the math that stalled the question last quarter. It helps to have someone outside the partnership run that meeting: a fixed agenda, the senior people who’ll live with the outcome, and a deliverable defined before anyone sits down. The deliverable is a plan the firm signs off on: the two or three offerings worth building, in order, with a pilot scoped for the first.
Nobody inside the partnership owns the agenda, so no one has to spend political capital keeping it on track. The deliverable is fixed before the first day starts, which removes the memo as an exit.
The senior people who’d sink a mandate handed down from management do most of the talking, shaping offerings they’ll have to sell rather than reviewing a document about their own future. They know the clients and the work.
The person running the session supplies the commercial framework that turns that knowledge into ranked offerings, and holds the group to a decision when the conversation drifts back toward consensus. Everyone in the firm, the managing partner included, has a stake in the compensation math, and that stake is what the job asks them to set aside. When the group splits, the managing partner makes the call. That’s the design behind the two-day plan we run for a whole firm.
Once they decide on an offering, they still have to put it in front of a real client with a price on it. That’s how it stays out of last year’s folder, and the firm learns something it can carry into the next one.
Before AI goes on the next quarterly agenda, write down what the meeting has to produce: one offering, its scope, its price and the client who’ll pilot it. If the meeting ends without all four, it’s worth taking the decision to a session run by someone outside the partnership.