Business
AI is breaking the billable hour, and outcome-based pricing is moving in
As AI compresses the hours behind knowledge work, consulting and professional-services firms are moving away from time-based billing, the Wall Street Journal and Business Insider report. McKinsey now says more than 30% of its global fees are tied directly to client outcomes; Deloitte has told its own consultants that the traditional labor-based model could shrink sharply as a share of the market by 2035. The emerging playbook pairs a stable base fee with a clear target, quality guardrails, and shared upside rather than a timesheet.
So what
If your business sells expertise by the hour, whether services, agencies, or advisory, AI is eroding the very unit you bill and clients will increasingly expect to pay for results. The shift carries a trap: single-metric outcome deals can quietly reward corner-cutting and headcount removal. The action: model at least one engagement on outcomes tied to a bundle of metrics, such as cost plus quality or speed plus accuracy, and start repricing before clients force the conversation.
⚠ Hype check: "The death of the billable hour" overstates it. The majority of firm revenue is still time-based, and reported outcome-tied shares cluster around a quarter to a third, not a majority.
Confidence: Informed Perspective — the trend and McKinsey's figure are reported, but the pace and end-state are contested.
Go deeper TheStreet · Curated: The Neuron
AI Strategy & Transformation
Ethan Mollick: the chatbot era is giving way to agents, across every function
In "The twilight of the chatbots," Wharton's Ethan Mollick argues the dominant mode of using AI is shifting from prompting a chatbot to assigning work to long-running, self-correcting agents. He cites an OpenAI internal study showing agent adoption spreading well beyond engineers, with Legal, HR, and other functions adopting at nearly the coder rate and a quarter of OpenAI staff running four or more agents at once each week. A separate study of Claude Code users found success depended less on profession than on domain expertise: experts extract more useful output per prompt and operate agents like a manager runs a team.
So what
The unit of AI value is moving from the assisted individual to the managed agent, and the people who win are domain experts who can direct and verify fleets of agents. Treating AI as a better chatbot understates where the leverage is heading. The action: find the workflows where you can hand an expert a set of agents rather than a faster autocomplete, and invest in the verification and oversight skills that turn agent output into work you can trust.
Confidence: Informed Perspective — Mollick synthesizes early studies and his own experiments; the trajectory is argued, not settled.
Go deeper One Useful Thing · Curated: Ethan Mollick
AI Strategy & Transformation
Every: name the four bets your AI strategy is quietly making
Writing in Every, Dan Pupius, CTO of The General Partnership, offers a four-axis framework to replace the lazy "AI wrapper" label with explicit bets. The axes: token economics (do costs rise, or fall toward zero), model self-sufficiency (does a more capable model absorb your product, or does it need your integration), platform structure (are you locked to one provider, or swappable), and trust and governance (do compliance rules stay light, or tighten). Each axis comes with practice questions: what is our bet, what must be true for it to pay off, what signal would tell us we are wrong, and how fast could we adapt.
So what
Most AI strategies rest on unexamined assumptions about cost, capability, lock-in, and regulation, and an overnight event like a government access ban can invalidate them. Making the bets explicit is how you avoid building durable features on temporary conditions. The action: run your top AI initiative through the four axes, write down the bet each implies, and define the leading indicator that would tell you the bet has gone wrong.
Confidence: Informed Perspective — a useful framework rather than empirical proof; its value is in the discipline it imposes.
Go deeper Every · Curated: Every (Thesis)
