McKinsey’s AI Problem Isn’t Technology, It’s Pricing

The consulting world used to run on one simple equation: throw a room full of smart twenty-somethings at a client problem, bill by the hour, then arrive three weeks later with a 140-slide PowerPoint deck and a confident smile. That formula built McKinsey, Bain, BCG, and an industry worth roughly $400 billion globally. Now AI just walked into the conference room, grabbed the junior analyst’s keyboard, and asked why the client is still paying for all those hours. That question is starting to hit consulting firms where it hurts most, pricing, staffing, and partner compensation. McKinsey’s recent internal pay overhaul, dubbed “Project Acorn,” may sound like something from a hedge fund retreat in Aspen, though it reveals something far more serious: the economics of consulting are changing fast, and even the industry’s most elite firms know the old billing model looks shaky. For accounting firms, advisory practices, and finance leaders watching from the sidelines, this is not just “consulting industry drama.” It is a preview of what happens when AI starts compressing billable time across all professional services.

Why bill 200 hours when AI does it before lunch?

For decades, consulting firms relied on the pyramid model. A small group of highly paid partners sat at the top, while armies of junior consultants handled research, benchmarking, Excel modeling, and presentation prep underneath them. Clients paid for the labor stack. AI just started flattening the pyramid. Tasks that once required weeks of analyst work now take minutes using generative AI tools. Data synthesis, competitor analysis, scenario modeling, and draft recommendations increasingly come from AI-assisted workflows. Clients know this. That is the problem. McKinsey clients are now pushing harder for outcome-based pricing instead of traditional hourly billing. They want consulting firms paid based on cost savings achieved, margins improved, operational efficiencies delivered, or market share gained.

Are consulting firms becoming the Uber drivers of white-collar work?

AI-native companies increasingly price services based on tasks completed. Salesforce charges per customer-record update. AI support platforms charge per case resolved. Identity verification firms charge per verification processed. Consulting clients are starting to ask the same uncomfortable question: if AI helps consultants finish work faster, why should clients still pay as though nothing changed? The pressure is not theoretical anymore. Microsoft reportedly rolled back broad usage of Anthropic’s Claude Code internally after adoption costs climbed rapidly. Uber’s CTO disclosed that the company burned through its entire 2026 AI coding tools budget within just four months.

Source: Financial Times

Goldman Sachs estimates that agentic AI systems could drive a 24-fold increase in token consumption by 2030. Gartner expects AI inference costs to fall sharply, though enterprise spending may still rise because usage grows even faster. Consulting firms now face a weird accounting problem. AI lowers the cost per task, but dramatically increases the volume of tasks clients expect to be completed. Anybody who has worked through a tax-season software rollout knows exactly how this movie goes. The minute processes become faster, leadership starts asking why ten more things cannot also get done by Friday.

Is the “slide deck economy” finally cracking?

Clients are getting more skeptical about the traditional consulting model. Many finance and strategy teams now use internal AI tools for research and early-stage analysis that once went straight to outside consultants. They still want expertise and execution help, just not massive invoices for information gathering. That shift is pushing consulting toward leaner models. Boutique firms with small teams of senior specialists are gaining traction, often using AI heavily instead of relying on large junior teams. The old leverage model starts looking shaky when AI handles much of the entry-level work first.

CPA firms should watch this closely. Tax research, financial analysis, audit documentation, and compliance testing increasingly involve AI-assisted workflows. Clients eventually notice when turnaround times shrink but invoices stay the same. This does not mean firms should slash fees overnight. It means they need stronger explanations of where human judgment, experience, and execution still create value. Because clients are no longer paying premium advisory rates for work AI can already handle in a few clicks.

So, what are clients actually willing to pay for now?

The answer appears to be implementation, accountability, and judgment. Clients still struggle with execution. AI can generate recommendations. It cannot yet walk into a tense boardroom, calm down a nervous audit committee, align six department heads, and manage political resistance inside an organization. At least not until ChatGPT starts asking for Marriott Bonvoy points and expense reimbursements. That is why outcome-based pricing may expand unevenly. Some consulting engagements naturally lend themselves to measurable targets. Procurement savings, operational efficiency gains, or working capital improvements are easier to quantify.

Others are murkier. A recession, tariffs, interest-rate shifts, or management turnover can derail even strong strategic recommendations. Consulting firms do not control all the variables affecting business performance. Still, the broader direction feels clear. Professional services firms increasingly need to prove value through results instead of effort. AI is accelerating that pressure because clients now see how much work can be automated.

The Final Call

The bigger takeaway for accounting and finance leaders is not whether McKinsey tweaks partner compensation by 3% or 5%. It is that the entire economics of expertise-based businesses are being renegotiated in real time. The firms that survive this shift probably will not look like the old consulting empires. They may look leaner, more specialized, more technology-heavy, and more directly tied to measurable business outcomes. In other words, the age of “trust us, we’re smart” billing may finally be running out of road.