McKinsey & Company, the ubiquitous management consultancy, is making noise about how AI is reshaping its business. The core of the shift? A move towards "performance-based arrangements," where fees are increasingly tied to client outcomes rather than billable hours. About a quarter of their global fees now operate under this model, according to Michael Birshan, a managing partner at McKinsey.
This sounds revolutionary. Consultants, traditionally incentivized to extend projects, now share the risk (and reward) with their clients. Kate Smaje, McKinsey's global leader of tech and AI, frames this as a natural evolution, driven by the demand for "deep implementation expertise" in multi-year transformation projects. She argues that clients undertaking "big career bets" appreciate this shared-risk approach.
But let's dissect the numbers. A quarter of global fees tied to outcomes. Is this truly because of AI, as the narrative suggests, or is AI simply a convenient justification for a trend that was already underway? Smaje admits the shift developed over "the last few years," with the 20% figure "probably more recent." (That's a vague timeframe, isn’t it?) What was the percentage, say, five years ago? That context matters. It's like saying a stock price jump is "because of" a new CEO when the entire market is booming. Correlation isn't causation.
The devil, as always, is in the details of those "outcomes." Success is measured against a "scorecard" that includes investor targets, revenue/profit goals, operational measures, and customer satisfaction. These are all lagging indicators, subject to a multitude of external factors that have nothing to do with McKinsey's specific interventions. Can McKinsey really isolate its impact on these metrics? And how are these targets defined? Are they genuinely ambitious, or are they set low to guarantee a payout? I’ve looked at enough corporate scorecards to know that the bar is often set on the floor.
And here’s a thought leap: How does McKinsey measure the impact of their advice? What's the methodology? The article doesn't say. Is it based on client self-reporting? Internal McKinsey assessments? Independent audits? The lack of transparency here is concerning. Without a rigorous, third-party verification process, these "performance-based arrangements" risk becoming elaborate smoke-and-mirrors exercises. AI is reshaping how McKinsey makes money confirms McKinsey's shift towards performance-based arrangements.
The claim that strategy advice now constitutes less than 20% of McKinsey's work is also interesting. This implies a shift towards implementation. But implementation is notoriously difficult to quantify. It's a messy, iterative process involving countless variables. Tying fees to tangible outcomes in such an environment seems… optimistic, at best.
It’s easy to declare that “We're not a supplier, we're not a vendor, we're a genuine partner," as Smaje does. But partnership implies shared control, and it's unlikely McKinsey is ceding significant decision-making power to its clients. The power dynamic remains firmly skewed in favor of the consultant, who ultimately defines the scope of the project and the metrics for success.
Raj Sharma of EY echoes a similar sentiment, suggesting AI agents may lead to a "service-as-a-software" model where clients pay based on outcome. This is where the AI hype truly kicks in. The idea is that AI can automate tasks previously performed by human consultants, leading to greater efficiency and measurable results.
But let's be realistic. AI, in its current state, is still a tool. It requires human oversight, training, and interpretation. It's not a magic black box that automatically generates value. Claiming that AI will fundamentally transform the consulting model is premature. It's more likely that AI will augment existing consulting services, allowing firms like McKinsey to deliver more sophisticated (and expensive) solutions.
And this is the part of the report that I find genuinely puzzling: If AI is so transformative, why are consulting firms still relying on billable hours at all? Shouldn’t AI be driving down costs and increasing efficiency to the point where hourly billing becomes obsolete? The fact that McKinsey still relies on this model, even for 75% of its revenue, suggests that AI's impact is being overblown.
In the end, McKinsey's "AI pivot" may be less about genuine transformation and more about adapting to changing client expectations. Clients are demanding greater accountability and measurable results. McKinsey is responding by repackaging its services under the guise of "performance-based arrangements" and leveraging the AI narrative to justify higher fees. It's a clever marketing strategy, but it remains to be seen whether it delivers genuine value. The burden of proof, as always, lies in the data.
McKinsey: What They Do and the AI Report
2025-11-21
