Download Intralogistics White Paper
(800) 959-8951

What McKinsey Just Admitted About Its Own Model

You don’t need to look far for consulting alternatives these days. McKinsey & Company just handed you one, although the firm probably doesn’t want you to connect the dots and understand this paradigm shift.

McKinsey now runs 20,000 AI agents alongside its 40,000 human employees. Global managing partner Bob Sternfels told Harvard Business Review he expects that number to roughly double within a year and a half, enough for every human employee to work alongside at least one agent.

Think about that. The world’s most famous consulting firm just admitted its billable-hour model can’t survive without a small army of software doing work junior consultants used to do.

McKinsey doesn’t want you to ask the next question: If AI agents can do much of the job, why pay McKinsey’s markup to access them? I’m getting some version of this question from more CEOs these days. I’d love to know what you’re thinking, because as long as you understand the Rentability Threshold, you don’t need big consulting.

The Rentability Threshold Forces Consulting Beyond the ‘Buy-It’ Stage

The Rentability Threshold is the point where a capability becomes standardized, digitized or AI-orchestrated enough that renting outcomes beats owning the staff who used to produce them.

I detailed the Rentability Threshold and its history in my latest analysis, The AI-Enabled Enterprise. Software crossed the Rentability Threshold in 1999, when Salesforce turned licensed CRM software into a subscription. Servers crossed in 2006, when AWS replaced owning data centers with renting cloud computing. Logistics crossed it back in the 1980s, when deregulation helped create the modern 3PL industry.

Every capability moves through the same four stages on its way across the Rentability Threshold.

You start by owning it. That means, for IT, hiring programmers, buying servers and building the internal team. This full-time headcount comes with a high fixed cost and is slow to scale.

Next, you buy licensed software and run it on-premise, which means you still need an IT department. That phase is still staff heavy and slow.

Then you rent that capability. Think SaaS, cloud computing and 3PLs. This paradigm allows vendors to spread the cost across many customers, and the capability becomes elastic instead of fixed: you pay for what you use this month, not what you might need someday.

Now, advancements in AI let you orchestrate the work. In The AI-Enabled Enterprise, a small core sets priorities, and AI scopes and routes the tasks. Leadership calls in fractional executives only where you need judgment.

Consulting has been stuck in the “buy it” stage for a century. You pay a fixed, enormous fee for a fixed engagement, and you own none of the capability once the consultants pack up their laptops.

And, as we at Tompkins Ventures have mentioned many times before, many consultants throw a book-sized report over the wall and leave. They don’t execute. And sometimes, they don’t even leave you with good details on how to execute, because those junior consultants haven’t spent years walking warehouses, ports, factories and more.

McKinsey’s 20,000 agents prove the firm knows this paradigm shift is coming. Their pitch is “hire us, and you get our AI.” But why rent McKinsey’s agents at McKinsey’s markup when you could own the AI orchestration and rent fractional human judgment as needed? 

Clearly, the big consulting companies want to keep you in the “buy it” phase forever.

Consulting’s AI Doesn’t Cut Costs for Clients

The AI-Enabled Enterprise detailed a $150 million consumer goods company that needed a vice president of supply chain. But the numbers couldn’t justify a $350,000 full-time salary.

A Tompkins Ventures partner placed a fractional executive instead. Supported by AI-enabled analysis, that executive got the job done with 200 hours of elite leadership. The company saved $280,000 a year and delivered results much faster than a new hire ever would have.

For another enterprise, a fractional executive delivered $15 million in cost reduction by stabilizing and optimizing deployment operations.

In another case, a national distributor wanted a redesigned transportation network, a project that normally takes six months. AI agents and fractional supply chain experts did it in six weeks. Transportation costs fell 14%. Service levels improved 9%.

And nobody paid a full consulting team six or seven figures for those results.

Run the same math for all three scenarios against a traditional consulting engagement, and fractional leadership paired with AI wins every time.

Big consulting bills you for a complete team. You pay for partners, engagement managers, associates and analysts. Now, with artificial intelligence behind them, they tackle the research and synthesis, pairing AI with human intelligence.

McKinsey’s numbers confirm it. Sternfels called the firm’s new approach “25 squared”: McKinsey is increasing client-facing roles by 25% while reducing nonclient roles about the same.

And you pay the bill, including all that overhead for a pyramid of labor the consulting firms are automating. Your fees don’t drop, but big consulting’s margins increase.

A fractional executive paired with your own AI orchestration flips that math. You pay for the judgment, not the pyramid underneath it. You own the AI doing the research and synthesis instead of renting access to someone else’s. And you keep the difference McKinsey used to pocket.

That’s a genuine consulting alternative.

Own the Orchestration. Rent the Judgment

Given that knowledge, why in the world would you pay a consulting firm to stand between you and their AI? Build or license your own AI layer to scope problems, gather data and draft the first pass of analysis. Then hire fractional executives, specialists who can review that work and make the calls only a human should.

I wrote about this exact question 20 years before AI had anything to do with it. In 2005, in Logistics and Manufacturing Outsourcing: Harness Your Core Competencies, my co-authors and I argued that leaders should sort every function into one of four buckets: primary core, secondary core, primary non-core or secondary non-core. Primary core is what customers know you for, what creates the most value, what competitors can’t easily copy. Secondary core still creates value even when customers never see it. Everything else, primary and secondary non-core, raises the same question every time: should you really own this, or should someone else provide it better?

Most of what you send out for a consulting engagement falls into that last bucket.

So the next time you’re considering big consulting, ask yourself three questions: Has the analysis you’re paying for become a commodity any competent AI system can produce? Does a fractional expert’s judgment actually move faster than a consulting team’s timeline, instead of just costing less? Does keeping a full engagement on retainer create drag, fixed fees, slow scoping, risk you don’t need to carry?

If you answered yes three times, your former consulting need has crossed the Rentability Threshold. Your leadership team really should look for consulting alternatives. Your company will get better outcomes without a bigger payroll or a huge retainer.

And, more importantly, what you build, you actually keep.

The Matchmaking Era Offers Consulting Alternatives, Not Heavy Markups

Tompkins Ventures began building the perfect ecosystem for this paradigm shift more than six years ago.

Back then, I was arguing that consulting firms were trying to be too broad and too deep. That combination only works with enormous overhead – think international offices laden with staff. Meanwhile, junior staff fresh out of Ivy League schools visited client sites to tackle the problems.

Clients ended up paying for an expensive recipe that wasn’t yielding results. Consulting firms had already started to shed staff as their own clients saw diminishing returns. And that was well before ChatGPT kicked off the artificial intelligence explosion.

Our matchmaking ecosystem offers you consulting alternatives by tackling the problem from the other direction. Instead of one firm trying to be broad and deep and profitable all at once, Tompkins Ventures connects you only with what you need. That could be a specific partner for a project, a fractional executive backed by AI-enabled capability, whatever fits your situation.

No pyramid. No markup for capacity you didn’t ask for.

McKinsey just proved the pyramid was never sustainable. I would rather help you skip it.

A Consulting Alternative Without the MBA Toll

Next time you have big issues, you have consulting alternatives. McKinsey, Deloitte, PwC, EY and KPMG are facing stiff competition.

So pick one project sitting on your desk right now, the kind you’d normally send out for a scoped engagement. Ask what AI orchestration and fractional leaders could do with it instead. Picture what would happen if you matched with the right partner instead of an army of consultants with MBAs standing between you and artificial intelligence.

I’d love to hear what you find. Drop me a line.