How to deploy AI agents in your consulting firm (without making the GM mistake)

There are two ways to deploy AI agents in a consulting firm. The GM approach replaces consultants with agents — same workflow, same tasks, fewer people. The Toyota approach uses agents to unlock capabilities the firm couldn't have before — more clients with the same team, faster onboarding, proposals built from complete context. Same technology. Entirely different outcomes. Most firms are making the GM mistake.

Key takeaways:

• GM spent $40–45B on factory robots and became Detroit's highest-cost producer. Toyota used identical technology, redesigned the system around it, and surpassed GM in profit per vehicle by 3x within a decade.

• The biggest AI agent opportunity in consulting isn't cost reduction — it's unlocking work that never got done: bids not submitted, clients not followed up, proposals not sent because there wasn't enough bandwidth.

• The question that separates a good AI deployment from a bad one: "What can my firm do now that it literally couldn't do before?"

In the 1980s, GM's CEO Roger Smith made a $40–45B bet on factory robots.

He placed them exactly where human workers had been. Same workflow. Same assembly line. Same logic. The robots sometimes painted each other instead of cars. Manufacturing costs increased. By the time Smith retired in 1990, GM had gone from Detroit's lowest-cost producer to its highest-cost producer.

Toyota, working with identical technology, asked a different question: What becomes possible when this new capability enters the system?

They redesigned the factory around the robots. Layouts changed. Quality feedback tightened. Human workers shifted from task execution to system oversight. Toyota surpassed GM in profit per vehicle by 3x within a decade, and by 2008 became the world's largest automaker — a position GM had held for 77 consecutive years.

Same tools. Entirely different outcome.

This is exactly what's happening with AI agents right now.

The GM approach: agents as substitutes

Most firms deploying agents are doing it the GM way.

They're dropping agents into existing workflows — one agent per role, one task at a time. Summarise this. Draft that. Replace this function. The pitch is cost reduction: fewer hours, lower headcount, tighter margins.

It's a compelling case on a spreadsheet. It's also leaving the biggest opportunity completely untouched.

The GM approach is seductive because it promises headcount savings. But it ignores the real problem: most firms are already scaling labor instead of profit, and replacing consultants with AI doesn't fix the structural inefficiency — it just moves it.

When you deploy AI as a human substitute, you compete on price. The Toyota question is different: what can your firm do now that it literally couldn't do before?

That's where the real value lives.

The Toyota approach: agents as capability unlocks

I'm building Aether, a client context platform for consulting firms. The original pitch leaned on agent-as-substitute framing: Recall Agent, Debrief Agent, Status Agent, Onboarding Agent — each one replacing a task someone used to do manually.

It was clean. Tidy. And it was the GM mistake.

Because the actual value isn't in what those agents replace. It's in what they unlock.

Right now, the principals at most consulting firms are carrying context in their heads. Every client relationship lives in a different Slack thread, email chain, or notebook. When a new team member joins, weeks of context get lost. When a client asks "where are we on X," someone has to dig.

The Toyota approach turns AI into a revenue multiplier because it eliminates the Recall Tax — the hidden cost consulting firms pay when client knowledge lives in people's heads instead of systems.

What becomes possible when context is never lost?

More clients taken on with the same team size. Faster onboarding that doesn't depend on the founding partner. Proposals written from a complete picture rather than a fragmented one. Follow-through that doesn't slip.
That's not a cost reduction story. That's a revenue growth story.

The "work not done" is the bigger market

The most compelling AI agent opportunity isn't replacing labor. It's unlocking the work that never got done — bids not submitted, clients not followed up, proposals not sent because there wasn't enough bandwidth.

This doesn't show up in a cost-reduction model. It shows up in revenue growth.

Toyota never positioned their automation as "robots replacing workers." They shipped better cars.

What this means if you're evaluating AI agents for your firm

Stop asking: which role does this replace?
Start asking: what can my firm do now that it literally couldn't before?

If the vendor can't answer that question clearly, their framing is wrong — and the implementation probably will be too. The agents that win in professional services won't be the ones that cut headcount. They'll be the ones that let a 10-person firm take on 15-person work.

The Toyota approach requires something most firms don't have yet: a context layer — a persistent knowledge system that captures everything a firm knows about its clients. Without it, even well-designed agents operate on incomplete information.

That's the Toyota deployment. That's the one worth building toward.
Want to see which approach your firm is set up for? The Context Leak Scanner reveals where your firm's context gaps are — the same gaps that determine whether AI agents amplify your team or just automate your inefficiencies.

Run the free Context Leak Scanner →

Frequently asked questions

What is the difference between the GM and Toyota approach to AI agents?

The GM approach drops AI agents into existing workflows as human substitutes — same tasks, fewer people, cost reduction focus. The Toyota approach redesigns workflows around AI capabilities, using agents to unlock work the firm couldn't do before. GM's $40–45B robot investment made them Detroit's highest-cost producer; Toyota used identical technology to surpass GM in profit per vehicle by 3x.

Should consulting firms use AI agents to cut costs or grow revenue?

The biggest opportunity is revenue growth, not cost reduction. The most valuable AI agent use cases in consulting unlock work that never got done — bids not submitted, clients not followed up, proposals not sent due to bandwidth constraints.

How do I evaluate AI agents for my consulting firm?

Ask one question: "What can my firm do now that it literally couldn't do before?" If the answer is only about replacing tasks or cutting headcount, the deployment is following the GM pattern. The agents that create the most value let a 10-person firm take on 15-person work — capability amplification, not human substitution.