Big Consulting Just Discovered How Agencies Work

Accenture has spent the last few years adding tens of thousands of AI and data professionals. EY put more than a billion dollars behind its AI platform. McKinsey keeps building out QuantumBlack and now puts AI at the center of how it delivers work.

The agency world looked up from its dashboards and said, "Welcome."

The big firms will do fine. They will win enormous contracts. They will write white papers that become industry standards and host events where CMOs nod along. That is not the point.

The point is to call the rebranding what it is.

Take away the language, and the big firms are describing a delivery model built on three things: rapid prototyping, practitioners embedded inside client teams, and working toward measurable outcomes instead of handing over a document and leaving.

This is called working at an agency.

I ran one for 13 years. We called it embedded. We called it iterative delivery. We called it showing up on Monday with a working version instead of a slide deck.

We did not need a billion-dollar announcement to do it.

Small agencies didn't build this model because they were ahead of a trend. They built it because they had no choice. When you are a 10-person shop competing for the same work as a 10,000-person firm, you win on specificity and speed. You embed. You iterate. You ship. You do not have the budget for theater.

AI didn't create that model. It plugged into it. Agencies that already worked this way were among the first to use AI on real client problems, because they were already in the room when the problems arose.

Tens of thousands of new AI hires sound like capability. In practice, it is an onboarding problem.

Many of those people come from adjacent roles inside large organizations. Becoming deployable on real client work, at the level the market expects, takes time.

Meanwhile, the agency that has spent the last few years using AI on live client work has a learning curve it can't hire away.

What clients are short on is not AI tools. It is judgment. They need someone who can walk into the room and say, "This use case, in your organization, with your data, here is what works, and here is what will fail." And here is why the thing the vendor is proposing will look great in a demo and break three months after go-live.

That kind of judgment does not come from a hiring announcement. It comes from shipping the same category of thing 40 times and learning what breaks.

The big firms cannot close this judgment gap with headcount.

Close contact with real clients. Navigating real organizational politics. Knowing that the CEO's stated priority and the actual priority are different things. These are judgment capabilities, and they compound through relationships.

Agencies with durable AI practices aren't the ones with the best capability decks. They are the ones where the founder picked up the phone when something broke. They knew the client's internal dynamics before walking into the meeting. Trust built up over months of showing up and delivering.

As the cost of intelligence keeps falling, this is what gains value. The trust between a practitioner and a client does not transfer to a new account manager. The big firms cannot manufacture it at scale.

The consulting land grab is not a threat. It is a clarification.

You cannot outstaff Accenture. You will not out-resource them. You will not write a better capability presentation than BCG X. So don't compete on their terms.

Build a position so specific (this client type, this problem, this method, this outcome) that you are not competing with them at all.

The big firms are going upmarket. Their model is built for enterprise budgets and long timelines. That leaves the midmarket wide open: companies that cannot absorb consulting overhead, cannot wait a year for a capability to come online, and need someone who shows up with real answers.

That is your ICP.

Here is what I think happens next. The big firms start buying boutique AI practices. They can build the capability themselves, but at their scale, the math on how long it takes to earn credibility just does not work.

The firms that get bought will not be the ones with the best marketing. They will be the ones with specific expertise, real client relationships, and a track record measured in production deployments, not pilots.

We have seen this pattern before with digital transformation, agile delivery, and UX as a practice. The big firms announced their plans, hired talent, and then acquired the shops that had built the real expertise.

If you are building something specific and making it work for real clients, you are building something someone will want to own.

The right response to all of this is not panic. It is not putting "AI-native" above the fold on your pitch deck. It is not about trying to get on a list next to Accenture.

The right response is to get every engagement to a working production deployment.

Most organizations have AI pilots. Far fewer have anything running in production. Recent surveys on AI agents keep finding the same thing: lots of experiments and only a small share in real use.

That gap is where the work is. It is where agencies that know what they are doing get called back and where the relationship shifts from vendor to advisor.

Small agencies that have shipped real work for real clients aren't behind the big consulting wave.

They were doing this work before the press release.