Almost everything written about AI and agency strategy is written for companies with thousands of employees and an account list that can absorb a bad year. A ten-person studio does not have that cushion. Which means the advice built for WPP or Omnicom does not just fail to apply to a small practice, it points in the wrong direction entirely.
There is a real industry split happening right now, and it is worth naming plainly because it is playing out in public, with numbers attached.
The replacers, in their own numbers
Across the major holding companies, the cost-cutting path has been dramatic. At least 18,000 jobs have been lost across Omnicom, Dentsu, and WPP in eighteen months. WPP alone cut its headcount by roughly 10,600 people between the end of 2024 and mid-2026, with another 1,000 planned. Its own CFO said the reasoning plainly: "most savings will come from people costs" (Joanne Wilson, WPP CFO).
That is the replacer strategy, stated without euphemism. Use AI to justify cutting the people who used to do the work, and reinvest the savings elsewhere.
Why that path doesn't exist for you
A holding company can lose 10,000 people and still have 97,000 left. It can absorb the trust hit, the culture damage, the account attrition, because the entity is large enough to keep functioning through it. A ten-person studio cannot do this math. If you cut the two people who actually hold your client relationships, your taste, your judgment calls, in exchange for a cheaper AI-run process, you have not made your studio more efficient. You have removed the entire reason a client chose you over a template.
At small scale, the maximiser path is also the only one that actually works. The point of AI in a small studio is the same people, freed from the parts of the job that were never the reason anyone hired you, with more room left for the parts that were.
The one honest counter-example worth naming
Not every holding company is cutting. Publicis added close to 6,000 employees through 2025 while its competitors shrank, on the back of real organic revenue growth. To be fair to the data, that growth is credited to business performance and acquisitions, not an explicit AI-maximiser strategy. But it is still the clearest live proof available right now that growing, not shrinking, is a real option in this same industry, under the same AI pressure everyone else is citing as a reason to cut.
This is the same test we apply to every system Shery builds through Aparato: if it makes your practice smaller or more replaceable, we don't build it. See what that looks like in practice →