Large agencies do not produce mediocre work because they are populated by mediocre people. They produce mediocre work because their economics require them to.
Once an agency crosses about forty heads, four things happen in roughly the same year. The senior people start spending more time selling and less time making. The junior people start producing more, supervised at a longer distance. The work begins to homogenise around whatever lowered the risk in the last client review. And the rate card stops reflecting the value of judgment and starts reflecting the cost of payroll.
None of this is sinister. It is simply what scale does to a service business whose product is taste.
The pyramid
An agency at scale is, financially, a triangle. A narrow band of partners at the top who win the work. A wider band of directors in the middle who run the work. A very wide band of associates at the bottom who do the work.
The maths only function if the partners spend most of their time on business development, the directors spend most of their time managing, and the associates spend most of their time producing. That distribution of attention is the model. It is not flexible. The minute a partner spends a week on craft, the firm's gross margin moves the wrong way.
What clients believe they are buying is partner-level judgment applied to their problem. What they are actually buying, on most days, is associate-level execution against a brief that a partner approved in a fifteen-minute review on Friday.
This isn't dishonest. It is the only way the business can clear its overhead. But it has consequences.
What happens to judgment in a pyramid
Judgment is the rarest input in a creative business. It is what separates a good idea from a finished one. It is also expensive, which is why the system above pushes it into a minority of hours.
When judgment is rare in an engagement, work tends to get either over- or under-cooked. Over-cooked when nobody senior dares cut the third deck. Under-cooked when nobody senior has seen the campaign before it shipped. Either way, the client buys hours that do not contain the thing they were paying for.
A studio of eight people with twenty years each does not have a pyramid. Every hour is a partner hour. The rate card is higher per head, lower per outcome. It also means the work cannot hide behind an org chart. If it is mediocre, there is nobody to blame, which is a productive form of pressure.
The Pareto problem at scale: 80 percent of the value lives in 20 percent of the hours. Big agencies bill all 100. Small studios sell only the 20.
The Pareto problem
Take any agency engagement and you can find, after the fact, the five or six hours that produced the actual value. The strategy session where the question got reframed. The thirty minutes where the headline arrived. The afternoon where the homepage hierarchy unlocked. The walk where someone realised the audience was wrong.
The rest is craft and coordination. Both necessary. Neither rare.
At scale, the agency bills for the craft and the coordination because that is what is countable. The five hours that produced the value get folded into the strategy line, which is the smallest line on the SoW, because strategy doesn't ship a file. At a smaller studio with senior people end-to-end, those five hours run through the same brain that drew the wireframe and wrote the campaign deck. You pay more per hour. You buy fewer hours. The hours are denser.
What clients should actually be buying
An afternoon with the right person beats a quarter with the wrong team. This sounds glib until you compare two engagements at the same dollar value. Do it on three dimensions.
- Cycles of judgment per dollar. How many times does a senior brain look at the work before it ships? In big shops, this is often two: kickoff and final approval. In small studios, it is closer to ten.
- Distance from decision to execution. The longer the chain between the person who has the insight and the person who executes against it, the more the insight gets diluted. Small studios have a chain of one or two. Network agencies often have chains of six.
- Replayability of the insight. Can you, as a client, repeat the insight at the board meeting without your account director in the room? If yes, the engagement was useful. If no, you bought a deck.
Where bigger is actually better
This is not a thesis that small wins every time. There are categories where scale is genuinely the right answer.
Multi-market production at speed. Global media buying. Pitch theatre for procurement-led RFPs. Compliance-heavy regulated work where the price of an error is enormous and you want fifteen people on it. Any time the project has more linearity than ambiguity, the pyramid is efficient.
Most strategic and creative work is the other shape. More ambiguity than linearity. The right team for it is small, senior, and short-chained.
How to spot the difference in a pitch
The cheap tell is who is in the room. If five people pitch and you never meet three of them again, you are buying a pyramid. If two people pitch and they will still be on the call in month five, you are buying judgment.
The slightly more honest tell is how the agency talks about their process. Pyramids talk about workstreams, deliverables, and gates. Studios talk about the question they want to answer, the people they need to talk to, and what they expect to be uncomfortable about. Neither vocabulary is wrong. They serve different products.
The studio bet
The bet of a small studio is that twenty years of taste compound, and that compounding cannot be replicated by giving the same problem to four people with five years each. The market mostly disagrees, which is why the small studio model stays small. Some of the best work in any given year comes from twelve-person rooms anyway.
Hire the room you need for the problem you have. Sometimes that room is forty people. Often it is six. The mistake is assuming the answer is always the bigger one.