Design education, and the agency world it feeds, runs on a particular kind of feedback: opinions, delivered in rooms. A crit is opinions from teachers. A pitch is opinions from clients. An award show is opinions from other designers, wearing better clothes. The work is judged by whether people who make work approve of it.
I lived in that world for years, at Memac Ogilvy and then at GranDesign, and I loved it. Then I started buying media, and the feedback changed species.
The verdict machine
Run an ad in the auction and you find out. Not in a quarterly review, not through a strategy director's interpretation. By morning. Real people, spending their own attention and their own money, either responded or did not. The number does not attend the meeting. It does not care about your process, your references, or how long the team worked. It is a verdict, and it arrives in hours.
The first campaigns humbled me. Work I would have defended in any client room, work that was objectively crafted, died quietly at full price. And scrappy variants I was almost embarrassed to run, plain type, blunt claim, no cleverness, pulled numbers the beautiful work never touched. The market was telling me something design school never had: I had been trained to impress the room, and the room was not the customer.
What the auction burns off
Honest feedback at that speed burns things off you.
Decoration goes first. Every element either earns its place or costs money, measurably, and most decoration turns out to be a tax the audience never asked for. Cleverness for its own sake goes next: the joke that needs context, the reference that flatters insiders. The customer is cold, mid-scroll, and owes you nothing. Consensus dies too. A committee can make work everyone tolerates; the auction rewards work somebody feels. And ego takes the longest, but it goes, because after enough mornings of numbers you stop having opinions about your own work and start having hypotheses.
What survives is short. Clarity: say the thing. An offer: give a reason to act, now, to this person. Speed: because the only way to be right often is to be wrong fast and cheap. That is the whole curriculum, and no one can teach it to you in a room, because rooms are the thing it unteaches.
An education in honesty
I think of performance marketing as the most honest education a maker can get, because it removes the gap between what people say and what they do. Rooms run on what people say. The auction only counts what they do. Years of that recalibrates you permanently. You stop asking "is this good" and start asking "what happened when it ran." You hold your taste more lightly and your evidence more tightly. Strangely, the work gets better, because taste trained against reality beats taste trained against applause.
Why this matters more now
I build AI systems now, and the auction's lesson turned out to be the operating manual for them.
A machine can generate endless plausible work, which means "does the room like it" has finally, fully collapsed as a filter. The only filter left standing is the honest one: run it, measure what survives, feed the answer back, go again. My systems are built in exactly that shape. Generation is cheap. Judgment is encoded. The market is the referee, and the loop never stops.
Design school taught me to make things well. The auction taught me to find out whether they work, and to prefer finding out over being flattered. Of the two educations, the second one built everything I run today.