TL;DR
- Marketers keep re-lettering themselves. T-shaped (broad, one deep skill), then V-shaped (deep plus hands-on adjacents), now Pi-shaped (two depths, your craft and AI).
- Strip the letters and every definition describes the same person, someone who went deep enough in something to have real judgment and can carry it into the next problem. That is not a shape, it is experience.
- AI made the joke undeniable. The broad horizontal bar everyone keeps re-naming is now something a model produces on demand, for free.
- The only part that was ever real under the letters is the part AI cannot fake, deep earned expertise and the taste to apply it. Go get that, and stop optimizing your headline.
Every couple of years, marketing reinvents the shape of the ideal marketer. First it was T-shaped. Then V-shaped. This year the smart money is on Pi-shaped. The alphabet is getting a workout.
Let me give the debate its due first, because the shapes do describe real differences.
It's all the same person
I watched a marketer post a heartfelt hiring call this week, hunting specifically for V-shaped people, and quoting a long definition to explain what that meant. The definition, written to describe the one kind of marketer AI supposedly cannot replace, used the phrase “seamless customer journeys.” Which is exactly the sort of line a human never writes and a language model always does. The call for the irreplaceable human was itself, unmistakably, generated.
The top comment said what everyone else was dancing around, and it got more likes than the post.
Marketers keep spending their time coming up with different terms to call themselves instead of getting better at marketing, and that is the reason nobody wants any of them.
The thread spiraled exactly where you would expect. One commenter resolved to become “less O-shaped and more W-shaped,” and Barron answered “More WEEEE, less WOOOO.”
Another reply put it flatter, that the whole framework was mental gymnastics to avoid saying “I want to hire people with more experience.”
They are all right. Strip the letters off any of these definitions and you are describing the same person every time, someone who has gone deep enough in one thing to have real judgment, and can carry it into the next problem.
We keep drawing new shapes for something that already has a name. Experience.
Why it suddenly matters
So why write another post about it. Because AI just put money on the joke.
The broad horizontal bar, the part every letter argues about, is the exact part a model now produces on demand. I open Claude and get a six-email lifecycle flow, a Google Ads account structure, a keyword cluster, and a landing page outline, all at a competent mid-level. The breadth that used to take a decade to build is the cheapest thing in the room. Optimizing for a wider bar in 2026 is optimizing for the part of the job a subscription does.
What the model cannot produce is the other part. The depth that comes from running the thing a hundred times, the taste to tell a good output from the average one it hands you, the judgment to point it. That is the part that was ever real under the letters, and it is the only part still worth paying for.





Going deep in one channel and stuck on the rest?
Book a 1:1 with a marketer who has the depth you are missing, not a new letter. One membership, unlimited calls, every mentor included.
What's left under all the letters
Three things survived the change, and not one of them is a shape.
- Earned depth. The pattern-matching you only get from doing the thing, not reading about it. AI has read everything and has done nothing. It can describe a pricing test, it has never watched one tank revenue for three weeks before the signal cleared.
- Taste. AI gives you the average, confidently. Knowing which of a thousand outputs to ship and which to bin is the whole job now.
- Articulation. The breadth is free, but only if you can command it precisely. Vague in, average out.
So what should you do?
Stop optimizing your LinkedIn headline into a new letter. The shape was never the point, and AI just made that obvious. Go unreasonably deep in one real thing, build the taste to judge the model's output, learn to direct it, and you become the person AI needs instead of the person it replaces. If you are hiring, the same thing flips into a test.
The honest version
I am a non-technical founder shipping production code this year by directing models, so I am not sentimental about human breadth. Mine is gone and I do not miss it.
The marketers worth hiring next year will not be the ones with the cleverest letter on their profile. They will be the ones who got very good at one thing, and can tell when the machine is wrong.
Deep where it counts
Forget the letter. For the stem, these mentors went deep enough in one place to lead it, and they will pressure-test the draft you are about to ship.
Michael Taylor
Growth marketer and prompt engineer, author of Prompt Engineering for Generative AI. The depth-plus-articulation combo this whole post is about.
Will Soprano
Product, SEO and content consultant. Deep technical SEO, the kind AI summarizes but cannot do.
Eden
Conversion copywriter. Messaging and positioning judgment, the taste layer a model does not have.
Frequently asked questions
Marketers who went deep before AI made it cheap
Forget the letter.
Get the depth AI can't fake.
Use the model for the breadth. For the call that decides the quarter, talk to someone whose depth you cannot prompt, and keep them for the next one.
Talk to a mentorKeep reading
More from the GrowthMentor blog
Growth Strategy · Apr 20, 2026
AI does the work now. It still can't tell you which work is worth doing.
Foti Panagiotakopoulos
Growth Strategy · Apr 11, 2026
Forrester says 15% of agency jobs vanish in 2026. Is yours one of them?
Foti Panagiotakopoulos
Growth Strategy · Jul 24, 2022
Everyone Sells “Marketing Advice.” Almost No One Sells Direction.
Foti Panagiotakopoulos





