Steven McConnell.
A few years ago, I stayed at a hotel near Soho in New York. What I remember most is not the room, the bar, or the room service. It was the doorman.
He wasn’t just opening doors or handling bags. His value was harder to pin down than that.
Impeccably dressed, articulate, with a British accent that somehow felt completely at home in Manhattan, he had a warm presence of a yoga teacher - yet he moved with the sharp precision of a Swiss railway worker.
He wasn't just opening doors and handling bags - he was setting the entire context for the experience.
And he made the hotel feel better than it probably was on paper.
The feeling of being genuinely welcomed - taken care of without a hint of pompousness - stayed with me long after I checked out.
A few years later I went back. The doorman was gone. No replacement. I just walked in. Nothing was wrong with the rest of the stay. But everything felt flatter.
I didn’t know it at the time, but I experienced what Rory Sutherland calls “the doorman fallacy” first-hand.
That is the problem many businesses are about to create with AI.
The first thing most businesses will do with AI is the dumbest thing: use it to justify headcount cuts.
Not because that's the smartest application of the technology, but because of the pressure to look good in front of boards and investors.
Marketing is unusually exposed in this environment.
Why? Because a lot of marketing value looks messy. It does not always show up neatly in a dashboard. It is often delayed, indirect, hard to isolate and easy to misread.
The Soho doorman was almost certainly not hired by the marketing department. But he might as well have been.
The was adding mountains of value to the things marketing gets measured on - return visits, satisfaction scores, word of mouth, lower CPAs.
And almost none of it would have shown up on a spreadsheet.
This is where the danger lies. In the first wave of AI-driven restructures, many businesses will confuse what is hard to measure with what is safe to remove.
That will expose two different types of marketers.
The first are marketers who create real but hard-to-measure value. I once hired a copywriter who charged $10K and took two weeks to write a single landing page. Preposterous, right?
That page has since converted more than $10M of product at an astonishing rate.
Marketers like that don’t fit neatly into boxes. They’re messy, creative, and often wrong. But when they’re right, they’re really really right. And their value shows up in unexpected ways.
Some of these people will be cut unfairly - because their contribution don’t fit the language a restructuring committee prefers.
The second group has been hiding behind complexity, process, and the appearance of busyness.
Always in meetings, always “warming client relationships”, always producing decks about decks - but when you strip the activity away, the commercial contribution is thin. AI will expose this group too. And it should.
If you're in the first camp, listen up. You need to learn to sell your value to key stakeholders. And the great irony is that you’ve excelled at marketing brands for decades - but you aren’t great at marketing your own brand.
It’s not a character flaw - it’s normal.
Most marketers are too close to their own career to see it clearly. They describe responsibilities instead of outcomes. They bury impact inside lists of activity. They assume someone else will connect the dots.
In a forgiving market, that's something you can get away with. In an AI-era restructure, it's a liability.
Spend this weekend putting yourself in the shoes of your CEO. Practice explaining your impact in commercial terms that will resonate with them.
It’s one of the best forms of career insurance you can get today.
Steven McConnell is the Chief Customer Officer and Co-Founder of Arielle Executive, a specialist advisory firm.