The deck and the drawer
Strategy used to be slow, heavy work. Half the time went on making the deck look the part. The other half went on making sure the thinking held up. In a busy agency with three clients on the go, the second half always lost, and you walked into the room hoping nobody asked the one question you hadn’t had time to answer.
The work I loved most was consultancy strategy. You sat beside the CMO and solved the problems they knew they had but never had the time to reach. No creative fluff. Just research, logic and an answer you could defend to people in the business who’d never sat in a marketing meeting.
Then it stopped. You handed over 150 pages and walked away. Whether the client had the time, the people or the skills to make it real was someone else’s problem. Ask any design agency that has inherited one of those decks.
Plenty of thinking. No creative manual.
One night, one website
I was early to AI. ChatGPT became my sidekick within weeks, a sounding board in a city where strategy teams are small and always busy. But the real shift came on a run, listening to a podcast about vibe coding. A man had built a website over a weekend. It sounded a world away from anything I did.
Then a young artist I know made a brilliant piece about making vaping as socially unacceptable as smoking. I wanted to give it a home online. So I opened Claude Code in the terminal, understood almost nothing, and screenshotted every error into Claude chat with one question: what does this mean, in plain English? By the end of the night, the site worked.
From scamps to a working product
Next came Elora, a health startup for women living with lipoedema, a condition medicine largely ignored for decades. I did the research, spoke to women who live with it, and found the insight: the millions of women medicine forgot. I showed the client early scamps of the site. They told me to keep going. So I built it. The website, the newsletter and the database behind it.
That was the first time I saw the job the way a CMO sees it. A good idea isn’t enough. Someone has to make it real, and make it something people actually want to sign up for.
What built strategy means
The strategist doesn't stop at the plan anymore. You do the thinking, then you build enough of it that people can see it working.
As in: the insight, the landing page and the newsletter, in the same week.These days I don’t work with one sidekick. I work with a team of agents for research, copy and design, set up like another agency at my back, if I need them. I set the direction and check every stage. Nothing leaves without my sign-off.
What hasn’t changed
The core of the job is exactly the same. Understand people. Find the insight. The best ones still come from sitting across a table from someone, and AI can’t do that. It only pulls from what already exists, so on its own it can’t find the new thing.
That’s why we’re seeing strategy turned around in two days. Fast, cheap and hollow, with a bit of human top spin on a spongy AI foundation. That isn’t the same as work made together. Clients pay for it because any strategy feels better than none. It’s our job to show them the difference.
So the strategist matters more now, not less.
Design, product, content and social all turn on one question: does this brand make people’s lives better? Nils Leonard at Uncommon talks about building brands people want to exist in the world. Someone has to know what that brand is. That’s still us.
Monday morning
I learned this trade in London, from talented people and good books. Underneath all of it was one lesson: stay curious. That hasn’t changed. It just points at new things now.
So on Monday, build something you don’t know how to build yet. Get stuck. Ask stupid questions. Screenshot the error.