There's a version of the future a lot of marketing teams are quietly betting on. AI makes content so cheap that you can just make more of it. Fifty variants of every ad, a post every day on every channel, personalized everything. Volume wins.
I don't think volume wins. I think volume is about to become the floor.
When everyone can make more, more stops mattering
When I started in social, the hard part was production. Getting a shoot approved, getting the edit done, getting it out on time. Brands with bigger budgets simply made more and better stuff.
That advantage is evaporating. Anyone can now produce a decent-looking video, a passable caption and a dozen ad variants in an afternoon. Feeds are already filling up with it, and people can tell. It has a look. Smooth, competent, forgettable.
So what's left to compete on? The idea. The point of view. The thing only your brand could say.
What I learned making a landmark feel new
I spent years leading social for the Empire State Building. It's one of the most photographed buildings on earth, which sounds like an advantage until you try to make anyone care about another photo of it.
Nobody needed more content about the Empire State Building. The internet had plenty. What worked was giving people a reason to see it differently: the people who work inside it, the weird history, the jokes only a hundred-year-old skyscraper could make, the building having opinions about what was happening in the city that day.
None of that came from production. It came from a voice. Once we had the voice, we could make things fast and cheap, and they still felt like us. Without the voice, no amount of polish would have helped. That's what took the accounts to a new audience on TikTok and Instagram.
Where AI genuinely helps creative teams
I use AI on creative work every day. Just not to have the ideas.
- Research and listening. Summarizing what people are saying about a brand or a topic across hundreds of comments, so you start from what the audience actually cares about.
- Getting past the blank page. Asking for thirty bad headlines so you can find the angle in the one that's almost right.
- Versioning. Once the idea and the hero piece exist, adapting them across formats, sizes and markets. This is where the time savings are real.
- Checking the work. Does this caption sound like our last ten? Does this script land the point in the first three seconds?
The pattern is the same as everywhere else. AI is excellent at the parts of the job that come after the decision. It's mediocre at the decision.
A simple test for your next brief
Before anything goes into production, I ask one question: if we took the logo off, would anyone know it was us?
If the answer is no, more variants won't fix it. Go back to the idea. It's the only part of the process that got more valuable this year, and it's the part most teams are trying to automate first.