The potential power of AI for advertisers is real, so this isn’t meant as an indictment of AI. But I have one primary concern about AI when it comes to Meta advertising.
Let me explain…
Make More Bad Decisions
But I fear AI’s main application for advertisers, particularly early, will be to make more bad decisions more quickly. Whether they are creating dashboards, subscribing to random apps, or even leveraging an AI agent or LLM, there is real risk here.
If you’ve obsessed over the wrong metrics before, you’ll have AI to do it for you automatically. But you’ll have that AI create colorful dashboards so that it looks more important.
If you micromanaged ad results based on small sample sizes before, you’ll have AI highlight these differences for you now. You’ll just have access to more of these meaningless results more quickly.
Advertisers will have AI generate countless ad copy and creative variations to “win in the era of Andromeda.” But too many won’t check the quality of those assets.
When AI Will Be Useful
AI is only useful when it’s based on best practices and good advice. Otherwise, it can do a really good job of creating convincing output that won’t actually help you.
Automating bad strategies will only help you lose money faster. AI will not fill your knowledge gaps if it doesn’t first have a foundation in best practices.
There will be excellent applications of AI for advertisers, but we’ll need to sort through a whole lot of bad ones first. Everywhere you look, someone will have an “Amazing AI App” that looks impressive at first glance. The proof will be in the long-term, sustainable results.
AI can be useful when in the right hands. But it’s otherwise going to be a cool shiny object that results in more frustration. It will be our job to discern value from pretenders.
We’ve been talking a lot about AI in our weekly strategy sessions. Join us at www.jonloomer.com/elite.
