AI Will Amplify Whatever You Already Are
The other day, Grant was explaining how we run a marketing diagnostic. Nothing flashy, just walking through how we look at a business, how we evaluate what’s working, what’s not, and how everything ties back to revenue.
Someone in the comments jumped in and said, “Why not just let AI run the diagnostic?”
It wasn’t a joke. It was a real suggestion. And honestly, it’s a pretty accurate snapshot of where people are right now. There’s this growing belief that if you can feed enough information into AI, it can figure out what’s wrong and tell you what to do next.
On the surface, that sounds efficient. If AI can analyze patterns and summarize data faster than a human, why wouldn’t you use it?
The problem is, it assumes the inputs are solid. It assumes the data is clean, that the right things are being measured, and that the person asking the question actually knows what they should be asking. That’s a big leap, especially for businesses that are already struggling to understand their marketing performance.
This is where a little common sense gets lost. If you hand bad inputs to a machine, you shouldn’t expect good outputs. That’s true in finance, operations, and just about every other part of a business. But when it comes to AI, people tend to treat it like it can override that reality.
It can’t. It just processes what it’s given.
Most of the issues we see in marketing aren’t really execution problems. They’re definition problems. The ICP is vague, the CRM is messy, reporting is inconsistent, and revenue goals aren’t clearly connected to what marketing is doing day to day. There’s activity, sometimes a lot of it, but not much clarity around what actually matters.
If you take that environment and layer AI on top of it, you’ll still get answers. They’ll just be built on shaky ground. The outputs might look polished, but they’re still rooted in flawed assumptions and incomplete data.
That’s the part people miss when they say, “Just let AI handle it.” They’re treating the diagnostic like it’s a data exercise, when it’s really a thinking exercise.
If the problem was obvious, it probably would have been fixed already. The reason businesses need a diagnostic in the first place is because something isn’t clear. Either the wrong things are being measured, or the right things are being misinterpreted, or internal assumptions haven’t been challenged in a long time.
A good diagnostic isn’t just pulling reports and summarizing trends. It’s stepping back and asking better questions. Why are we measuring this? What are we missing? Where are sales and marketing out of sync? What are we assuming that might not be true anymore?
That kind of work doesn’t happen automatically, and it definitely doesn’t happen just by handing data to a tool. It takes perspective. It takes experience. It takes someone willing to slow things down enough to actually understand what’s going on before trying to optimize it.
None of this is a knock on AI. Once there’s real clarity, it becomes incredibly useful. It can speed things up, help test ideas faster, and make execution more efficient.
But it needs direction.
Without that, it’s just amplifying whatever is already there, good or bad. And if the foundation is off, it doesn’t correct the course. It just helps you move faster in the wrong direction.
That’s why the first step isn’t automation. It’s clarity. And most of the time, clarity doesn’t come from a tool. It comes from asking better questions than the ones you’ve been asking yourself.