Random Acts of AI

The strategy problem we already had a name for, now with a new tool attached.

You know the phrase "random acts of marketing." A webinar because a competitor ran one. A trade show booth that was critical because XYZ would be there but no one supported the plan behind it. Individually, each one seemed reasonable. Together, they added up to a lot of activity and not much momentum.

We've been talking about that problem for twenty years. I think we have a new version of it now, and most of us haven't named it yet.

Random acts of AI.

A chatbot added to the website because sales wanted one. An AI tool for social captions because someone saw a demo. A pilot for personalized email because a vendor offered a free trial. Each one, again, seems reasonable in the moment. None of them connect to the others, or to anything you were already trying to do.

Same disease. Faster symptoms.

Why this version moves faster

Random acts of marketing used to take a quarter to pile up. A tactic here, an initiative there, eventually a messy portfolio of half-finished ideas.

AI compresses that timeline. A tool can be trialed, deployed and embedded into someone's workflow in a week. By the time anyone asks "wait, why do we have three different AI writing tools," three different teams have already built habits around three different answers.

The old version of this problem gave you time to notice. This one doesn't.

How to spot it

Here's the tell, the same one that gave away random acts of marketing all along: nobody can answer "why are we doing this, what strategy is it reinforcing?"

Ask it about any AI tool or pilot running in your org right now:

• What business outcome is this supposed to move?

• What would we stop doing if this works?

• Who owns whether it's actually working?

If the honest answer is "we're not totally sure, but it seemed smart to try it," you've found a random act. Not a reason to kill it necessarily. A reason to notice it.

The fix isn't a moratorium

I'm not suggesting a committee that slows everyone down for six months while it writes an AI policy nobody reads. That just trades one problem for another.

The fix is closer to what should have fixed random acts of marketing in the first place: A short list of what you're actually trying to accomplish, and a habit of checking new ideas against it before they launch, not after.

That's it. Not a framework with twelve boxes. A page, maybe a paragraph, that says what matters this year, and a standing two-minute question whenever something new shows up: does this serve the list, or are we just excited about the tool?

Some of what you're already running will pass that test easily. Some won't, and that's useful information, not a failure.

Enterprise vs. mid-market

For enterprise teams, the risk is scale: by the time leadership notices the pattern, a dozen teams have each adopted their own AI tool for the same job. Put one person in charge of knowing what's running where, not approving it, just knowing. You can't connect activity to strategy if you can't see the activity.

For mid-market teams, the risk is speed without a filter: a small team can say yes to every demo because nobody's stopping them, and "yes" to everything quietly becomes a strategy you never chose. The fix costs nothing but a habit: ask the three questions above before the next tool gets added, not after.

The question underneath the question

Here's what I keep coming back to. We spent two decades teaching marketing teams that activity isn't strategy. That busy isn't the same as effective. AI didn't change that lesson. It just gave us a faster way to forget it.

The leaders who get real value from AI this year won't be the ones who adopted the most tools. They'll be the ones who could explain, for every single one, what it was actually for.

So here's my honest question for you, not as a gotcha, as something worth two minutes with your own list:

How many AI tools are running in your organization right now that you could not explain the strategy behind?

If the number is more than zero, you're not alone. You've just found your next conversation.