If you’re a business owner, you’ve probably had this conversation more than once this year. Someone on your team, a peer, maybe your own board, asks some version of “what should we be doing with AI?”
It’s a fair question to ask. But as Devfinity CEO Shawn Barber has pointed out, it’s also a pretty generic one, and generic questions tend to get generic answers. Usually that answer is: buy everyone a subscription and hope something improves. Most of the time, not much does. The problem was never the tool. It was the question.
Why the “buy everyone a subscription” move doesn’t always work
A lot of owners feel pressure to do something with AI without being totally sure what that something should be. That pressure pushes people toward rolling it out everywhere at once instead of picking one real problem to fix. Our advice is simple: start with one real business problem. Pick one manual process, one unreliable dataset, or one team that is clearly overloaded. Fix that first, measure what changed, then expand.
A story you’ve probably lived some version of
We see this pattern often. A business pulls reports from several systems, exports them into CSVs, combines everything manually, then asks how to add AI on top. But AI does not fix fragmented systems or unreliable data. It just works faster on top of the mess that’s already there.
We’ve watched this pattern before
This isn’t the first time a technology has come with this kind of pressure attached. We have seen this cycle before. During the early cloud era, companies rushed to become ‘cloud-enabled’ before they had a clear strategy for why. AI is having a similar moment now. The technology may be different, but the risk is the same: buying tools before defining the business problem.
“The AI use case is the flashy penthouse everyone wants to talk about. It doesn’t mean much if the foundation underneath it isn’t solid.”
A better question to ask yourself
Before you put AI anywhere in your business, ask something simpler first: does this part of the business actually need it, and could you measure the return if you did it? It’s easy to point to one person being faster with AI. It’s a lot harder, and more honest, to ask whether the whole department or the whole business is running better.
So, skip “how can we use AI” for a minute. Start with “where does my business need to work better?” Once you know that, AI has something real to do. Until then, it’s just another subscription looking for a problem.
Before putting AI anywhere in the business, start with a simpler question:
Where does the business need to work better?
Maybe reporting takes too long. Maybe teams are entering the same information into multiple systems. Maybe managers don’t have a reliable view of what’s happening across the business. Maybe a process is consuming hours of manual work every week. Those are business problems worth solving. Then ask whether AI is the right solution. Sometimes it is. Sometimes the better answer is cleaner data, better integration, better automation, or a simpler process. And if AI is the right answer, don’t start everywhere at once.
Where Operational Intelligence Comes In
Start with where the business needs to work better. Understand where the business is losing time, where information is fragmented, where processes are breaking down, and what leaders need to see more clearly.
Then look at the systems, data, and workflows underneath that problem. Sometimes AI is the right answer. Sometimes the better answer is cleaner data, better integration, stronger automation, or a simpler process.
That’s where Operational Intelligence comes in. It gives businesses a clearer view of how the operation actually works so they can make better decisions about what to improve, automate, or build next. AI should be applied where it can create measurable value, not added simply because it’s available..
