Over the last couple of weeks, I’ve focused on where most businesses go wrong with AI—starting with the tool instead of the problem, and trying to automate processes that aren’t clearly defined yet.

This week, I want to shift gears a bit and talk about where AI is actually helping, because there are real opportunities if you approach it the right way.

In last month’s Ask the CFO session on AI with Ken Scales, we spent time breaking this down into practical areas where businesses are already seeing results.

What stood out to me is that the wins aren’t coming from broad “AI strategies.” They’re coming from very specific use cases tied to how the business runs.

Ken Scales talked about three areas where he’s consistently seeing impact.

The first is reducing manual work. This shows up in places where people are spending time moving data between systems, processing documents, or pulling together reports. These are usually not complicated problems, but they take up time and create delays. When you reduce that manual effort, you free up capacity pretty quickly.

The second is improving visibility. In one example Ken Scales shared, a business owner running multiple companies didn’t have a clear view of performance because everything was sitting in different systems. Once that data was consolidated into one place, it changed how decisions were made. Nothing about the business itself changed overnight—what changed was the ability to see what was actually happening.

The third is better decision support. This is where things get more advanced, but also more valuable. One example he shared was optimizing routing and timing in an operation, which ended up saving about $1.7 million over a few years. That didn’t start with “let’s use AI.” It started with a specific operational issue and a need to make better decisions with better information.

That’s the common thread across all of this. The value comes from solving something specific, not from implementing technology for its own sake.

If you want to make this practical, go back to the area you identified in Week 1 and refined in Week 2. Now ask yourself:

• Is there a repetitive task here that we could reduce or eliminate?
• Are we missing visibility into something we should be tracking more clearly?
• Are we making decisions based on information that’s delayed, incomplete, or hard to access?

You don’t need to solve everything. You’re just looking for one place where improving how the work gets done would actually make a difference.

That’s where AI can start to play a role.

If you start there, the technology becomes a tool that supports the business. If you start anywhere else, it usually becomes something you’re trying to force into the business.