Every CEO and board is asking the same question right now: what are we doing with AI? Sales leaders feel the pressure to show something, and show it fast.
I’m genuinely excited about what AI can do for commercial teams. I use it myself. But it’s a tool, not a superhero coming to save the day. It works best when you understand what it’s good at, what it needs from you, and where people still have to step in.
In commercial work, I think about AI in two ways: quantitative work with numbers, and qualitative work with words. It helps with both, and both still need people.
Quantitative: clean data comes first
Ask AI about pricing, margin or forecasts, and it will give you a clean, confident answer. The catch is that it can only be as good as the data underneath it. A person looking at a messy report can usually tell something is off. AI doesn’t hesitate. It’s wrong quietly, and sounds sure.
In a manufacturing business, the problems are rarely dramatic. They’re the same customer set up three different ways, with no link between the ship-to, the bill-to and the parent company. They’re SKUs with inconsistent descriptions and units of measure. They’re contract prices, rebates and one-off deals that live in spreadsheets and inboxes instead of the system. Ask AI for pricing advice when it can’t see your rebates, and you’ll get advice on the wrong number.
That’s why a sound master data plan has to come first. It’s a leadership decision, not an IT project. Every area of data needs a named owner and clear rules for how information is entered and kept up to date.
Even with clean data, AI can tell you where to look, but not always why. Years ago, consultants we brought in showed us that we had dropped price, mid-contract, with our largest customer. They went straight to the obvious conclusion: the salesperson was giving away price. When I dug into the bill of materials, I found the customer had switched to a less expensive material. We had passed part of that savings along and kept some of it, and our margin had actually improved.
An AI pricing tool would flag that same drop in seconds. It still couldn’t have told us the reason. That lived in the bill of materials and in knowing how that customer’s business worked. With numbers, the human element is knowing the business well enough to ask why before you act.
Qualitative: where AI really shines
The other half of commercial work is qualitative, and it isn’t in a spreadsheet. It’s in emails, meeting notes, call summaries and customer conversations. Until recently, there was no good way to work with all of that at scale. You can’t write an Excel formula to find which customer is frustrated, or which team member keeps getting praised.
Qualitative data is where AI has surprised me most.
Before a meeting with one of our largest customers, an account with several people supporting it, I didn’t just ask the sales and customer success reps what to expect. I asked them to run a Copilot prompt across their own emails and meeting notes. It pulled together the open projects and issues, identified next steps and roadblocks, and summarized it all as a status update.
Then came the important part. Before sending it to me, each of them reviewed the summary, corrected what was off and added their own thoughts. I walked into that meeting with a much fuller picture, and I had time beforehand to ask questions and understand the day-to-day issues that might come up.
The same approach works for any pile of words your team already has. Imagine thirty customer interviews, or a year of meeting notes. Reading every one would take days. AI can help you find the themes: what customers keep praising, where service is slipping, which people behind the scenes customers see as unsung heroes. Questions like that used to go unanswered simply because nobody had the time.
With words, the human element is review and context. AI gives you a strong first draft. The people closest to the work make it accurate and add what only they know.
Where to start
Start narrow, with one problem and one clear measure of success. Summarizing notes and preparing for meetings are good first uses, because they save time right away and keep people in the loop. Use AI to help clean your data too, by spotting duplicates and flagging gaps. Save pricing and forecasting recommendations until your master data is in shape.
And bring your team along. If your salespeople don’t trust the data or the summaries, they won’t use them. When they’re part of reviewing and improving the output, they trust it, and the tool starts doing what you hoped it would.
The human element is the point
AI won’t save the day on its own. With quantitative work, it needs clean data and people who know the business to ask why. With qualitative work, it needs people to review it and add what only they know. Used that way, it gives your team time back and helps you see more clearly. That’s what a good tool does.