Almost every distributor wants AI, and few have made it work.

A distributor makes money in the gap. It buys from makers, holds stock, and sells to thousands of customers at a markup. Out of that gap it pays for salespeople, warehouses and trucks. Margins are thin, so small changes in price or cost make a big difference to profit.

Almost everyone wants AI. Few have made it work. In one survey of 233 distribution leaders, 93% called AI a priority. Only 16% use it across more than one part of the business. Another survey of 300 distributors found the same gap: nearly all had AI projects, but only about one in ten had put it to full use.

Pricing is where the money is. In a business with thin margins, one point of better pricing can add more to profit than a year of cost cuts. Pricing is the top AI goal for distributors. Most of those trying AI pricing expect at least two points of margin. So far, only one in six has got there.

Order handling is where the time goes. Customers send orders by email, PDF and phone. People type them in. AI can read many of them. Wesco, a large distributor, now handles about 40% of its 3 million invoices a year with no person touching them.

The big distributors are going two ways with salespeople. Grainger uses AI to help its service staff, warehouses and website search. It still plans to add salespeople each year. Its sales grew 10% last quarter, with a 16% operating margin. Southern Glazer's, the largest alcohol distributor, moved some small accounts from sales visits to phone, website and AI. Both are betting AI makes each salesperson worth more. They differ on where people still pay off.

The data comes first. Most distributors have bought several systems that do not talk to each other. The ones using AI widely spent years cleaning up product and customer data before they scaled it. AI on messy data gives confident wrong answers.

Customers' software may start shopping for them. Large customers use purchasing software. If it learns to compare prices and stock across distributors and order by itself, service and relationships count for less. There are no public figures on this yet. It is a risk to watch, not a trend.

Time saved is not money saved. If AI reads orders faster, the saving is real only if you handle more orders with the same people, or hire fewer as people leave.

Here is what that looks like in practice. A distributor sells $300 million a year at a 25% gross margin, so it earns $75 million on the gap. An AI pricing tool raises prices a little on items customers do not shop around for. If that adds one point of margin, it earns $3 million more a year. If it raises prices on the wrong items and a large customer leaves, it can lose more than that.

What should a distributor's leaders do first? Three plain things.

1. Clean up product and customer data before buying more AI.
2. Test AI pricing on some product lines or branches, and watch both margin and customer loss.
3. Pick one slow, manual task, like typing in orders, and measure the hours before and after.

None of this needs a new team or a large budget. It needs one leader who owns each answer.

The full guide shows where AI touches the money, what has actually happened, who gains and who loses, and what to ask before you spend.