By John Wheeler, Wheeler Intelligent Systems LLC
Suppose you own a business that earns $500,000 a year. A buyer offers $2 million. You decide to keep the business, put AI to work in a few places, and get annual earnings up to $650,000. A year later, another buyer offers $1.95 million.
Why would someone pay less for a business that earns $150,000 more?
The first buyer was willing to pay four times annual earnings. The second was willing to pay three times. That drop is called multiple contraction. I think it’s one of the more important ways AI could change what a business is worth. The same tools that help you earn more can make it easier for someone else to compete for those earnings.
What a buyer is paying for
An earnings multiple tells you how much a buyer is willing to pay for each dollar of annual earnings. At four times earnings, the buyer pays $4 for every $1 the business earns in a year. That number reflects what the buyer thinks will happen after the sale. Can the business keep its customers? Can it grow? How much will it have to spend to keep earning that money?
A buyer looking at last year’s financial statements is trying to answer questions about the next several years. Two companies can show the same earnings and get very different offers if one has steadier customers, better growth prospects, or fewer ways for things to go wrong. Risk, growth, and the ability to turn earnings into cash all matter. Aswath Damodaran, NYU Stern
Here’s the math using the same earnings measure in each case. The numbers are illustrative.
| Scenario | Annual earnings | Multiple | Implied business value |
|---|---|---|---|
| Before AI savings | $500,000 | 4× | $2,000,000 |
| Higher earnings, same multiple | $650,000 | 4× | $2,600,000 |
| Higher earnings, lower multiple | $650,000 | 3× | $1,950,000 |
Earnings rose by 30%, but the multiple fell by 25%. At three times earnings, the business would need to earn about $667,000 a year to get back to the original $2 million valuation. So an owner can do a good job cutting costs and still have some ground to make up at the time of sale.
How AI could push multiples down
Imagine you run a distributor. You use AI to help answer common customer questions, prepare quotes, and look up product details. Customers get answers faster, and your team spends less time handling routine requests. That can make a real difference. In one study of more than 5,000 customer-support agents, workers using an AI assistant resolved 15% more issues per hour on average. That result tells us something about the potential for productivity in one workflow. What happens to profits depends on the business. Generative AI at Work, The Quarterly Journal of Economics
Now imagine a competitor puts those same tools to work. You both have lower costs, and both of you are trying to win the same account. The next time that account goes out for bids, one of you may give up part of the savings to win the order. Customers might also come to expect faster quotes and better product information as part of the usual service.
AI can also change the buyer’s choice between buying a business and building one. If it used to take two years and a large team to set up a service company, buying an existing one could save a great deal of time. If a capable operator can now set up much of the operation with a smaller team, starting from scratch becomes a more serious option. The seller may still have something the buyer wants, but the buyer has more room to walk away from a high price.
These changes can affect the multiple before they show up in earnings. A business can report its best year ever while a buyer is already figuring in slower growth, lower prices, or a higher cost to keep customers. If the buyer expects lower ongoing earnings, they may also use a smaller earnings figure in the first place. Those are two different judgments. One is about how much the company can earn under normal conditions. The other is about what those earnings are worth given the growth and risks ahead.
Where AI could support a higher multiple
Take a food manufacturer with a track record of shipping the right product, in the right quantity, on time. A new competitor might use AI to put together a website, write product descriptions, and prepare a draft spec sheet. Winning the account still takes a supplier that can make the product consistently and fill the order when the distributor needs it.
The manufacturer can use the same tools to respond to product questions sooner, spot gaps in its records, and serve more accounts with the same office staff. If buyers see that customers stay and the company can grow without costs rising at the same pace, they may be willing to pay more per dollar of earnings.
AI could also make a business easier to hand over. Think about a company where the owner is the only person who knows which jobs to quote, which customers get special terms, and what to do when a delivery goes wrong. A buyer has to figure out how much of that knowledge walks out the door at closing. If the owner can put those decisions into clear procedures, track exceptions, and show that someone else can run the work, the buyer has a better basis for trusting the earnings. AI can help carry out parts of those procedures. The sale value comes from being able to show that the operation works under new ownership.
The question I’d ask as an owner
At Wheeler Intelligent Systems, I’m interested in how far one person can go with AI doing much of the routine work. The valuation question adds another test. Suppose someone else can build a tool that looks like mine and performs many of the same tasks. What would customers still pay my company to do?
For a Business Central tool, the answer might be knowing how a manufacturer actually runs production, handling the cases that come up when a job goes off plan, and giving a manager a clear record of what happened. To a buyer of the business, I’d also have to show that this knowledge lives in the product and in the way the company operates. If it lives mostly in my head, a small payroll may make the business profitable while leaving the buyer with a hard transition.
I expect AI to make the gap between these kinds of businesses easier to see. Some owners will have a few good years from cutting costs and find that buyers already expect everyone in the field to catch up. Others will use the savings to improve something customers have a hard time finding elsewhere, and they’ll be able to show that the next owner can keep delivering it.
If I were preparing to sell, I’d want to be able to answer two questions with evidence. How much more does the business earn because of AI? How much of that extra profit can the next owner still count on after competitors have the same tools?