Give an owner an AI budget and watch where it goes. Almost every time: marketing. More content, more leads, more top-line. It's intuitive — the phone rings more, the dashboard climbs, it feels like progress. But ask what actually reached the owner's pocket at year-end and the answer is usually a thin slice of all that motion. Pointing AI at the top line is the most expensive, least certain way to make an owner richer. And almost nobody runs the math before they do it.
Run the math
Say your business nets a 10% margin. To put one extra dollar in your pocket through growth, you have to generate ten dollars of new revenue — and pay to acquire, serve, and deliver all ten. Now flip it. Use AI to cut one dollar of waste from operations and the entire dollar drops to the bottom line. No acquisition cost. No fulfillment cost. No risk the customer churns before you earn it back. One dollar saved is one dollar of pure profit.
At a 10% margin, that single saved dollar is worth the same as ten dollars of new sales. That isn't motivation — it's arithmetic. And it gets sharper the thinner your margins are. So when you're deciding where to aim your first AI dollar, you're really choosing between chasing ten dollars of revenue or banking one dollar of certain profit. They land the same. Only one of them is free.
$10 : $1
At a 10% net margin, ten dollars of new revenue and one dollar of cost cut put the same money in the owner's pocket — but only one of them costs you nothing to acquire. AI is better at the cheap one.
Marketing buys revenue you split with everyone who helped earn it. Efficiency buys profit you keep all of.
Why owners reach for the wrong lever
If the math is this clear, why does everyone point AI at marketing? Because growth is glamorous and efficiency is boring. A new AI campaign is a story you tell at the conference. Using AI to kill a redundant step, compress a three-month training curve, or catch leakage in the funnel — none of that makes a good slide. There's a deep bias toward addition over subtraction: research in Nature found people overwhelmingly try to improve things by adding rather than removing, even when removing is obviously better. Owners point AI at "more" because more feels like doing something. Pointing it at "less waste" feels like admitting there was waste.
There's also a visibility problem. Marketing spend has a dashboard. Operational waste doesn't — it hides inside "the cost of doing business," and nobody owns the line item for "things we do that we don't need to." So the visible lever gets the AI budget and the invisible one gets ignored, even though the invisible one is where the free money lives.
Where AI prints the cheap dollars
- Process you've automated badly or not at all. Every step a person does that AI could do is a recurring cost you're choosing to keep paying.
- Training overhead. If onboarding takes months and half your hires leave, you're funding a leaky bucket. AI that compresses the training curve saves money that compounds every hire.
- Rework and leakage. Files that come back, cases that fall out of the funnel, the gap between what you accept and what you convert. Every one is margin walking out the door.
- Vendor and tooling drift. Costs that made sense at one size and never got revisited at the next.
Top-line growth is a story you tell investors. Bottom-line savings is a check you write yourself.
The owner's sequence
None of this means never market. Growth matters, and there are seasons to press it hard. The point is sequence: before you point AI at chasing ten dollars of revenue, point it at the dollar of waste that gets you the same result with none of the risk. Most owners have years of those dollars sitting untouched because they chased the glamorous lever first. Warren Buffett built part of a fortune on exactly this asymmetry — relentless cost discipline at businesses everyone else was trying to grow. It isn't exciting. It's just where the money is.
Point AI at profit before you point it at marketing. The cheapest, surest, fastest dollar of profit an owner will ever make is the one they stop wasting — and AI is the best tool ever built for finding it.
Sources & further reading
- Adams et al., Nature — "People systematically overlook subtractive changes." nature.com The cognitive bias toward adding (more marketing) over removing (cutting waste).
- McKinsey & Company — "The State of AI." mckinsey.com Where companies are actually capturing AI's bottom-line impact versus where they're spending.
- Berkshire Hathaway — Buffett's annual shareholder letters. berkshirehathaway.com Decades on cost discipline as a durable source of owner profit.