Cost of Marketing

A Demo Costs $992 Now. It Is the Best Money on the Sheet.

We already published the lead-cost half of this story: cost per lead quadrupled while the business got healthier. This note is the appointment-level sequel, because the demo is where the money actually changes hands, and the demo economics tell the story even more sharply than the lead economics do.

Between 2021 and 2025, cost per demo at the eight-figure Toledo exteriors company where I direct marketing nearly tripled, from $337 to $992. Net revenue grew from $8.4M to $13.0M over the same window. If your unit costs are climbing, that pairing is the whole lesson: a rising cost per demo is not a failure signal on its own. It only becomes one when nothing downstream rises with it.

Why we let the demo get expensive

Some of the increase was the market. Every contractor in the region saw lead costs climb after 2021, and anyone buying Google Ads or aggregator leads felt it directly. But a meaningful share of the increase was chosen. Spend shifted away from cheap, low-intent lead volume and toward channels that produce fewer, better appointments.

The clearest evidence is in the funnel. Set-to-sale went from 15.8% in 2021 to 25.4% in 2025. The company was setting roughly half the appointments and converting a far larger share of them into contracts. Chasing the old cost per demo would have meant chasing the old lead quality, and the old lead quality is what produced a 15.8% set-to-sale rate in the first place.

The metric with the authority to declare a problem

Cost per demo is an input price. The metric that decides whether the price is acceptable is cost of marketing as a percentage of net revenue, COM%. Ours ran 12.1% in 2021 and 19.0% in 2025, against a target band of 10 to 15%.

That 19% is above where I want it, and I will not pretend otherwise. 2025 was a year of heavy reinvestment in brand channels whose payback runs longer than a calendar year. Through early May 2026 the company is back at 13.9% COM on the way to a $16M revenue goal. The point stands: COM% is the number with the authority to declare a problem. Cost per demo alone does not have that authority, because it says nothing about what the demo turns into.

The cheapest demos we ever bought were also the ones least likely to become revenue.

The three questions before you cut anything

First, is close rate rising with the cost? If the demos are getting more expensive and closing better, your mix is shifting toward quality and the blended math may be improving.

Second, is average ticket rising with it? A more expensive appointment that produces a larger job can be a better appointment.

Third, what is happening to COM%? If cost per demo doubled but marketing spend as a share of revenue is flat or inside your band, you do not have a cost problem. You have a mix shift. The full framework is here: COM%, the number that governs a contractor marketing budget.