How much marketing can a contractor add before the percentage breaks?

How much can a contractor add to the marketing budget before cost of marketing crosses the target?

It depends on the marginal return of the added dollars, and the arithmetic is one line: blended cost of marketing after the addition equals (current spend plus added spend) divided by (current net revenue plus added spend divided by the added spend’s own cost of marketing). A per-channel rule and a blended target are two ways of asking that.

Why a channel rule and a blended target are the same question

Leadership at the eight-figure Toledo exteriors company where I direct marketing put it to me two ways in the same month. First: we should spend as much as we can on any channel that runs at 22% cost of marketing or better. Second: what would happen if the whole budget ran at 20% instead of 15%. Those sound like different questions, one about channels and one about the total, and they are the same question, because every dollar added to a 22% channel pulls the blended number toward 22%, and how far it pulls depends only on how many dollars are added and what they return.

The 22% rule has a hidden assumption in it, which is that the next dollar into the channel returns what the last dollar did, and I have written about why it usually does not (Why does my cost per lead go up when I spend more on Google?). Set that aside for a minute and take the rule at face value, because the arithmetic is worth seeing on its own.

The formula

Call current spend S, current net revenue R, the amount to be added X, and the cost of marketing the new dollars will run at m, as a fraction. The new blended figure is (S + X) ÷ (R + X ÷ m). That is the whole thing. The X ÷ m term is the revenue the new dollars buy, and it is the term nobody has a report for, because the report only shows what the old dollars bought.

A worked example

Take a shop, not ours, doing $10M net on $1.5M of marketing, which is 15%. Add $200,000 to a channel the vendor says runs at 33%. The new dollars buy $200,000 ÷ 0.33, about $606,000 of net. The new blended figure is $1.7M ÷ $10.606M, which is 16.0%. Add $530,000 at the same 33% and it is $2.03M ÷ $11.6M, or 17.5%. Add $1.27M and the blend crosses 20%.

So on that shop’s numbers the 15% target admits nothing at a 33% marginal rate, the 17.5% line admits about half a million, and the 20% line admits about $1.3 million. That is what the question “what if the budget ran at 20%” actually means: it means adding well over a million dollars of spend at a third and installing the jobs that come with it, and nobody who asks the question the short way has that number in their head.

Why the answer is always “it depends on the new dollars”

Every term in the formula is on a report except m, and m is the one that decides everything. At m of 15%, the new dollars are as good as the old ones and the blend does not move. At m of 50%, a billboard on an assumption, the blend crosses 15% almost immediately. And m is not a rate that holds at any volume, because a channel’s return at twice the spend is not its return at the current spend, so the real answer is that m itself falls as X rises and the honest way to find it is in increments the channel has run before.

When the 22% rule and the 20% question came up, the useful thing was not arguing with either one. It was writing the formula down and asking what m was, because at that point the number was visibly a guess, including to me.

Where this is wrong

The formula assumes the jobs can be installed. Adding $1.3 million of marketing at a third is about $3.9M of net, and at our 2025 average net ticket of $9,457 that is roughly 400 more jobs, and whether the crews and subs can put on 400 more roofs is not a marketing number. Sales can grow by decision. Production cannot. Any answer to “how much can we add” that does not have a production number next to it is half an answer, and it is the half that does not cost anything to say.

What to do

• Tonight: put S, R and the target percentage on one line, and solve for the X that crosses the target at three values of m: last year’s blended rate, the vendor’s quoted rate, and double the vendor’s rate.
• This quarter: add spend to any channel in an increment it has run before, and read the marginal cost of marketing on that increment at ninety days.
• Ongoing: attach an install-capacity number to every budget increase before it goes to leadership. If nobody knows how many more jobs the crews can take, that is the first number to get.

Sources: Arithmetic is the author’s. Worked example is hypothetical and labeled as such. The 22% and 20% figures are internal thresholds under discussion at the author’s company, stated by leadership.

Related notes