The honest answer annoys people, so here it is first. There isn't one. The industry figures you'll find are averages taken across businesses that have nothing in common with yours, and no benchmark survives contact with a real margin. A customer who costs $300 to acquire is a bargain for one business and a slow bleed for the shop next door, and what separates them has nothing to do with what industry they're in.
What exists instead is a ceiling, and it belongs to you alone. Below it almost any acquisition cost works. Above it none does, and no amount of volume fixes that.
The reason this question usually gets answered with a percentage instead is worth a minute of your time, because that percentage is the thing standing where your ceiling should be.
Where does the percent rule come from?
It isn't made up, which is the part worth understanding before discarding it. Percent-of-revenue budgets are a real planning tool, and they work for the companies that invented them. A business with a marketing department has salaries to forecast a year out, an agency on retainer, a trade show it commits to in February, and a brand budget that can't be traced to any individual sale. For that business, a percentage is a sane way to size a standing commitment against a standing revenue base.
Almost none of that describes a five-person business. You have no department, no retainer, and no line item that resists attribution. The costs those rules were built to govern are somebody else's cost structure, rational for the people carrying it, and carried into your planning by a search result that never asked what you sell.
The ad platforms don't help, and they aren't villains either. Their pricing runs by auction against every other bidder who wants the same click in your county, which is exactly what an auction is for. That price is real, and it is also entirely uninformed about your margins. Nothing in the system knows whether the customer it just sold you is worth having.
What does a new customer cost you?
Start with the number you can compute today. Customer acquisition cost is your marketing spend divided by the new customers it brought in over the same stretch.
Say you run heating and air. You're putting $2,400 a month into ads, the truck wrap, and the directory listings, and eight new customers a month come out the other end. Each one cost you $300 before they've paid you anything. That's your CAC, and it's arithmetic rather than opinion. The tool's customer acquisition cost section runs it on your figures.
Whether that's fine or alarming is the whole question, and it can't be answered by comparing it to anybody else's $300.
So what is a good acquisition cost?
Here's the comparison that settles it. A new customer is worth what they leave behind in margin, not what they pay you, and the honest window is about a year.
An average job runs $480. After the parts, the truck, and the tech's hours, you keep 40 percent, so that's $192 of contribution. In a year the typical customer calls you 1.3 times, between the service call and whatever breaks. That's roughly $250 of margin from one customer in a year.
Which means the ceiling is $250, and you're paying $300. Every new customer arrives about $50 in the hole, eight times a month, and the business has been quietly funding $400 a month of that from somewhere else. Nothing in the P&L says so. Marketing shows up as a cost, new customers show up as revenue, and the two never meet on a page.
Now the interesting part. There are two ways out of that hole and only one of them is about marketing. You can cut the $300, which is slow and mostly means bidding less and waiting. Or you can move the 1.3. A maintenance plan that gets a customer calling twice a year plus a tune-up takes the yearly margin to about $420, and the same $300 that was losing money becomes a comfortable buy. The spend question section on the tool page makes the same point in the other direction, which is that the defensible budget falls out of your own unit economics and nowhere else.
The number nobody has
Notice what that calculation needed. Average sale, gross margin, purchase frequency, monthly spend, and new customers per month. Four of those five are sitting in your bank statements and your invoices.
The fifth one is why most owners never run this. Almost nobody can say how many of last month's customers were new. The point-of-sale knows, the scheduler knows, the invoicing software knows, because every one of them stored a name and a date. Nobody has ever asked it the question, so the answer exists and isn't available, which is a different problem from not having the data. That same silence keeps a POS and a marketing platform from ever comparing notes.
That single missing field is the reason the percent-of-revenue rule survives. It's the only marketing budget you can set without knowing anything about your own customers, and its real appeal was never that it's accurate. If you do one thing after reading this, get that number. Pull last quarter's invoices, count the names that hadn't appeared before, and divide by three. It's an afternoon, it's exact enough, and it converts every other number you already have into a decision.
A dental practice can do the same with new-patient forms. A law practice can do it with matter numbers. The specific report doesn't matter. Asking the question is what nobody does.
When the math works and the timing doesn't
One more thing the ceiling won't tell you, and it's the one that actually closes businesses.
The ceiling is annual. The invoice is monthly. If a customer is worth $420 across a year and you paid $300 to get them, that's a good trade and it takes seven months to become one. Sign forty of them in a quarter and the arithmetic is delighted while the checking account is in real trouble, because you paid $12,000 in March for margin that arrives through the following February.
That gap is a cash problem wearing a marketing problem's clothes, and it has known moves, none of which are "spend less." Stagger the acquisition. Take a deposit. Sell the maintenance plan at the first visit rather than the third, which pulls the payback forward without changing the annual value at all. If you're the one reading this at eleven at night with both numbers in front of you, that gap is worth a conversation.
Running your own numbers
The marketing break-even calculator takes the five fields above and hands back your CAC, what a customer is worth over twelve months, the ceiling you can afford, and how long the payback takes. It'll also tell you plainly when the answer is a guess, because a customers-per-month field you estimated makes every number downstream that same estimate, worked through.
There's a variable underneath all of it that this page has been circling and never priced. Every figure here leaned on the 1.3, and the 1.3 is retention wearing a statistician's hat. A customer who never comes back is worth exactly one job, whatever the average says, and no amount of acquisition math improves a business people don't return to. Which raises a harder question than the one you came here with, and it isn't answered by a budget.