Everybody in this industry talks about cost per lead. Almost nobody can tell you what a customer costs them.
Those are different numbers, and the gap between them is where most marketing budgets quietly fail. A lead is somebody who raised a hand. A customer is somebody who paid you. Only one of those pays the fuel bill.
The formula, and where it goes wrong
Customer acquisition cost is total marketing spend divided by customers acquired in the same period. That is the entire calculation, and it is trivial arithmetic.
Everything difficult about it lives in the phrase "total marketing spend," which most owners underestimate by a wide margin.
It includes the obvious: ads, lead platform fees, credits, sponsorship, print. It also includes the software you pay for monthly to support marketing, the wrap amortised across its life, the call tracking subscription, and anything you pay a person to write, design, or post.
And it includes your time, which is the input almost nobody counts.
Your hours are the largest line and the invisible one
If you spend six hours a month writing posts, chasing reviews, updating your profile, and answering platform enquiries, and your billable rate is $80, that is $480 of acquisition cost that never appears anywhere.
Owners resist counting this because it does not leave the bank account. But it is the most constrained resource in a small service business, and pretending it is free produces genuinely bad decisions. A channel that looks cheap in dollars and consumes eight hours a month is frequently more expensive than a channel that costs twice as much and runs itself.
Count the hours at whatever you would charge for them. The number will be uncomfortable and it will be correct.
A worked example
Say last month you spent $400 on Local Services Ads, $60 on a scheduling tool, $25 on credits at a lead platform, and roughly five hours of your own time on marketing, valued at $80 an hour.
That is $400 plus $60 plus $25 plus $400, which is $885.
If that produced six new customers, your CAC is roughly $148.
Now the second number that makes the first one meaningful.
CAC on its own tells you almost nothing
A hundred and forty eight dollars is expensive for a business whose average job is $200 and disastrous if those customers never return. It is cheap for a business whose typical customer is worth $1,400 over three years.
So CAC only has meaning next to lifetime value. The ratio people generally aim for is a customer worth at least three times what they cost to acquire, and for service businesses with genuine repeat work it should be considerably better than that.
There is a second question worth asking alongside it: how long until you get the money back. If your CAC is $148 and the gross profit on a first job is $90, you do not recover the acquisition cost until the second visit. That is survivable if they come back and fatal if they do not, and it tells you that retention is not a nice-to-have but the thing making the whole model work.
Blended CAC hides your worst channel
This is the most important part, and it is where the arithmetic actively misleads.
If you calculate one CAC across everything, referrals and repeat customers get folded in. Those cost you almost nothing to acquire, so they drag the average down and make the overall number look healthy.
Meanwhile a paid channel could be running at $600 a customer, and the blended figure conceals it completely.
So you need both. The blended number tells you whether the business as a whole is viable. The per-channel number tells you what to cut. Businesses that only track the first one keep funding the channel that is losing money, because the total looks fine.
Splitting it requires only that you know where each customer came from, which is the intake question and a spreadsheet column rather than software.
Give each channel its own clock
One complication worth handling deliberately. Channels report at wildly different speeds, so a naive monthly calculation punishes the slow ones.
Paid search shows up in the same month. Content and local SEO show up in three to six. A home show booth pays out over the following quarter. Word of mouth arrives whenever it feels like it.
If you divide this month's total spend by this month's customers, you are charging the content investment against results it has not produced yet, and the channel will look terrible right up until it does not.
The practical fix is to calculate CAC over ninety days rather than thirty, and to hold slow channels to a longer review window that you write down in advance.
What to actually do this week
Add one column to wherever you record jobs: where this customer came from. That single field makes per-channel CAC possible and it costs nothing.
Then add up last month's real marketing spend, including your hours, and divide by new customers. You will have a number within twenty minutes.
Then run it again per channel, and look for the one that is worse than you assumed. There is almost always one, and it is almost always the one that felt cheap.
The reason this matters more than lead volume
Most owners respond to a slow month by trying to generate more leads. That is the expensive answer.
Knowing your CAC lets you make three cheaper decisions instead. You can stop funding a channel that does not work, which frees money immediately. You can raise prices, which lowers CAC as a share of every job without generating a single additional lead. And you can invest in retention, where acquisition cost is close to zero and most service businesses have never tried.
None of those require more leads. All of them require knowing the number, and the number takes an afternoon to find.