What you can afford to pay for a customer

The usual approach starts at the budget: we’ll try $1,500 a month and see how it goes. That runs the chain forwards, and it makes the outcome a matter of opinion — the campaign either “felt worth it” or it didn’t. Run the same chain backwards, starting at the customer, and the outcome stops being an opinion.

• Average job value. The typical ticket.
• Gross margin on that job. What survives materials and the labour on the job. Margin, not revenue.
• Jobs per customer, lifetime. What your records can evidence, not what you hope for.
• Lead → job close rate. Of the people who call or send the form, how many you win.
• Click → lead conversion rate. What your landing page does with the traffic it is sent.
• Cost per click. The only one that comes from outside the business, and the platform will quote it to you for free.

Two lines connect them, and the second is the whole article:

• What a customer is worth = job value × margin × lifetime jobs. In gross profit, not revenue.
• The most you can pay for a click = that number × your click-to-lead rate × your lead-to-job rate.

Everything below is either how to get those inputs honestly, or what to do when the answer comes back badly. Run it on your own figures first.

Three things go wrong in that calculation, and all three go wrong in the same direction — they make you believe you can afford more than you can.

• Revenue put where margin belongs. A $450 job at 45% margin is worth about $202 to you, not $450. Enter the ticket instead of the margin and every number downstream doubles, including the one you use to justify the spend. This is the single most common break in the model, and it is invisible because the result still looks like a sensible figure.
• Gross profit read as take-home. What comes back up the chain is gross — before the van, the insurance, the phone, the software, and whatever you pay yourself. A campaign returning two dollars for one is frequently a net loss once overhead lands on it.
• Lifetime value used as a wish. Repeat business is where optimism hides, because nobody can disprove it this quarter. If you cannot pull the number out of your records, the honest input is 1.

This is also where the familiar “3:1” target comes from, and it gets quoted as though it were a law of physics. It is not. It is a margin for error. Every input above it is an estimate, and the ratio is what absorbs the estimating: at 5:1 you can be wrong about your close rate, your page and your repeat rate simultaneously and still make money. At 1.5:1 you have to be right about all three — and you will not be.

When the arithmetic comes back thin, the reflex is to lower the bid — or to pause the campaign, which is the same move at full strength. Neither is a lever. Bidding less does not change what a customer is worth to you; it just buys fewer of them. Only three things move the break-even price, and they are not equally expensive to move.

• Your close rate. Cheapest, and routinely ignored. Most owners treat this as a sales-skill problem. It is usually a logistics problem: who picks up, how long the caller waits, and whether anyone gives a price on the first contact. Ten points here is often available purely from answering faster, and it costs nothing.
• Your landing page. Second cheapest. The page has to repeat the promise the ad made — ideally in the same words — and then offer exactly one thing to do. Every additional choice on the page is a subtraction from that one. And count a lead as a completed form or a call that lasted long enough to be a real conversation, never a page view or a button press, or you will optimise the page towards traffic that never becomes work.
• What a customer is worth. Slowest, and largest. Raising the ticket or earning the second job moves the ceiling on everything else at once, and unlike the other two it improves every channel you have, not just the paid one. It is also the only lever that takes quarters rather than weeks.

The honest order is close rate, page, customer worth — fastest return on effort first. If none of the three can be moved far enough, that is not a failure of the campaign. It is the arithmetic telling you something true.

The model above is worth nothing if the inputs are guesses, and three of them are the kind of guess that flatters. Before spending, make sure you can actually see:

• A conversion event that fires on a real lead. A submitted form, or a call that passed a duration threshold. Not a page view, not a click on the phone number — those two are why so many dashboards report a healthy cost per lead for a month in which nobody was hired.
• Call tracking, if the phone is how people reach you. Without it the channel that works is attributed to nothing, and the channel that merely gets clicked takes the credit.
• The source, recorded at the moment you win the job. This is the only way a close rate by source becomes real instead of assumed, and it is usually where an owner discovers their tools do not talk to each other.
• Enough volume for the numbers to mean anything. At twenty leads, one extra win moves your measured close rate by five points. So treat any swing smaller than that as noise, and be suspicious of your own conclusions until the count is in the hundreds. Killing a campaign on twelve leads is a coin flip with a rationale attached to it.

Ads are a multiplier on a business that already converts. Applied to one that doesn’t, they multiply that instead — faster, and at a price per click.

• Your break-even click price is below what the auction charges. That is not a campaign problem and no amount of bid tuning touches it. Either move a lever until the number clears, or use the channels that are not priced per click: the business profile, reviews, referral, and the calls already coming in.
• Nothing answers the lead within the hour. A paid lead decays exactly like a free one. Buying attention you are not staffed to respond to is the most expensive available method of annoying people.
• You are already at capacity. More leads is not more revenue when the constraint is hours in the day. It is a longer waiting list and more people to disappoint.
• You are new, with no reviews yet. Worth stating plainly because it is genuinely unfair: you pay the same price per click as the established competitor and convert worse on identical traffic, so your break-even is harder to reach than theirs. That gap is not an advertising problem and cannot be solved inside an ad account. Build the profile first, then buy the clicks.

None of that is an argument against advertising. It is an argument for knowing which of these you are in before the money starts moving — because all four look identical from inside the dashboard, where every one of them shows up as a disappointing cost per lead.