You know your close rate. You know your average ticket. You know what a maintenance-agreement customer is worth over five years versus what a one-time repair customer is worth. We don’t know any of that about your business, and anyone who tells you a lead is “worth $47” without asking a single question about your numbers is guessing on your behalf. What follows is the method, not the verdict — you run your own figures through it.
Start with the chain, because “cost per lead” is the wrong number to anchor on and it’s the number every ad rep will hand you first. It’s also the number that hides the leak points laid out in the first piece in this series — a lead can be cheap and still be worthless if it dies somewhere between the click and the booked job.
Cost per lead is what you paid to get a name and a phone number. Cost per booked job divides your total spend by the number of those leads that turned into an appointment on the calendar — this is where your close rate enters the math, and it’s usually a fraction of your lead volume, not most of it. Cost per acquired customer divides spend by the number of booked jobs that actually happened and got paid for, since not every booking shows. Three different denominators, three very different numbers, and the first one is the least useful of the three because it says nothing about whether the lead turned into money.
Here’s the step almost everyone skips: cost per acquired customer should be measured against lifetime value, not against the first ticket. If a maintenance agreement or a repeat-service customer is worth several times the initial call-out over the years you keep them, then a lead that looks expensive against a single ticket can be cheap against what that customer actually pays you over time. Conversely, if you have no retention play — no maintenance agreements, no follow-up, nothing that brings a customer back — then your real return is capped at that first ticket, and a lead only makes sense if it’s cheap enough to pencil out on that number alone. This is the single biggest reason two contractors in the same city can have wildly different opinions about whether “$90 a lead” is a good deal. They’re not disagreeing about the lead. They’re disagreeing about what happens after the first job, whether they realize it or not.
You can’t get outside data on your own close rate or your own lifetime value — that’s yours to know. What you can get from outside your business is a sense of what other companies spend on marketing relative to revenue, as a sanity check on your budget, not a target to hit. The Duke Fuqua / Deloitte CMO Survey has tracked marketing spend as a share of company revenue across industries for years; service businesses have generally run somewhere in the high single digits to low double digits of revenue, with plenty of variation by company size and growth stage. Check their current report rather than trusting a number here, because this figure moves. It’s a reference point for whether your total marketing spend is in a reasonable range — it says nothing about what any individual lead should cost, which depends entirely on your close rate and your ticket.
The other outside data point worth knowing: what the shared-lead marketplaces charge, because it sets a ceiling, not a floor. Angi’s own help center explains that lead fees vary by task, location, and demand, and that you’re charged once both you and the homeowner show interest — not once you’ve won the job. That’s a fundamentally different economic model than exclusive lead generation, where you’re paying for a shot at being the only bidder rather than paying into a pool of competing bids. We cover that difference and its arithmetic in the next piece in this series. For now, the point is narrower: whatever a shared-lead platform charges you in your market is a reasonable upper bound for what you’d pay for a lead you’re splitting three or four ways. An exclusive lead should be priced differently, because you’re not splitting the opportunity.
Run the worksheet like this. Take your last quarter’s marketing spend and divide it by booked jobs from that spend — that’s your real cost per booked job today, not an estimate. Multiply your average close rate by your average first-ticket value to get a rough revenue-per-lead number. Compare that to what you’re paying per lead. If the lead cost is comfortably below revenue-per-lead even before you factor in repeat business, you’re probably fine and the conversation is really about volume, not price. If it’s close or above, the decision hinges entirely on retention — can you get more than one ticket out of this customer, and realistically, will you.
Here’s the honest exception: if your business runs on install and replacement work with little repeat service — new construction accounts, one-off big-ticket jobs, no maintenance program — buying leads at marketplace prices is frequently the wrong move entirely. Without a second or third ticket to spread the acquisition cost across, you need the first job to carry the full cost of getting it, and a lot of lead pricing simply isn’t built for that math to work. In that situation, referral relationships and repeat commercial accounts usually beat purchased leads on a straight cost basis, because you’re not paying a stranger for every single introduction.
None of this replaces knowing your own numbers. It just gives you a chain to run them through instead of anchoring on the first dollar figure a sales rep quotes you.