How many leads does a contractor need to book one job?

You know what you spent on marketing last month, and you know how many jobs you booked off it. What you probably can't say is how many leads it took to get one of those jobs, or whether that number is normal for your trade. Most owners who try to work it out are counting the wrong thing.

Here's the short version. Count a lead the way the benchmark data does, meaning a real prospect you actually spoke to, and leads per booked job is just your close rate upside down. On the home services average, that's a little over two. The number that actually separates one trade from another is how many raw inbound calls it takes to book one job. Run Invoca's 2026 home services benchmarks through the math and that's about 8 calls for a plumbing shop and about 15 for an HVAC company.

That gap isn't a selling gap. Once a person picks up, the HVAC shop actually books slightly more jobs per conversation than the plumber does. The whole difference is set before the conversation starts, by how many of those calls reach a person at all.

What counts as a lead before you can count anything?

Ask three owners for their lead-to-job ratio and you'll get three numbers that can't be compared, because none of them are counting the same things.

One counts every form fill. One counts every call that lit up the phone, robocalls and wrong numbers included. One counts only the people who got as far as asking for a price.

The benchmark data draws the line in a specific place, and it's worth borrowing. In Invoca's 2026 home services report, 38% of calls answered by a person are leads. A lead, in other words, is a conversation that turned out to be real demand.

So a lead isn't a ring. Everything upstream of the conversation is traffic, and mixing the two is what makes one shop's ratio look like a disaster next to another shop's that's doing exactly the same work. We took the dollar version of this apart in what a local service business should spend to capture one lead.

Why does it take three rates to get from a ringing phone to a job?

Because a call has to clear three separate steps, and each one has its own rate.

Does a person answer it. Across home services, 52% of calls are answered by a person (Invoca, 2026). The remainder isn't all lost work. It includes automated handling, misdials, and quick hangups. But it's the gate everything else sits behind.

Is it real demand. 38% of the calls answered by a person are leads, per the same report.

Does it book. And 45% of those leads convert on the call.

Which gives you the whole thing in one line:

calls × answer rate × lead rate × close rate = jobs booked on the phone

Push 100 inbound calls through the home services averages and you get about 20 leads and about 9 jobs booked on the call. That's a little over two leads per booked job, and about 11 raw calls to get there.

Now notice something about that first number. Leads per booked job is 1 divided by your close rate and nothing else. The answer rate cancels straight out of it.

So it isn't really a funnel measure at all. It's your closing percentage wearing a different hat, which is exactly why it makes a poor benchmark. The number with real information in it is the other one: how many calls you burned to get there.

What does the math look like for a plumbing shop?

Here's where the home services average splits by trade. Invoca breaks all three rates out by sub-industry, and they move a lot from trade to trade. The answer rate alone ranges from 32% to 74%, with plumbing near the top at 74%.

For plumbing, the same report puts the lead rate at 30% of answered calls and the close rate at 55% of those leads, converting on the call. So a plumbing shop taking 100 inbound calls:

• 74 answered by a person
• about 22 of those turn out to be real leads
• about 12 of them book on the call

Call it 8 inbound calls per booked job, and a little under two leads per booked job.

Why does the same math give an HVAC shop a different answer?

HVAC answers 34% in that same breakdown. Two trades that look identical from outside, both phone-driven, both emergency-heavy, both selling to the same homeowner at the same kitchen table, and one of them gets a person on the line more than twice as often.

The other two rates move as well, and not the way you'd guess. Invoca puts HVAC's lead rate at 45% of answered calls, well above plumbing's 30%, and its close rate at 43%, below plumbing's 55%.

The same 100 inbound calls, at an HVAC company:

• 34 answered by a person
• about 15 of those turn out to be real leads
• about six and a half of them book on the call

That's about 15 calls per booked job, nearly twice plumbing's 8. Same caveat as before: the calls that didn't reach a person aren't a stack of stolen jobs. They're automated handling, misdials and hangups mixed in with real people who didn't get through.

HVAC's higher lead rate does more than soften that. It more than covers the lower close rate: per 100 calls that reach a person, HVAC books about 19 jobs to plumbing's about 16 and a half. Once someone picks up, the HVAC shop is ahead. The whole difference in calls per job comes from the first step. The report doesn't say why the answer rates differ, and there's no sense inventing a cause it never measured.

Now look at leads per booked job, which moved too: about 1.8 for plumbing and about 2.3 for HVAC. That isn't a funnel effect. It's exactly the close rate, inverted: plumbing closes more of its leads on the call than HVAC does, and one divided by each close rate is the whole story.

Which makes leads per booked job the less useful half of the answer. It's a fine sanity check, and it won't tell you much about your shop that your close rate wasn't already saying. The number that separates one trade from the next is how many calls it takes to book one job, and that's the one you can do something about.

Where do owners get this number wrong?

Three ways, and all three point the blame at the wrong end of the funnel.

They count rings as leads. Divide jobs by total inbound calls and you get a number that looks like a closing problem. It isn't. You've buried the answer rate and the junk calls inside a ratio that's supposed to be measuring your sales conversation.

They stop the clock at the end of the call. That 45% is conversion during the call. A quote accepted on Thursday, a homeowner who texts back after the weekend, an estimate that finally lands on the second follow-up: none of those are in it. If nobody is working that tail, the jobs in it don't exist.

They treat the ratio as a verdict on their marketing. Usually it isn't. Two of the three rates happen after the call has already arrived and already been paid for. Which is the same trap we picked apart in why you're losing jobs you already paid to win.

Which of the three rates is easiest to move?

Not the one most owners work on.

Buying more leads multiplies the whole chain by a bigger number and leaves all three rates exactly where they were. It's the expensive lever, and it's the one that doesn't compound.

If the number you're trying to move is leads per job, the close rate is the only lever that touches it, and it's a cheap one to start on. In the same Invoca report, on 55% of the home services calls Invoca scored, nobody asked the caller to buy or book the job. That's a majority of companies leaving the one question the call was for unasked, and since leads per job is just the close rate inverted, it's the rate that moves that ratio at all.

Then there's the tail, the leads that don't book while you've got them on the phone. HBR's 2011 research is the standard reference here, and the caveat is that it studied web-form leads rather than calls, so read it for direction: in a study of 1.25 million sales leads at 29 B2C and 13 B2B US companies, firms that tried to contact the lead within an hour were nearly seven times as likely to qualify the lead as those that waited even an hour longer, and more than 60 times as likely as those that waited a day or more.

A separate audit of 2,241 US companies in the same article found 23% never responded to a web lead at all.

The answer rate is the easiest of the three to move, because it's the most mechanical. A call is either answered or it isn't. There's no technique in it and nothing to train. It's a coverage problem wearing a sales problem's clothes.

The rate you can change this month

Willison answers every inbound call, 24/7, in seconds, effectively on the first ring. It answers them all at once, so the fourth caller in a heat wave gets picked up as fast as the first. It asks your intake questions, qualifies the job, and books it straight onto the calendar.

Then it works the part the benchmark ignores: the follow-up text to a lead who went quiet, the check-back on an estimate nobody answered.

Be honest with yourself about the size of that, though, because the arithmetic above cuts both ways. Answering the calls you're currently missing also means answering more robocalls and more wrong numbers. Those don't become leads at 38% or at any other rate, and nothing on the phone creates a truck: if your August calendar fills by nine in the morning, a faster pickup gets you better information about which calls are waiting, not more capacity. Some callers also just want a person, which is what an escalation contact is for.

What does change is that the first step stops deciding which calls you ever hear about. Whatever your calls-per-job number is today, it was measured with some share of your calls never becoming conversations at all, and you don't know which ones.

If you want to hear what one of those calls sounds like, press play on a real call on willisonhq.com and follow it from first ring to booked.

Frequently asked questions

Count the conversations, not the rings. A lead is a caller you actually spoke to who turned out to want work done, so robocalls, wrong numbers, vendors, and anyone who hung up before a person picked up all sit outside the count. If you're pulling the number off a call-log total, you're counting traffic, and it'll make your booking rate look far worse than it really is.

It's roughly average, and average here is less useful than it sounds. Leads per booked job is really just your close rate inverted, so on Invoca's 2026 home services benchmarks, where 45% of leads convert on the call, it lands a little over two, and it shifts by trade with the close rate: about 1.8 for plumbing and about 2.3 for HVAC. That means it tells you almost nothing your close rate wasn't already telling you. What's worth watching is how many inbound calls it takes to book one job, because that one moves far more from trade to trade.

Partly, and it's the main thing the benchmark leaves out. That 45% measures leads converting during the call itself, so an estimate accepted on Thursday or a homeowner who texts back after the weekend never shows up in it. Treat the on-call figure as the ceiling, not the floor: about 1.8 leads per job for plumbing and 2.3 for HVAC is what it takes when only the phone counts, and genuine follow-up brings it down.

One month and three numbers will do it. Pull the total inbound call count off your phone bill or carrier log, every call whether anyone picked up or not. Note how many of those reached a person. Then count the jobs you booked while the caller was still on the phone. Total calls divided by those jobs is your calls-per-job number, the one to hold against this post's worked examples: about 8 for plumbing and about 15 for HVAC. Answered calls divided by total calls is your answer rate, and that split tells you whether the problem sits before the conversation or inside it. No software involved.