The one number every contractor should know about their phone

Most owners I talk to can tell me their average ticket without thinking. A lot of them know their cost per lead from the ads. Almost none of them know how many of the people who rang their business last month ended up as a booked job. That number sits between the ad and the invoice, and nobody owns it. The marketing company stops counting when the phone rings. The CRM starts counting when a job is booked. Everything in between, the calls nobody answered, the quote requests that went quiet, the customer who said “let me check with my wife,” falls through a gap that no report covers.

01Jobs booked divided by real calls in

That’s the whole number. Take every genuine customer call your business got in a month, count how many of those turned into a booked job, and divide. I call it the calls to jobs rate. You can call it whatever you like, as long as you work it out the same way every time.

It is deliberately blunt. It doesn’t care whether the call was missed, answered badly, answered well and then never followed up, or answered by the right person on the wrong day. It only asks one thing: of the people who picked up a phone and tried to give you work, how many did you actually get?

Two things make it useful. First, every part of it is yours. You don’t need a vendor, a benchmark or an industry report to calculate it. Your phone log and your job list are enough. Second, it moves when you change something real. Hire an office person, change who picks up at lunch, start calling back the quote requests, and this number tells you whether it worked.

I am not going to give you a “good” number to aim for. I don’t have an honest source for one across trades, and a roofer selling replacements and a plumber clearing drains shouldn’t be on the same scale anyway. The number you compare against is your own, last month.

02Why it beats the numbers you already watch

Most contractors watch two numbers closely. Leads from the marketing side, jobs and revenue from the office side. Both are fine. Neither one tells you what happened at the phone.

The ad dashboard counts clicks and calls generated. It will happily tell you that you got 90 calls this month. It has no idea how many of those 90 somebody actually spoke to. A lead company can deliver exactly what it promised while half those calls ring out in a truck cab.

The job list counts what made it through. It is a record of success, which is the problem. The calls that didn’t become jobs leave no row in your invoicing system, so the job list can look healthy while a third of your demand is leaking out before it ever gets there.

The calls to jobs rate is the only number that connects the two. If leads go up and jobs stay flat, this number drops, and you know the problem is downstream of the ad. If leads stay flat and jobs go up, it rises, and you know something you changed in the office is working.

There is some research that shows how easy it is for a business to be blind to this part. When Harvard Business Review audited 2,241 US companies on how they responded to inbound inquiries, the average first response took 42 hours and 23% never responded at all. That study timed web forms, not phone calls, so it is not a measure of your answer rate. I use it for a narrower point. Those companies weren’t careless on purpose. Nobody was counting, so nobody knew.

03Working it out from last month

Give this an hour, once. After that it takes ten minutes a month.

Start with the top of the fraction: real calls in. Export last month’s inbound call log from your carrier account, your VoIP dashboard or whoever runs your tracking numbers. Then clean it. Take out spam and robocalls, suppliers, your own crew, and existing customers calling about a job already in progress. Collapse repeats, so somebody who called three times in ten minutes counts once. What’s left is the number of people who tried to give you new work. If you have never done this cleanup before, I walked through it step by step in how to count the calls your business is missing.

Now the bottom: jobs booked from those calls. Go to your job list for the same month and count the new jobs that came in by phone. If your software records the source, use it. If it doesn’t, match phone numbers from the call log against customer records. It’s tedious the first time. It’s also the only honest way to do it.

Divide the second number by the first. That’s your rate.

A few rules keep it honest. Use the same date range for both lists. Count a booked job once, even if it took three calls to land it. Don’t include jobs that came in by email or walk in, because they never touched the phone. And write down what you excluded, so next month you exclude the same things.

04Splitting it into two halves

The single number tells you whether you have a problem. Splitting it tells you where.

Every call that turns into a job passes two gates. Somebody has to answer it, and then that conversation has to end in a booking. So the rate breaks into two parts.

Answer rate is calls answered by a person (or a system that actually talks to the caller) divided by real calls in. Booking rate is jobs booked divided by calls answered. Multiply them together and you get your calls to jobs rate back.

Here is a made up week to show the arithmetic, not a benchmark. A shop gets 60 real calls. Somebody answers 45 of them, which is a 75% answer rate. Of those 45 conversations, 18 become jobs, a 40% booking rate. Overall, 18 jobs from 60 calls is 30%, and 75% times 40% is also 30%.

Now the interesting part. Say the shop changes nothing about how it handles a conversation, but it starts answering 50 calls out of 60 instead of 45. At the same 40% booking rate, that’s 20 jobs instead of 18. Two extra jobs a week, without a single new lead and without anyone getting better at selling. Put your own average ticket on those two jobs and multiply by 52, and you’ll see why I care about this more than ad spend. The longer version of that cost math is in what a missed call actually costs.

The split works the other way too. If your answer rate is already high and your overall number is still low, the phone isn’t where you’re losing work. The conversations are.

05What moves each half

The two halves leak for different reasons, so they get fixed differently.

Answer rate falls when nobody is free. That’s lunch, the morning rush, the hour everyone is on a roof, and the evenings. Owners usually assume the gap is at night. When they sort missed calls by hour, a lot of them find the pile sitting in the middle of the working day, which I wrote about in the daytime missed call nobody talks about. The fix is somebody or something that picks up when your people can’t. That might be a part time office person, a rule that the apprentice answers when he hears it, an answering service, or a system like ours.

Booking rate falls for quieter reasons. The caller wanted a price and got “someone will call you back,” then nobody did. The person who answered couldn’t see the schedule. The customer asked a question only the owner could answer, and the owner saw the message two days later. Or the customer simply went quiet after the first call, and nobody followed up.

That last one is bigger than most owners expect. A caller who said “let me think about it” hasn’t said no. They’re busy and you’re one of three names on a list. One polite follow-up a couple of days later, asking whether they still want someone out, catches a share of those. Several over a couple of weeks catches more. Most shops send none, because nobody’s job is to send them.

One warning, since it touches this number directly. Don’t push booking rate up by booking things that aren’t real. An AI or a new hire that writes jobs onto the calendar without anyone confirming them will make the number look great for a month. Then a truck drives to an address for work you don’t do. I wrote about why that matters in why an AI should never book a job without confirming it first. A booked job, for this number, means one the business actually agreed to.

06One line a month

Put this on one line in a spreadsheet and update it on the first of every month. Five columns: real calls in, calls answered, jobs booked, answer rate, booking rate.

That’s all. No dashboard, no software. After three months you’ll know your normal range. After six you’ll know which months the phone gets away from you, and you can plan around them instead of finding out from a slow invoice month later.

When you change something, write it in a sixth column with the date. Hired somebody for the office, changed the lunch rotation, started using an answering system, started following up on quotes. Then look at the line the month after. If the number didn’t move, the change didn’t do what you thought, and that’s worth knowing too.

It also gives you a straight way to judge anyone selling you something for the phone. Ask what it’ll do to your answer rate and your booking rate, then measure it yourself afterward. If a vendor can’t tell you which half they affect, they’re guessing. If you can listen back to the calls, as I described in listening back to every call your business takes, you’ll also hear why a number moved, not just that it did.

Frequently asked questions

What counts as a real call?

A person who called to give you new work. Leave out spam, suppliers, your own crew and existing customers calling about a job already underway. Count repeat calls from the same number within a short window as one. Whatever rules you pick, use the same ones every month, or the trend means nothing.

Is there an industry average I can compare against?

Not one I trust enough to give you. Trades differ too much, and so do how shops count. A drain cleaning company and a roofing company selling full replacements will never look alike. Compare yourself to your own last month and your own same month last year.

My software doesn’t record where a job came from. Can I still do this?

Yes. Match phone numbers from your call log against your customer records for the same month. It takes longer the first time. After that, start asking every new customer how they found you and write it on the job, and next month is quicker.

Which half should I fix first?

Whichever one is lower. If you are answering most calls and still not booking, work on what happens during and after the call, especially follow up. If a lot of calls go unanswered, nothing you do in the conversation helps the people who never reached one.

Does answering more calls always mean more jobs?

No. Some answered calls were never going to book. That’s why the booking rate matters. But a call nobody answered has no chance of booking, so more answered calls usually gives you more chances, as long as the booking rate holds.

The bottom line

Your phone is the narrow part between the money you spend getting people to call and the money you make when they do. One number tells you how much gets through: jobs booked divided by real calls in. Split it into answer rate and booking rate, update it once a month, and you’ll know where the work is going missing and whether anything you change fixes it.

We book appointments for roofing companies: we find homeowners looking for roof repair or replacement, answer their replies and book them into your system. If that would help your team, see how it works at rushdial.com/roofing.