How to count the calls your business is missing

Ask ten contractors how many calls they missed last week and nine of them guess. The tenth says none. Both answers are wrong for the same reason: a call you never got to is a call you have no memory of. You remember the two voicemails because they made noise on your phone. The guy who rang out at 2:40 while you were under a house, waited four rings, hung up and called the next name on the list, left nothing behind. He is not in your head and he is not in your CRM. He is in one place, and you already pay for it.

01Your guess is low, and it is not your fault

I have never met an owner whose guess came in above the real number. Not once. The guess is always the same shape too: three or four, mostly after hours, mostly not serious.

The reason is structural. Missed calls are the only part of your operation that leaves no trace on you. A missed delivery gets a phone call. A missed inspection gets a letter. A missed call gets silence, and silence does not enter your memory as an event.

There is research that shows how wide this gap gets. Harvard Business Review timed 2,241 US companies responding to inbound web inquiries and found that 23% never responded at all. Not slowly. Never. Every single company in that group would have told you beforehand that of course they get back to everyone. That study measured web forms rather than phone calls, so I am not going to pretend it is your answer rate. What it is good for is the point about self-assessment. When somebody actually measured, a quarter of these businesses were doing something none of them believed they were doing.

So do not argue with your guess. Just go get the real number. It takes about an hour and you never have to wonder again.

02The log already exists and you are already paying for it

Every inbound call to your business is recorded somewhere by whoever carries it. Not the audio, the record: what number called, at what time, how long it lasted, and whether anyone picked up. Carriers call these call detail records. You do not need permission or a vendor to see yours.

Where it lives depends on what you run the phone through.

If your business number is a cell phone, log into the carrier’s website rather than looking at the phone itself. The handset shows you a short recent list that gets overwritten. The account shows you the full record, usually under usage or call details, and you can pick a date range and download it.

If you use a VoIP service (RingCentral, Grasshopper, OpenPhone, Dialpad, Google Voice, anything where the number lives in an app), the call log is in the admin dashboard and it will export to a spreadsheet. This is the easiest version. It usually has a disposition column that already tells you answered, missed, or voicemail.

If you have a desk phone system in an office, whoever installed it can pull the report, and if you have a receptionist she can probably do it herself in a few clicks.

If a marketing company runs tracking numbers on your ads, they have their own log for those numbers, and you should ask for it as a separate file. Those are your most expensive calls, because you paid for the click that made the phone ring.

One thing your call log is not: your CRM. Your CRM knows about the jobs you booked. It has no idea about the calls that never became jobs, which is the entire thing you are trying to measure. Jobs booked is an output. Calls received is the input, and almost nobody watches the input.

03Decide what counts before you start counting

Do this part first or you will spend the whole hour arguing with yourself.

I count four things as a missed call. One, the phone rang out and no voicemail was left. Two, it went to voicemail and they hung up without speaking. Three, somebody answered but the call lasted under about fifteen seconds, which is a hold abandon, a hang up, or a wrong person picking up. Four, it came in after hours and hit a recording.

Three things I do not count. Robocalls and spam, which you can spot by the pattern (same area code as you, no follow up, eight seconds long). Numbers you recognize as suppliers, sales reps, the parts counter, your own crew. And repeat calls from the same number inside a short window.

That last one matters more than it sounds. If a customer called you three times in eleven minutes, that is one person who needed you, not three missed calls. Sort your sheet by phone number and collapse the repeats before you count anything. Inflated missed call numbers almost always come from skipping this step, and I would rather hand you a small number you believe than a big one you argue with. Your own count is only useful if you trust it.

04The count itself, about forty minutes

Pull seven days. Pick a normal week, not the week you were on holiday and not the week after a storm, because you want your baseline, not your worst case. If you are in a seasonal trade, do this twice, once in season and once out.

Put the export in a spreadsheet and cut it down to inbound calls only. Most exports include your outgoing calls and they will double your row count for no reason. Then add four columns by hand: bucket (from the four above), prospect or not, called back yes or no, and became a job yes or no.

The first pass is quick. Go down the list, drop the spam and the suppliers, collapse the repeats, tag each remaining unanswered call with its bucket. Now you have a number. That number is how many people tried to reach your business last week and did not get you.

The second pass is the one that changes minds, and it takes twenty minutes. Take every unanswered number and search it in your invoicing system or your CRM. Some of them are existing customers, which is useful in a different way. The ones that are not in your system anywhere, and that you never called back, are the honest answer to what this is costing you. They called, nobody picked up, nobody chased, and they are gone.

If you want the uncomfortable version, call ten of them. Not to sell. Just ask whether they got the work done and who did it. You will hear “yeah, we got someone out that afternoon” enough times to stop needing a study.

05Turning a tally into a number that means something

A raw count of missed calls does not tell you anything on its own. Six sounds small. Six with a $900 average ticket is a different sentence.

The arithmetic is four numbers, all of which are yours. Start with the unanswered prospect calls from your week. Multiply by your close rate on the calls you do answer, because the ones you missed were not all going to buy either. Multiply that by your average job value. Multiply by fifty two.

I am deliberately not giving you sample numbers, because everybody’s are different enough that mine would just be noise. A plumbing call and a roof replacement are not on the same scale. If you want the longer version of that math, with the ways owners usually get it wrong, it is in what a missed call actually costs.

While you have the sheet open, do one more sort, and this is the part people find genuinely surprising. Sort your missed calls by hour of the day.

Owners expect the pile to sit in the evening. Often it does not. It sits between ten and two, in the middle of your working day, because that is when the phone rings most and also when every person who could answer it is on a job. If your pile really is after hours, that is a different problem with different fixes, and I wrote about that one separately in the after hours lead problem. But do not assume which one you have. Look at the column.

06Count it again in thirty days

The count is only half useful the first time. It becomes properly useful when you have two of them.

Whatever you change after seeing the number, and it might be nothing more complicated than telling your apprentice to pick up the phone when he hears it, you run the same count again in thirty days on a comparable week. Same source, same four buckets, same rules about repeats. That is the only honest way to know whether the change did anything.

Two warnings about the second count.

First, whoever you buy from will offer you their dashboard number, and their number will be bigger than yours. Not because they are lying. Because they count differently, usually including repeats and sometimes counting spam. Keep running your own count with your own rules. A vendor’s metric is a sales tool. Yours is a measurement.

Second, do not let anybody talk you into an answer rate benchmark. The only measured answer rate I have for a local service vertical comes from med spa software data, where businesses answer roughly 55% to 65% of their inbound calls. That is a different industry with a different phone pattern, so I would not hold your shop to it. It is worth knowing only because it tells you that a business answering two thirds of its calls is normal enough to show up in an industry average, and most of those owners would have guessed 90%.

Your own two numbers, thirty days apart, beat every benchmark anyone can sell you.

Frequently asked questions

How far back should I pull the log?

Seven days is enough to act on and short enough that you will actually finish it. Thirty days is better if the export is easy, because one week can be unusually quiet or unusually mad. If you can only face one week, pick a normal one and note the date so your follow up count uses a similar week.

My phone only shows a short list of recent calls. Is that all I get?

No, that is just the handset. Log into your carrier’s account on a computer and look for usage, call details, or call history. You can pick a date range there and often download it as a file. If you cannot find it, call the carrier’s business support line and ask for your inbound call detail records for the last thirty days. They are yours and it is a routine request.

Does a voicemail with a message count as a missed call?

It counts as missed for the tally, but track it separately, because it is a different problem. A message left is a lead you still have. A ring out is a lead you never knew existed. If most of your pile is voicemails you did call back the same day, your problem is smaller than you feared. If most of it is silent ring outs, it is bigger.

Isn’t this what my CRM dashboard is for?

Your CRM starts counting after somebody becomes a lead. The calls you are looking for never got that far, so they are invisible to it by design. That is why the count has to start at the carrier, upstream of every other system you own.

What if the count comes back at zero?

Then you have a genuinely good phone setup and you have spent an hour proving it, which is worth an hour. Do it again during your busiest month, since that is where the misses hide. Nobody drops calls in February.

The bottom line

Nobody can tell you what your phone is doing except your own call log, and it has been sitting in your carrier account this whole time. Pull seven days, collapse the repeats, sort by hour, and you will know more about the front door of your business than most owners in your trade know about theirs. It is one hour, once, and it costs nothing.

Then, once you have your number, call your own line from another phone and let it ring out. That is what a caller runs into when nobody picks up.

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.