Why Multi-Location Brands Never Find Out How Many Calls They Miss

Open your Google Business Profile dashboard right now. Pick one location. Look at the call button.

That number tells you how many people tapped it. It does not tell you how many of them reached a human.

Almost every multi-location brand we audit can produce the first number in seconds. Almost none can produce the second. The gap between those two numbers is where a serious amount of revenue quietly goes.

What is call answer rate?

Call answer rate is the percentage of inbound calls to a location that reach a human, measured against the calls that location actually generated. For a multi-location brand it is calculated per location and per hour, not as a network average, because an average hides both the locations losing calls and the hours in which they lose them.

Two things separate it from the call metrics most brands already have.

First, it is a ratio rather than a count. Call volume tells you how well your listings are working. Answer rate tells you how well your operation is working.

Second, it only means anything at the smallest unit. A network sitting at 78 per cent looks acceptable until you find eleven locations at 40 and everyone else at 95.

The number nobody has

Most brands measure demand with real discipline. Occupancy, footfall, enquiry volume, cost per lead. All tracked, all reviewed monthly.

Then a customer taps the call button on a listing, and the measuring stops.

Nobody at head office knows whether that call rang out. Nobody knows whether it was answered in four seconds or forty. Nobody knows whether the person who picked up could help, or took a message that went nowhere.

The listing did its job. It produced the call.

What happened next is invisible. And because it is invisible, it is nobody’s KPI.

Why the published benchmarks are worth ignoring

Search this topic and you will find confident numbers everywhere. Sixty-two per cent of business calls go unanswered. Missed calls cost the average business a specific figure per year.

We are not repeating them. We could not trace a single one to a primary source.

They appear on vendor blog after vendor blog, each citing the last, with no study anywhere at the bottom of the chain.

So the honest position is this. The published benchmarks in this category are not reliable and you should not size a budget on them.

Your own data is completely reliable, and it takes two weeks to collect.

Where your calls actually come from

Three sources, and they behave very differently.

• The listing call button. The highest intent of the three. Someone searched, saw your location, and chose to call rather than browse. In most networks this is the biggest single source of inbound calls and the least tracked.
• The website and microsites. Usually better instrumented, because marketing owns it.
• Directories, aggregators and paid ads. Varies a lot. Often already tracked, because somebody is paying per click.

The first one is the problem. It bypasses every system marketing controls, rings a phone at the location, and leaves no record anywhere central.

The four windows where calls go missing

In the audits we run, missed calls cluster into four patterns. They need four different fixes, which is why a single blended “missed call rate” is close to useless on its own.

1. After hours

The obvious one, and usually the largest.

For a hotel this is the guest in another timezone booking at what is midnight for you. For a diagnostics network it is somebody arranging a test before work. For a dealer network it is a contractor calling at nine at night, because that is when he stopped working.

These calls never come up in a staffing conversation. Nobody is there to notice them.

2. Peak hours

Less obvious, and often bigger than people expect.

The line is busy exactly when demand is highest. The front desk is checking someone in. Reception is working through a queue. The counter is serving a customer standing right there.

A physical customer always beats a ringing phone, and rightly so.

Which means you miss the most calls in the hour you generate the most.

3. Shift boundaries

The fifteen minutes around a handover, when nobody owns the phone. The outgoing shift is closing up. The incoming one is not at the desk yet.

This shows up as a sharp dip at the same time every single day, which makes it the easiest of the four to spot once you are actually looking.

4. Answered, but never handled

The most expensive category, and it appears in no call report anywhere.

Someone picked up. They took a name and a number. The enquiry went onto a pad, or was promised to a colleague, and stopped there.

On paper, a successfully answered call. In reality, a lost customer who believes they made contact with you.

How to measure it, in five steps

You do not need a platform to start. You need two weeks and one person willing to be systematic.

1. Pull call volume from your listings. Every Google Business Profile reports call button taps. Export it per location, per day, for ninety days. For a large network, pull it through the Business Profile API rather than location by location. This is your demand baseline, and it is free.
2. Pull answer data from your telephony. Whatever runs your phones can usually export inbound calls with a timestamp, a duration and an outcome. Ask for total inbound, answered, unanswered, and average time to answer. Per location, per hour.
3. Line the two up. Calls generated against calls answered, by hour, by location. Do it in a spreadsheet the first time. The pattern will jump out.
4. Sample the answered ones. Take fifty at random and trace each. Was an enquiry created anywhere? Did anyone ring back? This is the only way to find the fourth category, and yes, it is manual the first time.
5. Rank locations by answer rate, not volume. Your bottom ten per cent is where the problem concentrates, and it is rarely the locations anyone predicts.

What to do about each window

After hours is the one place automation is clearly the right first answer. There is no staffing fix at three in the morning that makes commercial sense. This is where an AI agent, or at minimum an automatic callback, earns its place.

Peak hours is an overflow problem, not an after-hours one. The call needs somewhere to go when the main line is engaged. A second line, a central desk, or a queue with an honest wait time.

Shift boundaries is pure process. Once you can see the dip, someone owns the phone through the handover. This one usually costs nothing at all.

Answered but never handled is the hardest, because it is a discipline problem wearing a technology costume. The fix is that every enquiry lands somewhere the location cannot quietly close, and somebody at the centre sees the ones going stale.

Human first, AI second

There is a loud assumption in this market that AI should answer the phone.

We would argue the opposite for anything premium. And we are saying that as the people selling the AI.

The technology director at a luxury hotel group put it to us plainly this year. He did not want AI picking up at all. He wanted every call routed to a human receptionist first, with AI taking over only when nobody responds, after hours, or when an overseas guest calls at an hour the desk is genuinely unstaffed.

He was right, and it generalises well past luxury hotels.

A guest, a patient or a large-ticket buyer who reaches a machine on the first ring learns something about how much you value them. It is not the thing you wanted them to learn.

One more rule follows. When the AI does pick up, it should say it is an AI in the first sentence. Every operator we have discussed this with reached that conclusion on their own, usually right after imagining the version where the customer works it out halfway through.

What good actually looks like

A network with this under control can answer four questions on any given morning.

• How many calls did each location generate yesterday?
• How many were answered, and how fast?
• Which hours are we losing calls, and is that staffing or process?
• Of the calls we answered, how many became a recorded enquiry?

None of those is an exotic question. Most large networks cannot answer a single one of them today, which is rather the point.

Start here

Take your ten highest-volume locations. Pull ninety days of listing call data and ninety days of telephony data. Line them up by hour.

You will almost certainly find that one of the four windows accounts for most of the loss, and that it sits in a handful of locations rather than spread evenly across the network.

That is two weeks of work, and it will tell you more about how your network actually runs than anything else you buy this year.

Frequently asked questions

What is a good call answer rate for a multi-location business?

There is no benchmark worth trusting. The figures circulating online cannot be traced to a primary source. Measure your own network, then set an internal target by comparing your best locations against your worst. The spread inside your own network is far more useful than any industry average.

How do I find out how many calls my Google Business Profile generates?

Every profile reports call button taps in its performance data. Export it per location, per day. For a network above roughly fifty locations, pull it through the Business Profile API instead of going location by location.

Will call tracking numbers hurt my local rankings?

Not if you configure them correctly. Keep the location’s real number as the primary and add the tracking number as an additional number. The damage happens when the primary is replaced and stops matching your citations elsewhere on the web.

Should AI answer calls at my locations?

Not first. Human, then overflow, then AI. And when AI does answer, it should say so in the opening sentence.

How long does this take to measure?

About two weeks for ten locations, most of which is waiting on your telephony provider to produce an export. The analysis itself is an afternoon in a spreadsheet.

So: do you know your answer rate by hour? Or only your call volume?