Key takeaways
- Count unique callers, not rings: one guest who tries three times is one missed guest, not three.
- Split unanswered calls into rang out, hung up early and went to voicemail, because each has a different fix.
- A seven-day hour-by-day grid shows whether the problem is peak service, the lunch handover or after close.
- Track how fast missed callers get a callback; it is the number that tells you whether the leak is closing.
- Guests whose first booking attempt fails often try another route, but some simply do not book[1].
Why missed calls are worth counting
Very few restaurants know how many calls they miss. The phone rings during service, someone eventually answers or it stops, and the night moves on. Yet the telephone has long been one of the most common ways diners make a reservation[2], and a missed call is often a guest deciding where to eat that night.
Cornell researchers asked diners what they would do if they could not book through their preferred channel. Most said they would try another route, but nearly one in five said they would not book at all[1]. You cannot see those guests in your covers report. You can only see them in your call logs.
Source: Cornell University Center for Hospitality Research (Kimes and Kies), 2012[1]
Show the data
| Item | Value |
|---|---|
| Would still try another channel | 58% |
| Would not book | 18% |
Define what you are counting
Before pulling any data, agree on definitions. Different phone systems label calls in different ways, and a missed-call rate means nothing if one week counts voicemails and the next does not.
| Outcome | What happened | Typical fix |
|---|---|---|
| Answered | A person or system picked up and the caller stayed on the line | None; use as the baseline |
| Rang out | The phone rang until the system stopped trying or the caller gave up | More coverage at that hour, or a backup ring group |
| Abandoned early | Caller hung up within a few rings or while on hold | Faster pickup, shorter holds, ask before holding |
| Voicemail | Caller reached the voicemail greeting, with or without leaving a message | Better greeting, faster callback, links by text |
| Busy or overflow | All lines were in use so the caller got a busy signal or overflow message | Extra simultaneous lines or overflow answering |
| Repeat attempt | The same number called again within a short window | Merge into one guest when counting |
For the headline number, use unique callers. If the same mobile number rings three times in ten minutes and gets through on the third try, record one guest who waited too long, not two missed calls and one answered call.
Pull one week of logs
Find the report
Log in to your carrier or VoIP portal and look for call detail records, call history or a usage report. POS-linked phone systems often keep a call log next to the order screen. If you only have a basic landline, ask the carrier for an itemized record of incoming calls.
Choose a normal week
Avoid holiday weeks and closures for the first measurement. Seven consecutive days gives you each weekday once, including the weekend peak.
Export the right fields
You need the date and time of each call, the caller's number, the call duration, and the outcome or disposition. Wait time before answer and whether a voicemail was left are useful if available.
Remove noise
Delete internal calls between your own extensions, calls from your own staff and supplier numbers you recognize, and obvious spam or robocalls.
Merge repeat attempts
Sort by caller number and time. Treat calls from the same number within fifteen minutes as one attempt, and record the final outcome.
Build the grid
In a spreadsheet, put days across the top and hours down the side. In each cell, count unique callers and how many were not answered.
The three numbers to calculate
- Missed-caller rate: unanswered unique callers divided by all unique callers, for the whole week and for each hour.
- Peak missed hour: the single hour on the single day with the most unanswered unique callers. This is where your first change should go.
- Callback speed: for each missed caller, how long until someone from the restaurant called back. Report the share called back within thirty minutes and the share never called back.
Callback speed matters because guests cannot always reach you when they need to. In a Cornell survey of frequent diners, respondents often found it hard to contact a restaurant to change a reservation, even though they believed they should call when plans changed[3]. Cornell's advice to restaurants includes making it easy for customers to reach you when they are running late or cannot make it[4]. A missed call from a guest trying to cancel is a table you could have resold.
Example: one week at a neighborhood bistro
Example: a fictional 70-seat bistro exports seven days of call records. After removing staff and supplier calls and merging repeat attempts, it has 380 unique callers. Of those, 57 were never answered by a person: 31 rang out, 14 hung up within four rings and 12 reached voicemail. That is an example missed-caller rate of 15% for the week.
The example grid shows where they cluster. Friday and Saturday between 6 and 8 pm account for 22 of the 57. Another 15 arrive between 2 and 4 pm on weekdays, when the lunch host has left and the dinner host has not started. The rest are spread across late evening, mostly after the kitchen closes.
In this example only 19 of the 57 missed callers received a callback, and just 8 of those within thirty minutes. The manager now has three specific problems to work on instead of a vague sense that the phone is busy: weekend peak coverage, the afternoon handover gap and a callback routine that does not exist yet.
Read the pattern and pick one change
Different patterns point to different fixes. Resist the urge to solve everything at once; change one thing, then measure the same way for another week.
| If missed calls cluster… | Likely cause | First change to try |
|---|---|---|
| Friday and Saturday evening | Host stand overloaded with walk-ins and wait list | Name a second person who answers when the host is busy, or route overflow |
| Mid-afternoon on weekdays | Shift handover gap | Overlap the lunch and dinner host by thirty minutes |
| After the kitchen closes | No one assigned to the phone | Rewrite the voicemail and text a booking link |
| Before opening | Calls arrive while prep is underway | Answer from the office or forward to a manager's line |
| Evenly through the day | Too few lines, or the phone is in a noisy spot | Add a simultaneous line or move the handset |
After-hours patterns deserve special attention. In earlier Cornell work, an OpenTable estimate put a quarter of online reservations between 10 pm and 10 am[5], when most restaurants are closed or prepping. If your late calls are mostly booking questions, a better voicemail and a texted booking link may recover more guests than extra staff ever could.
Mistakes that distort the number
A missed-call count is only useful if it is measured the same way each time. These are the errors that most often make a restaurant look better or worse than it really is.
- Counting rings instead of people. A single anxious guest who redials four times can make one bad minute look like a disaster.
- Leaving spam in the data. Robocalls and listing scams are often short and unanswered, so they inflate the missed rate at random hours.
- Ignoring calls that were answered too late. A caller who waited through eight rings was answered, but they were close to hanging up. If your system logs time to answer, track the share of calls answered after the fourth ring as a warning sign.
- Measuring a holiday week first. Mother's Day or a local festival week will show a pattern you cannot staff for the rest of the year.
- Counting voicemails as handled. A voicemail is only handled once someone calls back, so keep it in the missed column until the callback happens.
Turn the missed list into callbacks
Measuring is the first half of the job. The second half is a routine that turns every missed number into a returned call. Print the list of missed numbers from the phone system, or keep a paper sheet at the host stand, and assign one person per shift to work it.
- Check the list at fixed times: just before service, after the first seating and at close.
- Call back in the order calls were missed, oldest first, skipping numbers that already got through on a later attempt.
- Open with the reason: “Hi, this is Copper Pot Bistro returning a call from this number. How can I help?”
- If no one answers, leave a short message with the best time to reach you and a way to book or order online.
- Mark each number as reached, message left or not needed, so the next shift does not call the same guest twice.
Measure the callback routine with the same weekly pull. If the share of missed callers who never hear back is falling, the routine is working even before staffing changes.
Make it a monthly habit
Once you have a baseline, repeat the same seven-day pull every month, using the same definitions and the same cleaning rules. Keep a one-line log: date range, unique callers, missed-caller rate, peak missed hour and callback speed. Over a few months the log shows whether staffing changes, a new phone menu or a rewritten greeting actually moved the number.
Share the numbers with the hosts. People who answer the phone usually know which hours feel impossible, and a grid that confirms it gives them a reason to suggest fixes. For next steps, see our guides to tracking call volume by hour and handling the dinner rush.
Frequently asked questions
What counts as a missed call at a restaurant?
Any caller who did not reach a person or a system that handled their request: calls that rang out, hung up early, hit a busy signal or went to voicemail. Count unique callers, not individual rings.
How do I see missed calls on my restaurant phone?
Check your carrier or VoIP portal for call history or call detail records, or the call log in a POS-linked phone system. Export at least date, time, caller number, duration and outcome.
What is a good missed-call rate for a restaurant?
There is no reliable public benchmark. Measure your own baseline for a normal week, fix the worst hour first and track whether the rate and your callback speed improve month by month.
References
Cornell University Center for Hospitality Research (Kimes and Kies) (2012). The Role of Multi-Restaurant Reservation Sites in Restaurant Distribution Management (Cornell Hospitality Report, 2012) (opens in a new tab) ↑
Cornell University Center for Hospitality Research (Kimes and Kies) (2012). The Role of Multi-Restaurant Reservation Sites in Restaurant Distribution Management (Cornell Hospitality Report, 2012) (opens in a new tab) ↑
Cornell University Center for Hospitality Research (Sheryl E. Kimes) (2008). A Consumer's View of Restaurant Reservations Policies (Cornell Hospitality Report, 2008) (opens in a new tab) ↑
Cornell University (Sheryl E. Kimes) (2011). Customer Attitudes towards Restaurant Reservations Policies (Kimes, Journal of Revenue and Pricing Management, 2011) (opens in a new tab) ↑
Cornell University Center for Hospitality Research (Sheryl E. Kimes) (2009). How Restaurant Customers View Online Reservations (Cornell Hospitality Report Vol. 9, No. 5) (opens in a new tab) ↑