A note on where these numbers come from.
They come from you. This post is a model, not a study — we have not run a survey of restaurant phone records and we are not going to quote one we did not run. Every figure below is an assumption, labelled as an assumption, and the calculator further down lets you replace all of them with your own numbers. The point of the exercise is the shape of the arithmetic, not our defaults.
Start with one missed call.
The first input is your average phone-order ticket. You already know this one exactly — it is sitting in your POS. If you want a placeholder to read along with, we'll use $32, but use yours; a pizzeria and a steakhouse are not the same business.
The second input is the share of missed calls that turn into a lost order. Not all of them do — some callers try again, some leave a voicemail, some go straight to the next result on the page. We'll assume 55%. That is a judgement call, not a measurement, and it is the single number in this model most worth arguing about. If you think your regulars always call back, drop it.
On those two assumptions, a single missed call is worth $32 × 0.55 ≈ $17.60 in immediate lost revenue. That excludes the lifetime value of a customer who never tried you — more on that below.
Multiply by the shift.
The third input is how many calls you actually miss in a dinner shift. This is the one owners consistently guess wrong, in both directions, and it is the one you can measure this week: pull the call log off your carrier or your handset and count the inbound calls between 5 and 9pm that nobody picked up.
Take 8 missed calls a shift as a worked example. At $17.60 of expected value each, that is $140.80 per shift — about $845 a week across six dinner shifts, or roughly $44,000 a year.
Set against R.ai's $1,750/mo flat fee — $21,000 a year — that example clears the cost with room left over. Change any input and the conclusion can flip: at two missed calls a shift it does not clear, and you should not buy it. Run your own numbers before you believe ours.
Plug your own numbers in.
Drag the sliders. Use your honest peak-hour missed-call count, your actual ticket size, and the number of locations you run.
The hidden line items nobody puts in the spreadsheet.
There are two costs the simple math leaves out:
1. Customer lifetime value.A first-time caller you never reach is not one lost order — it is every order that customer would have placed afterwards. We don't publish a repeat-order multiple, because ours would be made up: the honest version is that you can compute it from your own POS in about ten minutes, and whatever number you get, the per-miss cost above is an undercount.
2. Reputation drag.A caller who couldn't get through is a plausible source of a “tried to order, no one answered” review, and is not recommending you to anyone. We don't put a dollar figure on it, and we'd be suspicious of anyone who did.
What to do about it.
You have four options to stop missing calls:
- Hire someone to answer the phone.Phone-shift labor at $20/hr × 30 hrs/week ≈ $2,600/mo per location. Three to four times the cost of R.ai. And they can't take the order while serving a table.
- Use a human answering service.$400–$900/mo per location. Can take messages. Can't take orders or run cards.
- Install an IVR. $50–$150/mo. Callers hate them. (Detailed comparison in this post.)
- Use an AI phone agent. R.ai, basically. Answers every line at once with no hold queue, takes the order, and runs the card on the call. $1,750/mo per location.
The math on options 1, 2, and 3 doesn't close. Option 4 closes by ~2 recovered orders a day. That's why we built this.