The Friday-night phone problem: what a missed call really costs a takeaway

I've stood in enough takeaways at seven o'clock on a Friday to know what the phone sounds like when nobody can get to it. It rings, it stops, it rings again. Somebody shouts "get that" and nobody does — the fryer is up, the driver is at the door, there's a queue at the counter. Then it goes quiet. Nobody writes down that a customer just gave up.

That last part is the problem. A missed call leaves no trace. There's no line on the till, no receipt in the bin, nothing to count at the end of the night. So it gets treated as free. It isn't.

This is the note I wish someone had handed me before I started building NeverMissed: a plain way to put a number on the phone you're not answering. No survey results, no industry averages I can't stand behind. Just arithmetic you can do on the back of a menu, with your own numbers in it.

Why the phone loses

On a quiet Tuesday the phone is fine. The trouble is that the phone is busiest exactly when you are. Orders cluster into two or three hours, and the same two or three people have to cook them, bag them, take payment and answer the door. The phone is the only one of those jobs that can be ignored without anyone in the shop noticing, so it's the one that gets ignored.

The customer doesn't see any of that. They hear ringing, and then they do one of three things: ring back later, walk in, or open a delivery app and order from whoever picks up. In my experience the third option is the one that costs you, and it's the one that's getting easier every year. The next shop is one thumb-swipe away, and they don't need a reason to prefer you when you're not answering.

A missed call leaves no trace. There's no line on the till, no receipt in the bin, nothing to count at the end of the night. So it gets treated as free. It isn't.

The arithmetic

Here's the whole model. Four numbers multiplied together:

Missed calls per week × share that were genuine orders × share that go elsewhere instead of ringing back × your average order value.

That's your lost revenue per week. Everything else is detail. Let me walk through each number, because the point is to use yours, not mine.

Missed calls per week. You probably don't know this one, and that's fine, because you can find out. Most phone systems and mobiles keep a missed-call log, and a landline provider will usually show it in the account. Count Friday and Saturday for a fortnight and you'll have a real number. If you can't get at a log, tally them for one busy night with a pen on the wall. Owners I've spoken to usually guess low. The NeverMissed page uses "ten or more on a busy Friday" as an illustration, and I stand by that as a reasonable order of magnitude for an independent shop, but it is an illustration. Yours could be four; it could be twenty-five.

Share that were genuine orders. Not every call is a customer trying to spend money. Some are "are you open", some are "where's my food", some are suppliers. Rather than guess, take the same fortnight and note what the answered calls were. If seven in ten of the calls you do pick up are orders, it's reasonable to assume the missed ones are similar. Absent your own data, I'd use somewhere between 50 and 80 per cent, and I'd call that an estimate, not a fact.

Share that go elsewhere. This is the softest number and the one people argue about. Some customers do ring back. Some come in. But a meaningful share have a second option loaded on their phone and use it. I'm not going to give you a percentage dressed up as research. I'll say that in the shops I've watched it is clearly not a small minority, and I'd model a range from a third to two thirds and look at both ends.

Average order value. The one number you already have. Take last month's phone or collection sales and divide by the number of orders. For a lot of independent takeaways it sits somewhere in the teens or twenties in pounds. Use your own.

A worked example, with the assumptions showing

Say a shop misses 12 calls across Friday and Saturday, 70 per cent of them orders (about 8), half of those going elsewhere (4), at an average order of £22.

That's roughly £90 a week walking out of the door, or getting on for £5,000 a year. Run the same shop at the gentle end of every range (8 missed, half of them orders, a third lost, £18 orders) and you get about £24 a week. Run it at the harsh end (20 missed, 80 per cent orders, two thirds lost, £26 orders) and it's over £270 a week.

Every figure is an estimate built from the assumptions above, shown so you can argue with them. Even the gentle version is not nothing, and it recurs every week the phone stays unanswered.

What the model leaves out

Three things, and they all push the true cost up rather than down.

First, the repeat order. A customer who orders from you every fortnight isn't worth one order; they're worth a year of orders. If the night you didn't pick up is the night they find the shop round the corner is decent too, some of them don't come back. I won't invent a figure for how many, but the direction is obvious.

Second, the answered calls that go badly. When the phone is finally grabbed mid-service, the order gets taken on a scrap of paper and the wrong pizza goes out. That's a refund, a remake and a customer less likely to ring next time — the same problem wearing a different coat.

Third, your own time. Every interruption to answer the phone is time not spent cooking. If you're the owner and the cook, the phone is competing with the thing that actually makes the money.

What it costs to fix it the old way

The traditional answer is a person. Someone on the phone and the counter for the two or three peak hours on Friday and Saturday.

At current UK wage rates that's somewhere north of £12 an hour before employer costs, and you can't realistically book someone for three hours — you're looking at a four or five-hour shift. Two nights a week, that's roughly £100 to £130 a week, or £400 to £550 a month, before you've even found someone reliable who'll turn up at 5pm every Friday.

What I built instead

NeverMissed is my answer to that gap. When your line is busy or unanswered, the call diverts to an AI receptionist that knows your menu and prices, takes the full order, reads it back, and prints it in your kitchen like any other ticket the second the call ends — every hour you're open, £499 a month after a free first week (3-month minimum), and I set it up in your shop myself, machine included. You can read exactly how it takes an order, and where it hands to a human, in the next note.

But you don't need NeverMissed to get value from this article. Count your missed calls this weekend and put your own numbers in the formula. If the answer is a couple of pounds, forget it. If it's a couple of hundred, you now know what the ringing costs, and you can decide what to do with your eyes open.