The 6pm Problem: Why Small Businesses Miss Calls When They Matter Most

The calls you lose are not random

Ask any owner when they miss calls and they will say "when we're busy". That is right, but it understates the problem. Missed calls are not scattered evenly across the trading day. They pile up inside a narrow window, and it is the window where each call is worth the most.

For a restaurant that window is roughly 5pm to 8pm. For a clinic it is the first hour after opening. For a salon it is Saturday morning. The pattern is the same everywhere: the hour customers most want to reach you is the hour you have the fewest free hands.

Why hiring does not fix it

The instinct is to put another person on the phone. It rarely works, for a simple reason.

The conflict is not between the phone and idle staff. It is between the phone and a customer standing in front of you. When a guest is waiting at the counter and line two rings, somebody has to lose. The person in front wins, every time, in every business, and they should. That decision gets made a dozen times a shift and it is the correct one each time.

Adding a person moves the line. It does not remove it. Your new hire is also standing in front of a customer at 6:40pm on Friday.

The arithmetic nobody runs

Most owners never count the cost because the loss is invisible. Nothing appears in the till. No complaint gets made. The customer simply calls somebody else and you never learn their name.

Try the count for one week. Ask whoever is on the floor to make a tally mark each time the phone rings out. Most owners are startled by the number, and startled again when they multiply it by their average order value.

Christian Wiens, founder of the voice AI company Loman, published an estimate in QSR Magazine in 2026 putting the annual loss across US restaurants at around twenty billion dollars, built from an assumed 150 missed calls per restaurant per month and a thirty-eight dollar average takeaway order. Treat that figure as a vendor's estimate rather than independent research, because it is one. The useful part is not the headline. It is the method. Run the same arithmetic on your own tally and your own average order, and the number you get is real in a way the industry figure never is.

What actually helps

Three things move the needle, in order of effort.

Shorten every call. If the average call drops from three minutes to ninety seconds, you handle twice as many with the same staff. This is the highest-return change available to a small business and almost nobody works on it deliberately.

Most call time is not order taking. It is identification. What is your name. Can I get your number. What was the address again. Sorry, can you repeat the street. For a returning customer, every second of that is a second spent recovering information you already had.

Know who is calling before you pick up. A returning customer whose name, address and last order appear on screen when the phone rings does not need to be identified at all. The call starts at the order instead of the introduction. This is the whole reason caller ID software exists for small businesses.

Make the callback cheap. You will still miss calls. What matters is whether a missed call is recoverable. If you have the number and know who it belongs to, a two-minute callback after the rush often saves the order. If the number is a stranger, it usually does not.

The honest limit

None of this makes the phone stop ringing at your worst moment. Nothing does. A busy Friday at 7pm will always be a busy Friday at 7pm.

What changes is how much each interruption costs you. A call that takes ninety seconds instead of three minutes, from somebody you already recognise, is an interruption you can absorb. That is the whole gain, and it is enough.

Start by counting for one week. You cannot fix a leak you have not measured.