Rent is fixed and the cost price moves slowly — wages, though, are the item you set yourself week by week, through the rota. And that is exactly what nobody works out in most shops; it gets written on the basis of “that's how it usually is”.
The question isn't whether to give fewer hours. That saving works for a while, then a good person leaves, and that costs far more. The question is which part of the day those same hours should fall on.
Measure wages as a share of turnover
In money, wage costs always rise, and you learn nothing from that. The usable figure is what percentage of monthly turnover the gross wages and contributions together make up.
| What you look at | How you work it out | What it tells you |
|---|---|---|
| Wage share | (wages + contributions) ÷ monthly turnover | Makes months comparable |
| Turnover per hour worked | monthly turnover ÷ hours worked | Shows how well the rota lands |
| Margin per hour worked | monthly margin ÷ hours worked | The better measure — this pays the wages |
Measure when customers actually come
Almost every shopkeeper can tell you when the shop is busy. And almost every shopkeeper is surprised the first time they look at the actual numbers: the peak is usually shorter than it feels, and often somewhere else.
- 11. Note transaction counts for two weeks, not takingsThe count — how many separate purchases there were — shows the staffing need better than the money. One big shop is one customer; five small purchases are five services.
- 22. Split the day into no more than three stretchesMorning, late morning to early afternoon, late afternoon to close. A finer split is noisy in a small shop, a coarser one says nothing.
- 33. Compare like daysTuesday with Tuesday, Saturday with Saturday. The difference between days of the week is so big that a mixed average washes out every pattern.
- 44. Find the two weakest hoursDon't look for the peak first, look for the dead time. That is where the surplus hours are — and where they can be trimmed with the least pain.
- 55. Look at what happens on delivery dayChecking in and shelving goods are working hours too, just not at the counter. If that isn't in the rota, somebody will be missing from the peak instead.
Four ground rules for the rota
- 1Plan around the peak, not around the opening hours. Opening hours are the frame; the rota fits the trade inside it.
- 2Have an overlap at the handover. Fifteen minutes for the cash and the open matters to be passed on — without it every difference ends up ownerless.
- 3Put work that isn't at the counter into the dead time: shelving, checking dates, cleaning. That way the paid hour doesn't stand idle.
- 4Publish it two weeks ahead. A predictable rota is the cheapest staff benefit there is — and it cuts turnover of staff better than a pay rise.
When are two people cheaper than one?
The second person is worth it when they bring more than they cost — and that isn't a theoretical question, it can be worked out. Look at how much margin arises in the stretch you are wondering about. If it is consistently more than the second person's hourly pay and contributions, then the second person isn't a cost but an investment.
| Situation | One person | Two people |
|---|---|---|
| A queue at the counter in the peak | A customer leaves — visible nowhere | The second person's margin covers their pay |
| Dead time in the late morning | Enough | A surplus hour |
| Delivery day | Either the counter or the stockroom loses out | The goods go in, the counter works |
| Opening and closing alone | Risky and tiring | Safer, but only where there is trade |
The customer who leaves is a dangerous item precisely because it never appears in any report. You see the wage hour you paid; you don't see the sale that didn't happen — which is why trimming the rota always LOOKS successful, even when it isn't.
What not to do
- Don't publish the rota at the last minute. Unpredictability is the commonest reason good people leave.
- Don't plan a shift with no break. It isn't only a legal question: a mistake made in the eighth hour costs more than the break.
- Don't put the hardest shift on the weakest person. Saturday morning isn't training time.
- Don't solve a structural shortage with overtime. Overtime costs more than one well-placed extra person.
- Don't look at the wage hour alone: if a customer leaves meanwhile, the saving turns into lost turnover.
The biggest improvement in wage costs doesn't come from trimming the pay rate but from having the same hours fall on a better part of the day. That takes knowing when customers come — and this is the point where instinct is wrong most often.
Two weeks of hourly transaction counts, the day split into three stretches, like days compared: that is enough to see the pattern. From then on the rota is built not from habit but from what the shop actually does.
Frequently asked questions
What share of turnover should wages be?
There is no single correct figure: between a bakery, a tobacconist and a greengrocer the difference is an order of magnitude. Your own share becomes useful because you measure it the same way month after month — if it suddenly jumps, there is something to look at. Comparing with an average found online is misleading.
How do I find out when the peak hours are?
For two weeks note the transaction count — how many separate purchases there were — by hour or split into three stretches of the day, and compare like days. Takings mislead, because one big shop is also a single service.
When is it worth putting a second person on?
When the margin arising in that stretch is consistently more than the second person's hourly pay and contributions. Don't leave the customer who walks out of the calculation: that never appears in a report, which is why cutting staff always looks successful.
How far ahead should I publish the rota?
At least two weeks. A predictable rota is the cheapest benefit you can give, and it often cuts staff turnover better than a pay rise — while training a new person costs far more than what you gain from flexibility.
Is an overlap needed at the shift change?
Yes, around a quarter of an hour. That is what it takes to hand over the till and the open matters. Without it a till difference can't be traced to a shift, and the conversation becomes a question of trust instead of process.
What should staff do in the dead time?
Work that isn't at the counter: shelving, checking dates, cleaning, restocking. These take working hours anyway — if they fall in the dead time, then nobody is missing from the peak because of them.
Is it better to cover a shortage with overtime?
Simpler in the short term, dearer over time — and tiredness brings mistakes you pay for at the till and on the stock. If overtime comes back every month, that isn't a peak but a structural shortage of staff.
How do I measure the rota without till reports?
Daily and per-stretch transaction counts can be noted by hand for two weeks, and even that shows the pattern. If you close the day in shifts — each with an opening float and its own close — that split comes about by itself and stays comparable months later.
Let the rota follow when people actually come in.
Boltom App works in shifts: every shift starts with an opening float and ends with its own close, and the daily turnover shows when there really are customers.
- Shift close with opening float and difference, by employee — not discovered the next morning.
- Daily turnover by day: you can see which day and which stretch brings the money.
- With several shops, separate shifts and a separate report per shop.
No card needed · 2 minutes
