Why change runs out and cash goes missing
Change shortage is not random. It follows your price points. A shop whose common sale is KES 70 paid with a hundred needs thirty-shilling combinations all day; a shop selling at 45 and 95 needs fives constantly. The float that works is shaped like your prices, which is why a round KES 5,000 of mixed notes often fails while a smaller, deliberately chosen mix does not.
Mobile money has changed the shape of the problem rather than removed it. Many shops now take most of their value through M-Pesa and most of their transactions in cash, so they need less float in total and still need the same small denominations. The two streams also have to be reconciled separately, which is where a lot of unexplained differences actually originate.
Cash goes missing for four reasons, in roughly this order of frequency: change given wrongly in a hurry, a sale not recorded, money taken out of the till for a legitimate purchase without a note, and theft. The first three are all fixed by the same thing, which is a count the same day, and the fourth is only detectable once the first three have been eliminated.
Want to see it running in your own business first?
Sign UpSizing and controlling the float
A week of observation first. Guessing the float is how shops end up with both too much money and not enough change.
- 1
Step 1: Write down every time you run short, for a week
Denomination, time of day, what you did instead. A week of this is enough to see the pattern, and the pattern is usually one or two denominations rather than a general shortage.
- 2
Step 2: Build the float from denominations, not a total
Decide how many of each coin and note you need to open, based on step one. The total is whatever that adds up to. A float specified as "KES 4,000" will drift in composition until it stops working; a float specified as a list of denominations can be checked at a glance.
- 3
Step 3: Count it at open and at close, by two people
Both counts, written down, signed or initialled by whoever counted. This is not about distrust; it establishes the boundary of each shift so that a difference belongs to a known period and a known pair of hands, which is as much protection for staff as for you.
- 4
Step 4: Bank the surplus rather than letting it accumulate
Takings above the float should leave the drawer on a schedule you keep to. A drawer with the week in it is both a cash flow problem and a temptation, and the amount at risk grows with every day it stays.
- 5
Step 5: Require a note for every non-sale withdrawal
Paying a supplier, buying lunch, giving a boda fare: all legitimate, all invisible by close of business. A slip in the drawer saying what left and why turns tomorrow’s mystery into today’s record.
- 6
Step 6: Reconcile cash and mobile money separately
Netting them off hides problems in both. Cash takings should match the recorded cash sales; M-Pesa received should match the recorded M-Pesa sales. A combined figure that balances can contain two errors cancelling each other out.
How float control fails
Treating the float as a round number
A float defined by its total drifts in composition. By Thursday it is the right amount of money in the wrong denominations, which is functionally the same as not having one.
Counting only at close
A single daily count cannot tell you whether a shortfall happened on the morning or the evening shift. Opening and closing counts cost two minutes and halve the size of every question.
Letting one person count alone
A count by one person is an assertion. Two people counting is a record, and it protects the honest staff member who would otherwise carry every unexplained difference personally.
Taking money out without a note
This is the single largest source of unexplained differences in small shops, and it is almost always the owner who does it. A drawer slip takes five seconds.
Chasing small differences and ignoring the trend
A shop that balances to the shilling most days and is short KES 300 every Friday has a Friday problem, not a cash problem. Record the daily difference so the pattern can appear; chasing each one in isolation finds nothing.
A worked example: a duka that kept running out of fifties
A duka in Ruiru held a KES 5,000 float, replenished when it looked thin. Change ran short most afternoons and the owner assumed the float was too small.
A week of notes showed something else. Of 23 recorded shortages, 19 were KES 50 notes and coins, all between 1pm and 4pm, driven by a lunchtime trade paying for KES 150 purchases with 200 notes.
The float was rebuilt as a denomination list: forty KES 50 notes, thirty KES 20 coins, forty KES 10 coins, twenty KES 5 coins, plus KES 1,000 in larger notes. The total came to KES 4,100, around a fifth less cash in the drawer than before, and the afternoon shortages stopped.
The second finding was unrelated and worth more. Once opening and closing counts began, the daily difference averaged about KES 120 short, every day, not varying much. That consistency pointed away from theft and towards a pricing or change-giving error, and it turned out to be a line priced at 85 on the shelf and 80 in the till. The count did not find a thief. It found a label.
Without clean daily records, tax time turns into guesswork, financing applications stall, and you cannot tell a genuinely good month from a lucky one.
Veira turns every sale into an organised record and a clear report, so your numbers are ready for KRA, a lender or yourself.
Knowing what the drawer should contain
A float count is only meaningful against an expected figure, and that expected figure is opening float plus recorded cash sales minus recorded cash paid out. Veira records sales by payment method, so the cash and mobile money figures stay separate and each can be reconciled on its own.
Because sales are attributed to a shift, a difference lands in a known window rather than against the shop in general. That is what turns step three from a ritual into a diagnostic.
The denomination mix in step two is specific to your prices and your customers, so the list is yours to build. The week of observation is the part nothing can shortcut.
Frequently asked questions
How much float should a small shop hold?
How often should the till be counted?
Why is my till always slightly short?
Should cash and M-Pesa be reconciled together?
What do I do about money taken from the till for purchases?
How often should takings be banked?
Does a shortfall mean somebody is stealing?
A float should be the smallest set of denominations that never runs out, and its value as a control depends entirely on being counted the same day. Build it from a week of observation, count it twice daily with two people, and reconcile cash and mobile money separately. See how Veira records takings by payment method so each one can be checked on its own.
Need help, or want to talk it through first? Chat with us on WhatsApp