Operations statistics: the numbers that matter and the ones that mislead
31. 7. 2026
Once your rides live in an application rather than on paper, you get numbers for free. The trouble is that numbers on their own decide nothing — and some of them actively push you towards the wrong call. Here is what has worked for us.
Average versus median
The average fare is treacherous. Three long trips to the county town in a month are enough to pull it up, even though the overwhelming majority of journeys are short hops around town. The median — the value in the middle — shows the ordinary ride far more honestly.
A simple rule: budget with the average (because budgets add up), judge what a normal day looks like with the median.
Always ask how many rides a number is built on
A figure without a sample size is just an impression. "Average fare £11.40" means one thing when it rests on 400 rides and something else entirely when it rests on nine. Distance is trickier still: the driver enters the kilometres and may not always do so, so the statistic can only count the rides where they did.
A good overview tells you how many rides sit behind each figure. If you cannot see that anywhere, ask.
The cancellation rate is the most useful single number
Revenue tells you how you are doing. The share of cancelled rides tells you where your problem is. When it jumps from five per cent to fifteen, something happened — and it is worth finding out what while you still remember the month.
That is why it pays to note a short reason when cancelling. On its own it changes nothing, but after three months the totals reveal whether it was "no show", "no car available" or "booked twice by mistake". Each of those three sentences leads to a different fix.
Demand by hour plans your shifts for you
The spread of rides across the day is surprisingly sharp in senior transport: a morning peak around surgery hours, a quiet spell after lunch, a smaller second wave in the afternoon. You can shape the driver's shift and the hours somebody sits by the phone around that — and it costs nothing.
What not to trust
- Short periods. One week in January tells you nothing about a year.
- Comparisons without context. December is never like November; holidays, flu season and weather all interfere.
- Data that gets deleted. We delete call records after 90 days for privacy reasons, so the booking-source statistic cannot reach further back. That is not a fault, but you need to know it.
- Numbers with no decision attached. If nothing you do would change because of a metric, feel free to ignore it.
How much time to spend on this
Ten minutes once a month. Genuinely. Open the overview, look at the cancellation rate, the demand by hour, and whether revenue matches what you expected. If it all looks normal, close it and go and do something more useful.
Statistics are visible to the company administrator only — dispatchers and drivers do not get the tab. Click through them in the demo, no registration needed.
Running a taxi or senior transport service? Send us a no-obligation enquiry or try the demo.