Ask a fleet owner where they lose money and they will usually say fuel, insurance or driver retention. Ask them how many calls they dropped between midnight and 4am last Saturday and most cannot tell you. That gap is the point of this article.
Why nights matter more than you think
Night work in this industry is disproportionately valuable. Fares are often higher, competition from public transport disappears, and the customers are frequently the ones with least choice — which means the fleet that answers gets the job and often the repeat custom.
It is also where fleets are weakest. We routinely see operators dropping 30–40% of calls after midnight. Every one of those is a fare that went to a competitor or an app, plus a rider who learned that you do not answer at night.
That last part is the expensive bit. A dropped call is not a lost fare, it is a lost customer relationship. People do not retry a number that rang out at 1am; they save a different one.
The rota maths nobody does
Here is the calculation that changes minds. To keep one seat continuously staffed, 24/7, you need:
- 168 hours a week of coverage.
- A full-time employee works around 37.5 hours.
- That is 4.5 FTEs before anything goes wrong.
- Add 28 days statutory holiday — roughly another 11%.
- Add sickness, training days and turnover gaps.
Realistically five full-time people per concurrent seat. If your Friday peak needs three seats, that is fifteen people to cover one operation properly — and that is before you have thought about who covers the person covering.
Most small fleets never do this sum. They staff nights with one person, no depth behind them, and absorb the failures personally. The owner becomes the absence cover, which works right up until it does not.
If your night rota has no float capacity, you do not have a night shift. You have a single point of failure that happens to work most weeks.
Why night shifts break
- Recruitment. Few people want permanent nights. The pool is small, so you compromise on quality or pay a premium that wrecks the economics.
- Retention. Night workers burn out. Sustained night work carries real health costs, and people leave — often suddenly.
- Performance. Human alertness bottoms out around 3–5am. Even a committed dispatcher is measurably slower then.
- Isolation. One person alone overnight has nobody to escalate to. When something unusual happens, they either guess or call you.
- No cover depth. One person off sick means the owner works the desk, or the phone rings out.
Notice that only one of those is about the individual. The rest are structural, which is why replacing the person rarely fixes it.
Designing a rota that holds
If you are keeping nights in-house, these are the things that actually help:
- Staff to the curve, not the average. Pull a month of call data by hour. You will likely find a sharp Friday and Saturday 9pm–2am spike and a very quiet Tuesday 2–5am. Flat staffing gets both wrong at once.
- Build in float capacity. Roster above minimum on peak nights. It feels wasteful until the first time someone calls in sick at 8pm on a Saturday.
- Rotate rather than fix. Permanent nights burn people out. Rotating patterns retain better, at the cost of more scheduling work.
- Never leave one person alone at peak. Two on a Friday night is not luxury; it is the difference between managing a surge and drowning in it.
- Write down the escalation path. Overnight staff need to know exactly what to do about an accident, an abusive caller or a driver who stops responding — without phoning you.
- Measure the night separately. Blended daily averages hide overnight failure completely.
What to measure
Four numbers, reported for the night window on its own:
- Answer rate — calls answered within SLA, as a percentage of calls offered.
- Abandonment rate — callers who hung up waiting. Above 5% overnight, you have a staffing problem.
- Median answer time — the median, not the mean. One 20-minute outlier distorts an average badly.
- Booking conversion — of the calls answered, how many became a job. Low conversion with high answer rate points at pricing or availability, not the desk.
If your phone system cannot give you these split by hour, that is the first thing to fix. You cannot manage a night shift you cannot see.
Your realistic options
1. Staff it properly in-house. Five FTEs per seat, float capacity, rotating patterns, documented escalation. It works, and it is expensive.
2. Voicemail or an answering service after midnight. Cheap, and it loses you the most valuable calls of the week. Generic answering services take a message; they do not dispatch a car.
3. Automate with IVR or an app. Helps with simple bookings. Falls over on exactly the messy overnight cases — an unfamiliar pickup point, a rider who has had a few drinks, a driver who has not arrived.
4. Outsource the night window. This is what we do, and the reason it works is not effort — it is arithmetic. Your 11pm–5am is 4am–10am in Lahore. The back half is a normal morning shift and the front half is one antisocial block rather than the entire operation. Rested people answer better than exhausted ones, and we carry the float capacity so absence never reaches you.
Most fleets that come to us start here: keep the day desk, hand over nights and weekends. It is the shift where in-house economics are worst and where the time zone advantage is largest.
Whichever route you take, start by pulling last month's call data by hour. Whatever you find will be more informative than any article, including this one.
Ride Dispatchers Operations Team
We run outsourced taxi dispatch and contact-centre desks from Lahore for fleets across the UK, Ireland and North America. Everything here comes from work on live fleet accounts, not from a keyword tool.