Five numbers a small fleet should look at every Monday

The week that just ended is knowable. Every job was booked, driven, signed for and invoiced, and all of it is written down somewhere. Yet in most small fleets Monday starts with a feeling rather than a figure: it was a busy week, the phones did not stop, and the bank balance disagrees with all of that in a way nobody has time to investigate before the first driver rings in.

The problem is not a shortage of data. It is that everything is measurable, so nothing in particular gets measured. A number you look at once, during a bad month, tells you nothing — it has no trend behind it and no habit in front of it. Five numbers, looked at every Monday, tell you what changed and usually why.

Why five, and why Monday

Five is roughly what a person keeps in their head without a printout. Beyond that you are reading a report, and reading a report is a different activity from running an operation — it happens less often and it changes less.

Monday, because the cadence has to match the decision. Daily is noise: one cancelled load and a truck stuck at a border move any of these figures far enough to look like a crisis. Monthly is a post-mortem — by the time the month closes, the four weeks that made it are gone and nothing about them is still editable. A week is short enough that the next one can still be changed and long enough to average out a bad Tuesday.

1. Utilisation

The share of available vehicle time that was sold. Of the days each vehicle could have worked, how many carried a revenue-earning job.

The whole number lives in the word available, so decide what it means before you measure it and then do not move it. If a truck sitting in the workshop counts as available, utilisation drops and dispatch takes the blame for a maintenance problem. If it does not count, the workshop stops appearing in your numbers at all — which is why unplanned downtime is on this list separately.

Utilisation is the number that answers do we need another vehicle? When it sits in the sixties, another truck buys you the same problem twice: a second asset earning nothing on the same days as the first, and one more insurance premium.

2. Empty share

Empty kilometres as a share of total kilometres — or, if your data will not support that yet, the count of legs that ran with nothing on them.

Some empty running is structural. A lane that flows one way, out to a region that ships nothing back, produces empty returns no planner can argue away, and chasing that share to zero means taking backhauls that price below cost. What the weekly figure is good for is the change, and the distribution behind it: two lanes usually own most of it, and those two are a commercial conversation rather than a planning failure. The rest is sequencing, and sequencing is fixable this week. We wrote about the trade-off in backhauls and empty kilometres.

3. Revenue per vehicle-day

Revenue for the period divided by the vehicle-days available in it.

This is the number that catches the week that felt busy and paid badly. Revenue per kilometre flatters long, cheap trunk work; revenue per job flatters whoever takes the small ones. Revenue per vehicle-day prices the thing you actually own — a truck, a driver and a day — and it moves when the mix shifts, which it does quietly and continuously as customers change what they send.

Treat it as a mix indicator, not a scoreboard. It falls for good reasons (deliberately taking volume to hold a customer through a soft quarter) and rises for bad ones (a single outsized job that will not repeat). Its job is to make you ask which.

4. Days sales outstanding

The average number of days between invoicing and being paid — receivables divided by credit sales for the period, multiplied by the days in it.

A small carrier finances its customers. Fuel, tolls and wages leave on a weekly rhythm; the money for the work that consumed them arrives on somebody else’s payment terms. That gap is the most common reason a profitable haulier runs short of cash, and it is a number rather than a mood.

The useful part is that the lever is rarely chasing harder. It is invoicing sooner and invoicing right the first time, because most of the tail is self-inflicted: the proof of delivery that took four days to reach the office, the invoice that went out without the reference the customer’s system requires and sat in a query queue for a fortnight. Both are yours to fix, and both show up here before they show up in the bank. See what paper proof of delivery costs you.

5. Unplanned downtime

Hours or days a vehicle was off the road without anybody having planned it — roadside failures, a repair that could not wait, a statutory date that lapsed and grounded the truck. Deliberately separate from scheduled service, which is a cost you chose and a slot you booked.

It is the leading indicator of the other four. Downtime this month is utilisation next month, overtime the week after and a customer conversation the month after that. It is also the number nobody records, because it does not arrive as a metric: it arrives as a bad morning, gets solved by phone, and leaves no trace except in the memory of whoever solved it. If you start recording one new thing this quarter, record this.

Reading a bad week without panicking

Small fleets have small numbers, and small numbers are volatile. With six vehicles, one truck in the workshop for three days moves utilisation by around seven points on its own. That is arithmetic, not a trend — and treating it as a trend is how a maintenance week turns into an unnecessary argument with dispatch.

So read them together rather than one at a time. Utilisation down and downtime up is a workshop week; nothing about sales or planning explains it. Utilisation flat and revenue per vehicle-day down is a mix or a pricing problem, and the customer list will say which. Empty share up with everything else steady is a sequencing problem, and probably a fixable one. Four weeks of context beats any single delta.

Then write down what you decided and why — one line is enough. Next Monday reads as the result of a test rather than as a fresh surprise, which is most of the difference between measuring and managing.

Why more dashboards is usually less management

The instinct, once weekly numbers start being useful, is to add more of them. A screen with thirty tiles feels like control and delivers the opposite: no tile has an owner, none of them implies an action, and the two that would have changed a decision are lost among the twenty-eight that merely moved.

Two honest caveats. These five are not the only defensible five — on-time performance, claims, driver turnover and safety all earn a place, and a fleet running mostly contract work would reasonably swap one out. The discipline is the count and the cadence, not this particular list. And a number you cannot trust is worse than no number: if empty kilometres are estimated by whoever happens to be asked, the Monday meeting will spend its time arguing about the figure instead of acting on it. Take the four you can source honestly over five you cannot.

navichain’s part in this sits upstream of the reporting. Bookings, runs, stops, signed deliveries, workshop records and invoices live in one system on one set of data, so figures like these are derived from the operational record rather than assembled from three exports and a spreadsheet — with live dashboards and a fleet health view on top, and invoicing that runs off the same delivery the driver signed. Every feature is on every plan and pricing starts at 995 kr a month including five users — see the platform page and the pricing page. Related reading: service intervals in kilometres and days.