Ask a transport manager for a carrier scorecard and you often get a page of twenty rows, each with a colour. Nobody can act on twenty rows. A review that cannot lead to a decision is a meeting about numbers.
A Thai study shows how long these lists grow. Researchers at Mahasarakham University built a performance index for domestic truck operators with 23 indicators: 6 about what the operator has, 9 about how it works, and 8 about what it delivers. They tried it on one operator, which scored 4.88 out of 5. That tells you how the index is used. It is a tool for an operator to mark itself, and an average of 23 scores near 5 is not something a shipper can use on a Tuesday.
This is how to cut it down to five measures you can run from one simple record, and what to leave out.
Count trips, not impressions
Look at the 8 result indicators in the study. Every one is built the same way: the number of trips in the month where something went right, divided by all trips in the month. That is the right shape for a shipper too. A trip is the unit. Each measure below is a yes or no on one trip, and the monthly figure is how many trips said yes.
That has a useful side effect. Because each measure is a yes or no on a trip, the trips that said no can be listed and discussed one by one. A percentage tells you there is a problem. The list tells you which truck, which day, and whose fault.
The five
1. Did they take the job? Count the jobs you offered and the jobs a carrier accepted the first time, without you phoning around.
This is the one measure that no trip record can show you, because a refused job never becomes a trip. A carrier that takes only the easy lanes can look perfect on every other measure. The Ministry of Industry's handbook shows how easily the base number slips: for its transport delivery rate, the table line calls the base figure deliveries made, and the note under it calls the same figure delivery orders received. Decide it for your own books. The base is what you offered. Why a carrier turns a job down explains what usually sits behind a refusal, and the count by lane and weekday is how you find out which of those it is.
2. Was the truck at your dock inside the booked window? The Thai study scores this as on-time collection against the time agreed with the customer. It is the early warning of the five, because a truck that loads late is already late. It is also partly your measure. If trucks arrive on time and then queue at your gate, the delay is yours. Record both the arrival and the time loading started, so the number can be split.
3. Was it delivered inside the booked window? This is the headline measure, and the one both the Thai study and the Ministry's handbook include. How to define on time properly, with the six choices that move the number, is in the article on measuring on-time delivery honestly. Keep the minutes late as well as the yes or no. Ten minutes late and five hours late are both a no, and they need different conversations.
4. Was it delivered right? Right place, right goods, right quantity, no damage. The study keeps these as separate indicators. The Handbook of Logistics and Distribution Management groups damage, shortage and refusal together as delivery problems, next to accurate documents. For a shipper one column is enough, with the exception written next to it: short, wrong, damaged, refused, or paperwork missing. The note is worth more than the tick.
5. Was the invoice right? Does the invoice match the agreed rate card: the trip rate, the fuel surcharge worked out by the agreed rule, the waiting charges agreed, the drops agreed. The Handbook lists accurate invoicing among the parts of the perfect order, and its worked example gives it the weakest score of the five, 94%. Invoice errors are also a standard measure in the supply chain reference model it describes. A carrier can drive well and still cost you hours every month in corrections.
One number from the five: the perfect trip
The Handbook's perfect order is simple to state. It is the share of orders where everything went right at once. For a carrier, that is the trips where all five measures say yes.
Its worked example shows why one headline figure is useful. Take five measures at 95%, 98%, 99%, 97% and 94%. Each looks fine alone. Multiplied together they give 84%. The Handbook's point is that the customer's experience is worse than any single line suggests.
There is a catch. Multiplying assumes the failures land on different trips. If one bad day causes a late arrival, a late delivery and a wrong invoice on the same trip, there are fewer bad trips than the product suggests. With your own record you do not need to guess, because you count the trips with no "no" in the row. For the example above the true figure has limits you can work out: it cannot be better than 94%, the weakest single measure, and it cannot be worse than 83%, which is 100% minus all the failures added together (5 + 2 + 1 + 3 + 6 = 17). At 80 trips a month, that is somewhere between about 5 and 14 trips with at least one problem. Counting tells you which.
These are the Handbook's illustration, not Thai data. Use them to see the shape, then use your own trips.
What to leave out
The study's other indicators are not wrong. They belong somewhere else.
- Things you check once. Contracts with customers, share of drivers holding the right licence, a quality certificate, the age and make-up of the fleet. These describe whether to hire a carrier, not how a carrier performed this month. Check them at hiring and at renewal.
- The carrier's internal cost controls. The study counts trips where fuel burn was above the operator's own standard, and trips with a traffic fine. Both are measured against the operator's own rule, which you cannot see, and both are the operator's own cost. You pay a trip rate and a fuel surcharge. If bad driving costs you something, it shows up as a late trip, a damaged load or a refusal, and you are already counting those.
- Your own transport spend as a share of sales. The Ministry's handbook has this as a company indicator. It is a measure of your budget, not of a carrier, and it belongs in your own cost review.
- A single average score. The study's 4.88 hides everything in the previous section. An average of 23 can sit near the top while the one measure you care about slips. Do not score carriers out of a hundred.
- The study's score bands as targets. Look at how it scores. For on-time collection, on-time delivery and the right goods, anything from 80% to 100% earns the full 5 points. A carrier late on one trip in five scores the same as a perfect one. Yet for damage and for the right place, 80% to 84% earns only 2. These bands were set by experts for a self-assessment tool, not for a contract. Writing an SLA a carrier can meet shows how to set a floor from a lane's own record instead.
The record that produces all of it
One row per trip, with these columns:
- Date and lane
- Time the job was offered, and whether it was taken first time
- Booked collection and delivery windows
- Actual arrival at your dock, loading start, and delivery time
- Minutes late at delivery
- Exceptions: short, wrong, damaged, refused, paperwork
- Invoice matches the agreed rate card, yes or no
Seven columns, filled in by the people who already handle the trip. A spreadsheet is enough.
Once a month, do three things:
- Count the five measures and the perfect trips.
- List every trip with a "no". Write one line of cause next to each, and mark whether it was your side or theirs.
- Give each measure an action. A falling acceptance count means look at the lanes, days and notice you are offering. Late collection means look at your own dock first. Late delivery means look at those trips' routes and times. Damage means look at the loading. Invoice errors mean the rate card needs tightening. A measure that has no action attached is the first one to drop.
On a small lane, one trip is a big swing. At 20 trips a month, one trip is 5 points. A bad month on a good carrier is normal at that size, and the table in the SLA article shows how often. On small lanes, review the list of trips rather than the percentage.
What to do this month
- Start the seven-column record on your busiest lane. Eight weeks of it is enough to see a pattern.
- Fill in the five measures and the perfect-trip count at the end of each month.
- Read the trips with a "no" before you read any percentage.
- Drop any measure that no one can act on.