You have chosen a new carrier. The tender is done, the rate is agreed, and the old carrier has been told. The risky part is only starting.
Most switches that go badly were not a bad choice. They were a bad start: all the volume moved on one day, no agreed targets for the first weeks, and nobody sure whose fault the first late truck was. This article is for the person who has to make the switch work, usually the logistics manager, and it assumes full truckloads on a set of lanes.
Short answer
- Do not move everything on one day. Move a slice of the lane, check it against a gate, then move the next slice. Keep the old carrier on the rest until the new one has earned it.
- Set start-up targets lower than steady-state targets, and write both down before day one. The textbook calls expecting full results on the first day a classic client mistake.
- Keep a daily exception log for the first two weeks. Each line says what went wrong and whether the cause was on your side or theirs.
- Name one owner on each side. The textbook recommends a named manager from the client, and one point of contact for the provider.
- Decide how the start-up phase ends. A fixed run of weeks at the steady target, then you move to the normal review rhythm.
Why the start fails, in the textbook's own list
The textbook lists the ways an implementation goes wrong, and says they come from both sides. Read the client column first, because it is the one you control.
- Unrealistic goals at start-up. Teething problems are certain, so a provider cannot be expected to hit every target from day one.
- Last-minute requests and a moved start date. New lanes, new rules or a new start day, added late.
- A hands-off attitude. Signing the contract and waiting.
- Metrics used as a whip. Numbers that punish the provider instead of helping both sides manage the operation.
- Reacting too severely to the first quality breakdown.
- Expecting total adherence to the contract from the first day.
The provider's column has its own items: not managing the client's expectations, not agreeing start-up metrics, not measuring performance during start-up, and moving to steady operation before the start-up phase is complete. You cannot fix their column, but you can ask for each of those things in writing before you start.
Before day one: what has to exist
A switch needs a small project plan, with tasks for both sides and a date that does not move. Before the first truck:
- The handover pack. Every pick-up and drop address with its opening hours, gate rules, vehicle restrictions and a named contact. What the new carrier needs to quote accurately is also what it needs to run. See what a carrier needs to quote for the list.
- The paperwork loop. Which document travels with each load, who signs it, and how the signed copy comes back to you. Settle this now, because it is where invoicing later gets stuck.
- The old contract's exit terms. If the old carrier needs notice, a switch that is too sudden means paying for both or leaving the lane uncovered. The contract clauses article shows where notice and exit terms usually sit.
- One named owner on each side, with phone numbers, and an agreed time each day when exceptions are reviewed for the first two weeks.
- A start date that stays put. Avoid the week before a long public holiday, and avoid moving the date once it is announced.
Move the volume in slices
The aim is to find the problems while they are small. Take one lane with 60 trips a month, about 15 a week. This is an illustration, not a Thai statistic.
| Week | Share to the new carrier | Trips to the new carrier | Trips left with the old carrier |
|---|---|---|---|
| 1 | 20% | 3 | 12 |
| 2 | 40% | 6 | 9 |
| 3 | 60% | 9 | 6 |
| 4 | 100% | 15 | 0 |
Over the four weeks the new carrier handles 33 trips (3 + 6 + 9 + 15) and the old carrier the other 27. If the first week shows that the new carrier arrives late at one particular plant, you have a problem on 3 trips, not on 15.
The gate between steps is the point. Do not move up a slice on the calendar alone. Move up when the previous slice met its start-up targets, or when the cause of a miss was found and fixed.
Start with the lane that is easiest to recover. A lane with a flexible delivery window teaches you more cheaply than a lane feeding a production line that stops when a truck is late.
Two sets of targets
The textbook says start-up metrics and steady-state metrics should both be agreed, ideally during negotiation, and that start-up metrics will need adjusting once you see the real operation. So write two tables.
- Steady state. Use the old carrier's actual record on the lane as the base, and see writing a delivery SLA a carrier can meet for how to set the floor from a lane's own history.
- Start-up. A deliberately lower and stepped target, written in advance. For example, a lower on-time figure for the first two weeks, a higher one for weeks three and four, then the steady figure. These numbers are yours to choose. What matters is that they exist before the first load.
In the first weeks, measure the things that explain a miss as well as the result. Which of these mean something is covered in the five carrier measures worth tracking. For a start-up, add: the arrival time against the booked time at the pick-up gate, whether the documents were complete when the truck left, and how many days the signed delivery copy took to come back.
The daily exception log
For two weeks, keep one page. A line for every miss: the date, the lane, what happened, and a cause under one of four headings.
| Cause | Examples |
|---|---|
| Ours | Load not ready, wrong address in the pack, order changed after booking, dock occupied |
| Theirs | Truck late with no warning, driver not briefed, wrong vehicle type |
| Shared | Instruction unclear on both sides |
| Outside both | Road closure, weather, a breakdown that was reported promptly |
The log serves three purposes. It stops the first argument from becoming a blame fight, because the cause is written down at the time. It shows how many misses are yours, and the textbook says a client's own inaccurate information is one of the usual causes. And it gives the review meeting facts instead of impressions.
How to react to the first failure
The first late truck will happen. The textbook names over-reaction to the first quality breakdown as a client mistake, and using metrics as a whip as another.
- Find the cause with the carrier within a day. Write it in the log.
- Fix the cause in the handover pack or the instruction, so it cannot repeat.
- Hold your response for a repeat of a cause that has already been fixed.
If the same cause comes back after it was fixed, the conversation changes. Then it is a performance issue, and you already have the dates.
Ending the start-up phase
Decide the exit before you begin: for example, four weeks in a row at the steady target on the full lane, with no unexplained miss open in the log. On one lane that is about eight weeks with the ramp above. Switching several lanes one after another is why we use 90 days as the planning allowance. It is a planning figure, not a Thai finding. When you reach the exit:
- Stop the daily exception call and move to the normal review rhythm. The textbook describes weekly operational meetings between supervisors and a regular strategic review between senior managers.
- Tell the old carrier whether it keeps any share. If you decided to keep a second carrier on the lane, see how many carriers one route needs.
- Replace the start-up targets with the steady targets in the contract's service terms.
What to do on Monday
- Write the handover pack for the first lane, and check it with someone who has never driven there.
- Draw the four-week ramp for that lane with a gate at each step.
- Write the start-up and steady-state targets in two columns.
- Agree the one owner on each side and the daily exception time.
- Decide, now, what ends the start-up phase.
Where the evidence stops
The method here comes from the textbook's chapter on managing an outsourcing arrangement. We did not find a Thai study that measures what happens when a company changes road carriers, so the ramp example and the sample targets are illustrations to set against your own record, not Thai findings.