Changing a supply chain while it keeps moving

Nobody gets to switch a supply chain off for a few weeks while the change is made. Orders arrive, vehicles leave, and customers expect what they were promised. Moving a warehouse, replacing a transport system, changing a supplier base or redrawing a network all has to happen around live operations, and that shapes how the work should be run.

Begin with what must not stop

Before designing the change, write down the service commitments that cannot be broken: cut-off times, delivery windows, regulatory obligations, and the customers whose loss would cost most. Agree them with operations and commercial colleagues, not just the project team. These become the boundaries the plan works within, and they stop the project quietly trading service for schedule.

Then map what the change touches. Follow an order from the customer through systems, stock, people and transport and back again, and mark every point the change affects. The dependencies nobody mentioned in the workshop usually appear in this walk-through, which is far better than meeting them on the night of go-live.

Change in slices, not in one move

A single cutover is attractive on paper: one date, one plan, one celebration. In a live operation it concentrates all the risk in one weekend. Wherever it can be done, split the change into slices that can each be started, checked and, if necessary, reversed: one site, one customer group, one product family, one lane at a time. Each slice should have:

• A clear scope and a named owner.
• Measures that show whether service has held.
• A rollback route that has been rehearsed, not just written down.
• A decision point where someone with authority says go or stop.

Choose the first slice to teach you the most at the lowest cost of being wrong. A representative but forgiving part of the operation is better than the easiest one, because an easy pilot proves little. Agree in advance what a good result looks like, and resist the pressure to speed up after a good first slice. Early success is when confidence rises faster than evidence. Let the measures, not the mood, set the pace.

Plan the way back as carefully as the way forward

A rollback plan that has never been tried is a hope. Test it where you can, and be honest about the point beyond which going back is no longer realistic. That is the moment for the most senior decision, made with the facts in front of the people making it.

Define the triggers in advance: for example, service falling below an agreed level for a set period, or a defect that affects safety or compliance. If the triggers are agreed before the pressure arrives, stopping becomes a procedure rather than a confrontation. Keep parallel running only as long as it is useful, since running two ways of working doubles the load on the same people.

Look after the people and the suppliers

The people who run the operation are being asked to keep service up and to learn something new at the same time. Involve supervisors and planners early, because they know where the workarounds are. Train close to the date, on the real process, and put experienced people on the floor in the first weeks.

Suppliers and carriers need the same honesty. Tell them what is changing, when, and what you need from them, and agree who to call when something goes wrong. A great deal of disruption in a transition comes from something one party assumed the other had told them. Keep communication short and regular: a daily ten-minute call in the first weeks, with a single list of open problems, beats a weekly report no one has time to read.

After each slice, measure service against the commitments you started with, write down what you learned and change the plan for the next one. That small loop is what keeps a change from drifting.

If you are planning a change to a live supply chain, talk to us and we will help you work out how to run it without stopping the business.