When to redesign your supply chain network, and when not to

Few decisions in a mid-sized business are as hard to reverse as the shape of its distribution network. Where the warehouses are, how many there are, which ports and carriers feed them and how far the trucks travel to customers determines a large share of logistics cost and service for years. Leases run long. Moves are disruptive. A badly timed redesign can absorb a year of management attention and deliver a network that suited last year's problem.

Equally, plenty of businesses run networks designed for a customer base, product range or sourcing pattern they no longer have. They pay for it every day in extra miles, double handling and stock in the wrong place.

So the question is less about "should we redesign?" and more about "has something structural changed?"

Good reasons to redesign

The demand footprint has moved: if the mix between channels has shifted, if a large customer has changed how it wants to be served, or if growth has come from regions your sites were never placed to reach, the network is solving the wrong problem.

Sourcing has moved: a network built around one port of entry and one origin region may not suit a new one. The UK-India trade agreement, which entered into force on 15 July 2026 and removed UK tariffs on 99% of Indian goods, will lead some businesses to shift volume. A change in origin can change the best inbound port, the right stockholding point and the case for a consolidation hub.

The cost of a material input has changed for good: the UK's carbon border adjustment mechanism starts in January 2027 for imports of aluminium, cement, fertiliser, hydrogen and iron and steel, for businesses importing £50,000 or more of those goods. Where that changes the economics of where components are sourced or finished, it may change where they should be stored.

A lease event is approaching: lease breaks and expiries are the natural window to test alternatives. If you do not model options 18 to 24 months ahead, the renewal will decide for you.

An acquisition has doubled the footprint: two networks bolted together are rarely the best single network. Integration is often where the largest savings sit.

Poor reasons to redesign

A fuel spike: Logistics UK reported diesel up 36% in the year to April 2026, and fuel makes up around a third of the operating cost of an HGV. That hurts, and it should be in your scenarios. But a network designed around one year's fuel price will be wrong when the price moves again. Test sensitivity to fuel; do not build for a single number.

A shipping disruption that is unwinding: when lines diverted around the Cape, many businesses added inland buffer stock and in some cases extra space. Seatrade Maritime reported in October 2026 that Suez transits had recovered to more than half of pre-war levels, with a full return expected by early 2027. Shorter, more reliable ocean transit may reduce the stock you need, so committing to new space for it now may be premature.

An attractive property offer: rents are rising. Logistics UK put prime rents for large units at £13.38 per square foot, up 13% in a year. A good deal on a building is a reason to look closely, not a strategy.

One bad peak: if the network failed at Christmas because of a labour shortfall or a system problem, fix that first. The sites may be fine.

A business wanted a new national distribution centre because its existing site felt full. The data showed the space was full of slow-moving stock and not active lines. Clearing and relaying the existing site bought three years of growth, and when the move did come, the business case was built on real demand rather than a property opportunity.

What a proper study looks like

A credible network study starts with a baseline that reconciles to the accounts. If the model cannot reproduce last year's actual logistics cost within a few percent, nothing it says about the future can be expected to be reliable.

Next comes demand by location and by service requirement, at postcode level if possible, and cost-to-serve by customer type. Then scenarios: the current network optimised, one or two structural alternatives, and the options the leadership team already has in mind, so they are tested rather than argued about.

The part most often skipped is the cost and risk of getting there. Dual running, stock transfer, recruitment at the new site, service dips during cutover and redundancy at the old one. To take illustrative numbers, a scenario that saves £1.5m a year but costs £3m and a bad peak to implement is a different decision from one that saves £1m with little disruption.

Ownership should sit with the COO, with finance validating the baseline and the commercial team confirming the service assumptions. A focused study takes weeks. And if the answer is to move, the implementation takes much longer.

Cheaper alternatives to check first

Before redesigning, test the smaller levers. Re-slot fast movers. Revisit stock placement between existing sites. Re-tender carriers on the lanes that have changed. Change delivery days or minimum drops for low-value customers. Many networks that look wrong are actually poorly operated rather than poorly located.