Don't build supply chains. Build value chains.

A supply chain moves goods. A value chain moves goods and captures the value created by moving them.

Most organisations have built the first and assumed they were building the second. For years that assumption held, because moving goods well was genuinely scarce. It is not scarce any more. The networks are mature, the software is commoditised and the remaining efficiency is smaller than the cost of chasing it.

The advantage has moved somewhere else.

I. The material sets the terms

I trained as a materials scientist before I worked on markets, and the order matters.

Every commercial decision a business makes is eventually a claim about the physical world: that a thing can be made at a cost, moved at a speed, held at a temperature, sold before it degrades. Strategy that has not been tested against weight, distance, time and chemistry is not strategy. It is a wish with a slide deck.

This is why so much value chain work fails. It begins at the commercial end and reasons backwards toward a physical reality that will not negotiate. The material has already decided several things for you. Work forwards from it, and a set of options appears that were invisible from the other direction.

II. Efficiency concentrates. Alignment distributes.

Here is the part that is rarely said plainly.

Optimisation, pursued far enough, concentrates. It consolidates suppliers, shortens supplier lists, favours the counterparties who are already largest and already trusted, and quietly removes everyone who cannot yet meet a standard they were never given the means to meet. Every individual decision is rational. The aggregate outcome is a global economy in which participation narrows even as volume grows.

Alignment behaves differently. When you align what a market genuinely demands with what a distribution system can physically carry — and make both legible to everyone in the chain — you do not only reduce cost. You make it possible for parties who were previously illegible to participate at all.

That is not a moral addendum to the commercial argument. It is the commercial argument. A value chain with more qualified participants is more resilient, more competitive on input cost, and less exposed to the failure of any single node. The last five years have made that case better than any consultant could.

Value chain engineering should not only optimise efficiency. It should widen trusted participation in the global economy.

III. Trust is infrastructure

Two parties who do not know each other, in two jurisdictions with different rules, cannot transact at scale on goodwill. They transact on infrastructure: shared standards, interoperable data, verifiable provenance, and a way of establishing facts that neither side has to take on faith.

We are used to thinking of infrastructure as physical — the port, the road, the plant. But a trade corridor with no data standard is as impassable as one with no road. Most of the friction now sits in the second kind of infrastructure, and almost none of the investment does.

Building it is not primarily a technology problem. The technology mostly exists. It is a coordination problem between parties with different incentives, different regulators and different reasons to be cautious. Which is why the work looks less like consulting and more like diplomacy.

IV. The missing component is the decision

A value chain has an unlisted component that appears on no network diagram: the speed and quality with which the organisation’s leadership can decide.

I have watched businesses with excellent networks lose to businesses with mediocre ones, for a single reason — one could resolve a question in nine days and the other took six weeks. In a market that moves in two, a six-week decision is not a slow decision. It is the wrong decision, arrived at carefully.

Decision-making is treated as culture, or leadership, or personality. It should be treated as infrastructure: designed, instrumented, given a cadence and a set of decision rights, and measured like anything else that determines throughput.

Most organisations have never seen this number for their own business. It is usually the largest single source of leaked value they own.

V. What follows

If the argument holds, four things change about how this work should be done.

1. Start at the material, not the market. Establish what the physical world permits before deciding what the commercial plan requires.
2. Design for participation, not only for cost. Ask who your standard excludes, and whether excluding them is a decision you actually made.
3. Treat data standards and traceability as capital projects. They belong in the business case beside the asset, not in a change budget afterwards.
4. Instrument the decision. Measure how long a significant decision takes, how many parties must agree, and how often it returns unresolved. Then design that down.

Where this is applied

This is not a theory I hold at a distance. It is the method behind both parts of my practice.