Why Supplier Diversification Does Not Fix Customer Concentration
Published September 22, 2026
Altsets
Research by Altsets Research
Micron's supplier breadth does not offset its 17.62% Nvidia revenue exposure. Separate customer demand, procurement costs and qualified replacement capacity.
Dataset reference:Altsets supply chain relationships dataset covers 115k+ entities, 550k+ relationships, and history since 2006 across the full dataset; individual analyses use a narrower scope.
Key findings
- Micron’s multiple supplier relationships do not reduce its displayed 17.62% Nvidia revenue dependence.
- An ASML cost exposure and a Nvidia revenue exposure cannot be subtracted into a net concentration score.
- Demand and supply-continuity hypotheses require different outcomes and can interact through manufacturing investment.
Supplier diversification does not fix customer concentration because the two describe different sides of a business. More upstream vendors may help a company obtain inputs; they do not replace revenue from a major downstream buyer. Altsets makes that distinction concrete for Micron: the mapped suppliers include ASML, Lam Research, Applied Materials, KLA and Shin-Etsu Chemical, while Nvidia still represents an estimated 17.62% of Micron revenue.
That is enough to reject a single diversification label, but it is not enough to call Micron's production resilient. Several suppliers can perform different manufacturing functions rather than provide alternatives for the same input. In the displayed estimates, ASML accounts for 11.91% of Micron's cost base. That procurement measure cannot be subtracted from Nvidia's revenue share, and the number of other suppliers cannot be treated as an offset to it.
For an investor or stock model, the useful output is therefore two separate assessments: how customer purchasing could change sales, and how input availability could constrain production. The relationship data changes the analysis by identifying the named counterparties and the correct financial denominators. Additional product, inventory and qualification evidence determines whether either dependency can actually be reduced.
Two exposures that cannot be netted
| Observation | What it measures | What would reduce the dependency? |
|---|---|---|
| Nvidia: 17.62% of Micron revenue | Micron sales associated with a named customer | Other customer sales, a changed customer mix or less dependence on that purchasing stream |
| ASML: 11.91% of Micron costs | Micron procurement associated with a named supplier | Relevant qualified alternatives, changed input requirements or other supported operating changes |
| Several equipment and materials suppliers | Breadth of observed upstream relationships | Evidence that multiple vendors can perform the same required function |
The first two percentages refer to different financial bases. The third observation is a count or classification of relationships. None provides a common unit in which to calculate a net concentration score. Even two suppliers with similar cost shares need not provide the same process capability.
This also explains why supplier diversification and customer diversification require separate evidence. Winning another buyer does not qualify an alternative lithography tool. Adding an equipment vendor does not demonstrate that another customer will buy the memory previously sold to Nvidia.
A demand shock has a direct revenue bridge
A hypothetical 10% decline in Micron sales to Nvidia contributes -1.762% to total Micron revenue if all other customer sales are unchanged. The calculation uses the 17.62% starting revenue share and a shock to the sales on that relationship. It does not assume that a 10% fall in Nvidia's own revenue produces the same purchasing change.
A supplier interruption needs a different bridge. ASML's 11.91% cost share does not mean an interruption removes 11.91% of Micron revenue. The analyst needs to know which production or investment activity is affected, whether installed equipment remains usable, what inventory is available and how quickly the input can be replaced. A high procurement share helps prioritize investigation; it does not supply the lost-output estimate.
The Micron revenue sensitivity analysis develops the purchasing calculation. Keep that demand scenario separate from a supplier-continuity scenario even when both eventually affect the same revenue forecast.
Four operating conditions, rather than one diversification score
| Relevant sourcing condition | Revenue spread across customers | Revenue concentrated in a major customer |
|---|---|---|
| Multiple qualified sources for the affected input | Evidence may support less dependence on both one input source and one buyer | Input alternatives may preserve production, but do not replace the major buyer's demand |
| Limited replacement options for the affected input | Customer breadth leaves the production bottleneck unresolved | Production continuity and concentrated demand both need stress tests |
This is a conceptual framework, not a measured classification of Micron. The mapped supplier breadth and Nvidia revenue share establish different observations. They do not identify qualified alternatives for every manufacturing step. The shared-supplier analysis explains why function matters when interpreting supplier breadth.
Let capacity explain an interaction
The two exposures can interact through operating conditions. Strong customer demand may make a supplier delay more expensive when production is already constrained. Weak demand may leave inventory or spare capacity that absorbs some interruption. Those are hypotheses about utilization, timing and affected products, rather than reasons to subtract one exposure from the other.
A quantitative model should first retain separate customer-demand and procurement features and test them against outcomes suited to each channel. Customer purchasing belongs with supplier sales revisions; input availability belongs with production or delivery outcomes. An interaction needs evidence that the relevant operating state was observable at the decision date and that its contribution survives later test periods.
The conclusion supported by this Micron example is precise: supplier breadth does not neutralize the mapped Nvidia revenue dependence. Whether either side becomes less risky depends on a change in its own economics. Altsets methodology defines the directional estimates and their limitations.
Sources
Related research
Methodology
Cite this research
Altsets Research. "Why Supplier Diversification Does Not Fix Customer Concentration." Published September 22, 2026. https://www.altsets.com/research/supplier-diversification-does-not-fix-customer-concentration