How to Rank Supplier Disruption Exposure by Customer Cost Share

Key findings

• Customer cost percentage provides a first-pass ranking of how economically visible a supplier relationship is within each customer's cost base.
• SK Hynix to Nvidia at 27.33%, SK ecoplant to SK Hynix at 18.81%, and ASML to Micron at 11.91% show why disclosed supplier relationships should not be treated equally.
• The ranking is a research triage tool, not proof of operational dependence or replacement difficulty.

To rank exposure to a supplier disruption, start with each customer's cost percentage for that supplier, then verify product criticality, substitutability, inventory, and timing. The percentage prioritizes which customers appear most economically exposed; it does not predict the final operational impact.

Customer cost percentage is one useful starting measure. It estimates how much of the customer's cost base is associated with a mapped supplier relationship. Several real Altsets relationships show the range:

Those values do not tell you which disruption would stop production. They do tell you that treating every disclosed supplier relationship equally throws away economically useful information.

The question customer cost percentage answers

For disruption research, the useful question is: How much of the customer's mapped cost base is associated with this supplier relationship?

A larger percentage can justify earlier investigation because the relationship occupies more of the customer's economic input base. That is not the same as saying the supplier is irreplaceable.

A supplier can represent a meaningful share of costs while still being replaceable. Another supplier can represent a smaller share of costs but provide a highly specialized component that is difficult to substitute.

Customer cost percentage helps rank exposure. Operational criticality still requires additional evidence.

Example: the same supplier can matter very differently across customers

SK Hynix provides a clean example.

Altsets estimates SK Hynix to Nvidia at 27.33% of Nvidia COGS. The SK Hynix to Apple relationship is estimated at 1.76% of Apple COGS. If a hypothetical SK Hynix event affected both relationships, the displayed cost percentages give a clear reason to investigate Nvidia before Apple.

That conclusion is narrow. It does not predict production losses or stock returns. It tells you where the mapped economic exposure is larger.

This is one reason a supplier list is not enough. Both companies can name the same supplier while carrying very different levels of economic exposure.

Example: upstream infrastructure can matter too

Disruption analysis should not stop at famous component suppliers. The SK ecoplant to SK Hynix relationship is estimated at 18.81% of SK Hynix COGS and 3.1B USD in relationship size. That relationship sits one step upstream from SK Hynix's memory relationships with Nvidia and Apple. A disruption workflow can therefore move in stages:

1. identify the disrupted entity;
2. rank its direct customers by customer cost percentage;
3. identify economically important customers;
4. inspect what sits downstream from those customers;
5. decide whether second-order paths deserve investigation.

The goal is not to assume contagion. It is to prevent the analysis from stopping at the first visible layer.

Example: ranking several suppliers into one customer

The Micron network provides another use. Altsets estimates: ASML to Micron at 11.91% of Micron COGS, Lam Research to Micron at 5.52%, Applied Materials to Micron at 3.84%, KLA to Micron at 2.84%, and Shin-Etsu Chemical to Micron at 1.34%. Those percentages create a first-pass ranking of mapped cost exposure.

They do not prove that ASML is the hardest supplier to replace or that Shin-Etsu is operationally unimportant. The values simply tell you that the displayed economic shares differ materially. This can help decide which supplier event deserves the deepest immediate work.

What to investigate after the ranking

Once customer cost percentage identifies a high-priority relationship, the next questions are operational: Is the product specialized? Are alternate suppliers qualified? How much inventory is available? Are there geographic alternatives? What are lead times? Does the customer dual-source the input? Is capacity constrained across the industry? Does the relationship depend on one facility? Are there export controls or regulatory constraints? Can the customer redesign around the supplier?

Supply-chain relationship data is the triage layer. Those questions determine whether the economic exposure can turn into an operating problem.

Why relationship size still matters

A percentage without absolute scale can also mislead. A high customer cost percentage attached to a small relationship may deserve a different interpretation from the same percentage attached to a multi-billion-dollar relationship.

The SK Hynix to Nvidia edge combines 27.33% of Nvidia COGS with a 21B USD relationship estimate. The KLA to Micron edge combines 2.84% of Micron COGS with 693M USD.

Both can matter, but they represent very different scales. A useful disruption screen therefore uses customer cost percentage and relationship size together.

A repeatable disruption workflow

A disciplined workflow is:

1. Define the disrupted supplier, facility, geography, or input.
2. Map direct customer relationships.
3. Rank customers by customer cost percentage.
4. Add relationship size as an absolute scale check.
5. Preserve missing percentages as unknown.
6. Research substitute suppliers and switching constraints.
7. Expand to second-order paths only when the first-hop relationship is economically meaningful.
8. Separate observed data from disruption hypotheses.

The SK ecoplant and SK Hynix analysis illustrates the second-order step. The SK Hynix, Nvidia, and Apple comparison shows how the same supplier can have dramatically different mapped cost exposure across two customers.

For metric definitions and interpretation limits, read the Altsets supply-chain data methodology. Browse Supply-Chain Data Use Cases for other research workflows.

Methodology

Read the methodology for this research.

Related research

• SK Hynix, Nvidia, and Apple: Which Relationship Is More Economically Important?

• How to Screen for Replacement Suppliers After a Disruption

• What Supply-Chain Data Adds to Credit and Bond Analysis

• What Sits Upstream of SK Hynix? The 3.1B SK ecoplant Relationship