How to Screen for Replacement Suppliers After a Disruption
August 23, 2026
Altsets
Research by Altsets Research
Use the supply-chain graph to find candidate alternatives, then filter by product function, qualification, capacity, geography, and existing customer relationships.
Data used:Altsets Supply Chain Intelligence: 90k+ entities, 400k+ relationships, 20+ years of history.
Key findings
- A shared customer does not establish functional substitutability because suppliers connected to the same manufacturer can perform very different technical roles.
- Relationship size and existing customer connections can prioritize replacement candidates only after a valid product or process substitute set is defined.
Finding replacement suppliers is not the same as finding every other company that sells to the same customer. A supply-chain graph can identify candidate alternatives, but a credible replacement screen has to separate network proximity from functional substitutability. The Micron supplier network makes this distinction clear. ASML, Lam Research, Applied Materials, KLA, and Shin-Etsu Chemical all appear upstream from Micron, but they do not all perform the same role. Their displayed relationships include:
| Supplier to Micron | Relationship size | Micron cost percentage |
|---|---|---|
| ASML | 3B | 11.91% |
| Lam Research | 1.2B | 5.52% |
| Applied Materials | 990M | 3.84% |
| KLA | 693M | 2.84% |
| Shin-Etsu Chemical | 59.1B JPY in the broader network view | 1.34% |
The first conclusion is therefore a negative one: a shared customer does not make two suppliers substitutes. That is why replacement-supplier research needs several filters.
Start with the disrupted function
The first question should not be, "Who else supplies Micron?" It should be: What exact product, process, service, or manufacturing function is disrupted?
Semiconductor equipment, process control, silicon wafers, chemicals, packaging, and other inputs can all sit upstream from the same manufacturer while being functionally unrelated. A replacement search that ignores the product layer can return a long list of useless candidates.
Use the graph to find the candidate universe
Once the disrupted function is known, the relationship graph becomes useful. It can identify other suppliers already connected to the customer, suppliers connected to comparable customers, companies operating in the same input category, geographic alternatives, and relationships large enough to indicate existing commercial scale. The graph narrows where to research. It does not decide that a supplier is qualified.
Existing customer relationships are especially useful
An existing relationship with the affected customer can be an important signal. If a candidate already supplies the customer in another product category or facility, there may be an established commercial relationship, procurement process, or operational interface. But that still does not prove the candidate can replace the disrupted input.
Product qualification remains specific. The safest conclusion is that an existing relationship can improve the candidate's research priority, not that it proves replacement capacity.
Relationship size can screen for commercial scale
A candidate with an existing multi-billion-dollar relationship may deserve different attention from a company with only a small or unquantified connection. Relationship size can therefore be used as one filter for commercial scale. Customer cost percentage adds another perspective by showing how much of the customer's cost base is associated with the relationship.
Neither metric answers spare capacity, technical compatibility, qualification status, lead time, pricing, intellectual property, or facility availability. Those questions have to be researched separately.
Why Micron's suppliers are not one replacement pool
The Micron example is useful because it prevents a common graph-analysis error. These suppliers all participate in semiconductor manufacturing, but they occupy different technical roles. That means a disruption at one supplier cannot be modeled by simply reallocating its relationship value across the other names.
A valid substitute set should be defined at the product or process level. The relationship network should then be searched inside that functional set.
A stronger replacement-supplier workflow
- Identify the disrupted supplier.
- Identify the exact product or process affected.
- Define the technical substitute category.
- Find other suppliers in that category.
- Check which candidates already connect to the affected customer.
- Check relationships with comparable customers.
- Add relationship size as a commercial-scale signal.
- Research qualification, spare capacity, geography, and lead time.
- Separate "candidate" from "qualified replacement."
This prevents the graph from producing false confidence.
How the economic metrics help after the candidate set is valid
Once the substitute set is technically plausible, relationship metrics become more useful. For each candidate, an analyst can compare existing relationship size with the affected customer, customer cost percentage, relationship history, customer breadth, and geographic presence. A candidate already serving the customer at scale may deserve a higher research priority than a technically similar company with no visible relationship. That is still a probability judgment, not a guarantee.
Replacement suppliers and disruption analysis are different stages
The supplier-disruption exposure guide asks which customers deserve attention after a disruption. Replacement-supplier screening begins after that. First identify the exposed customer.
Then investigate whether alternate supply exists. Keeping those stages separate makes the analysis easier to audit.
A bounded conclusion
Supply-chain data is valuable for replacement-supplier research because it shows existing commercial relationships and economic scale. It is not enough to establish substitutability by itself. The strongest replacement screen combines the network with product taxonomy, qualification evidence, capacity, lead times, geography, and customer-specific technical requirements.
For relationship definitions and missing-value treatment, read the Altsets supply-chain data methodology. Browse Supply-Chain Data Use Cases for other methods.