Frequently Asked Questions

What is a cluster? Why are some companies in the same cluster?

Given our data comes from a number of sources, we discover supplier relationships at different levels of granularity. For example, a source might tell us that a product is shipped to a subsidiary of a global company and another source may tell us only that a relationship exists in some form between the common name of two companies. This challenge means that in order to ensure our data is well connected we group related companies into clusters and nominate a company to represent that group. This group is called a cluster.

When using /{vid}/cluster_suppliers/ or /{vid}/cluster_customers/ with a vid, relationships will be returned which connect to vid and all of the other companies in the same cluster as vid (these companies are grouped by the cid, the cluster id)

When using /{vid}/suppliers/ or /{vid}/customers/ with a vid, only relationships connected to vid will be returned. This gives a more granular view of the connections to and from a company.

Why are there more relationships in /{vid}/cluster_suppliers/ vs /{vid}/suppliers/? Which one should I use?

Since the {vid}/cluster_suppliers/ endpoint returns relationships to companies in the same cluster, it will return relationships connected to the parent company and its subsidiaries and this will yield the most results.

If you are interested in the relationships between a specific company and its suppliers, you should use the {vid}/suppliers endpoint

What is evidence?

Evidence is the term for unique pieces of information we find across a number of sources that supports relationships between companies. For example, if we find a company listed as a supplier on a website, and we also find a company listed as a customer on a website, we can infer that the two companies are related. This relationship is supported by the evidence we found on the websites.