What is a CAGE Code?
A CAGE Code — Commercial and Government Entity code — is a five-character alphanumeric identifier for a specific legal entity at a specific physical location. It is assigned automatically when an entity completes federal registration, and it costs nothing.
Where the Unique Entity ID answers “who is this organization,” the code answers “which entity, operating from where.” That distinction is why both exist and why a single company can legitimately hold several codes.
Where it gets used
The code has a longer history than most of the modern federal identifiers and it is embedded in a wider range of systems:
- Contract and payment records, alongside the entity identifier.
- Logistics and supply systems, where it identifies the source of a particular item from a particular place.
- Facility security records, where clearances attach to an entity at a location rather than to a company in the abstract.
- Pre-award checks, where a contracting officer confirms the specific entity a proposal came from.
For a pure software company, the logistics uses rarely come up. The security and pre-award uses do, usually at an inconvenient moment.
One company, several codes
Because the code is location-bound, a company operating from multiple locations can hold multiple codes. That is normal. It becomes a problem only when nobody inside the company knows which codes exist.
The recurring failure is a company that has acquired another, inherited its codes and registrations, and never reconciled them. Award history and facility records sit under codes the current team does not track, which makes the company’s own federal footprint harder for it to describe than for an outsider to look up.
Keeping an internal list of your entities, their identifiers, and their codes is ten minutes of work that saves days during a proposal.
Foreign entities
A company without a domestic presence obtains the international equivalent through its own country’s codification authority rather than through the domestic registration path. The concept is identical — an entity at a location — and the code interoperates with federal systems.
This matters to non-domestic software companies more than they expect. Selling into the federal market from abroad is possible, and the identity plumbing is the first thing to sort out, because nothing downstream works without it.
The practical advice
Three things are worth doing once:
- Find out what codes you already have. Registration assigns one automatically, and companies routinely do not know theirs.
- Reconcile inherited entities. After any acquisition, list the entities, identifiers, and codes that came with it and decide which stay active.
- Treat an address change as a records project. Because the code is location-bound, moving offices touches more systems than the registration itself.
What it does not tell you
The code carries no qualitative meaning. It does not indicate size, capability, past performance, or approval. It is a pointer to an entity at a place — nothing more — and any inference beyond that is being made by the reader rather than by the government.