Every conversation about an expensive classified environment starts the same way. Someone produces the annual figure. Everyone agrees it is large. Then the conversation goes looking for things to cut.
That is the wrong denominator.
An air-gapped estate is provisioned for an organization. It is used by a subset of that organization. The distance between those two numbers is usually large, and it is usually nobody's job to measure it.
The number nobody computes
Take the annual cost of the estate. Divide it, not by the seats you provisioned, but by the number of people who genuinely get work done inside it in a normal month.
That figure is often startling. It is also the honest one, because it is the only version that describes what the budget line is actually buying.
You do not need a study to estimate it. The signals are already in your environment:
- Seats provisioned against concurrent logins over a quarter.
- Work that gets done somewhere else and transferred in, rather than done in place.
- A queue for something that ought to be self-service.
- Capability requests that were abandoned rather than fielded.
If those signals are present, the estate is not serving the organization it was sized for. It is serving a core of people who have learned to live with it.
Why this reframe matters more than it sounds
Because it changes what kind of problem you have.
A high total cost invites a cost-cutting exercise. Cost-cutting exercises against an estate like this fail, for reasons this series will get to: almost none of the cost scales with how much of it you use.
A high cost per productive user invites a different question, which is what the money should buy instead. That is a design question, and design questions have answers.
The two framings also land differently with the people who control the budget. "This is expensive" is a complaint. "This serves a sixth of the organization it was built for, and the same money in a different shape would serve most of them" is a proposal.
The question underneath, which is the better one
Low utilization has two possible causes, and they lead to opposite conclusions.
Friction. The environment is hard enough to get into, hard enough to get data into, and slow enough to change that people route around it. Demand is real, and it is being suppressed.
Demand. Those people genuinely do not need a high-side environment to do their jobs. The estate is correctly used by the people who use it, and everyone else was never a real user.
These call for entirely different responses. The first says the environment is the bottleneck and fixing it releases work. The second says the estate is over-scoped and the organization needs a different tier for the majority of its work.
Here is the uncomfortable part: in most organizations, nobody knows which one is true. The utilization data exists, the reasons do not, and the question has never been put to the people who stopped using it.
That is answerable in a week of conversations, and it is worth more than any infrastructure analysis you could commission. Ask twenty people who have access and do not use it why not. The answers sort themselves into those two buckets very quickly.
A note on numbers, for the whole series
This series carries no cost figures.
That is deliberate. We have modeled estates of this shape in detail, and the modeling is useful, but a dollar figure that describes someone else's estate is worse than no figure at all when it lands in front of a program office. It invites an argument about whether the number is right instead of a conversation about whether the structure is.
What generalizes between estates is the structure: where the cost sits, which levers move it, and which ones do not. Those findings survive being wrong about the parameters. The dollars do not, and they belong to whoever owns the estate.
So the rest of this series is about mechanisms.
Robert Burckner is the founder of Millabs Corporation, a Service-Disabled Veteran-Owned Small Business. He has served as ISSM and ISSE for legacy weapon systems and as Division Chief at the Space Warfighting Analysis Center (USSF/NRO).