Every organization that manages engineering software licenses makes decisions. The question is what those decisions are based on if you don’t have evidence-based software usage visibility. For most organizations, the honest answer is: assumptions derived from headcount, previous purchase history, and the loudest complaints from engineering teams.
That is not a criticism of the people making those decisions. It is a structural observation about what happens when usage data is not available. And it is the norm, not the exception: Flexera’s 2026 State of ITAM Report found that only 36% of organizations have complete visibility into their IT estate, leaving the majority deciding on partial insight. Assumptions fill the gap. They feel reasonable in the moment and produce outcomes that feel acceptable until someone does the math.
When you can’t see usage, assumptions fill the gap. This piece explores how software usage visibility closes it, beginning with the assumptions most organizations don’t realize they’re making.
Software Usage Visibility: The Assumption Inventory
Before addressing what evidence-based software license management looks like, it helps to inventory the specific assumptions most organizations are currently running on.
“We need at least as many licenses as we have engineers using the tool.”
This assumption treats software license need as a function of headcount. The reality is that license need is a function of concurrent demand – how many engineers need the tool simultaneously, not how many engineers use it in a given month. Peak concurrent demand typically runs below the total user count for any given tool, meaning organizations that license based on headcount are carrying structural excess capacity as a direct consequence of how they procure.
“If engineers are not complaining about access, the license pool is fine.”
Denial events that resolve quickly often never generate a formal complaint. The engineer queues for a few minutes and gets access, accepting the friction as normal. Meanwhile, the denial table is accumulating a pattern that, if analyzed, may show 20 to 30 denial events per day concentrated in a two-hour window. The point being is absence of complaints is not evidence of adequate provisioning. It is evidence of tolerance.
The point being is absence of complaints is not evidence of adequate provisioning. It is evidence of tolerance. And that tolerated friction carries a real cost, one we’ve traced in detail in how poor management of license assets affects engineering productivity.
“We buy extra licenses to avoid audit risk.”
Over-purchasing as a compliance strategy is common in large engineering organizations. The flaw is that over-purchasing does not actually establish compliance – it just reduces the probability of finding a gap. And it does so at full annual license cost, rather than through accurate deployment tracking that would provide genuine audit confidence at a fraction of the price.
Software Usage Visibility: What Evidence-Based Actually Looks Like
The shift from assumption to evidence is about giving human judgment better inputs, not replacing it. Not only do these data points provide software usage visibility, they also equip you to make strategic licensing decisions.
Concurrent demand data replaces headcount assumptions. When an organization knows from 90 days of concurrency history that their peak concurrent demand for a specific tool runs at 42 users out of a 60-seat pool, they have a basis for a renewal conversation that headcount data cannot provide. The vendor proposes renewing 60 seats. The utilization data supports renewing 48 with a small buffer. The difference, at $20,000 per seat (an illustrative figure based on enterprise CAE pricing ranges), is $240,000 in avoided annual spend.
Denial frequency data replaces the complaint assumption. An organization that runs denial monitoring knows how often engineers are blocked, for which tools, during which hours. This data identifies whether the problem is genuine capacity shortage or checkout duration, a distinction with very different remedies.
Continuous deployment tracking replaces over-purchasing as audit strategy. An organization that maintains an accurate, current record of what is deployed against what is licensed has genuine audit confidence. When a vendor initiates a review, they respond with documentation rather than reconstruction.
The Goodhart’s Law Problem in License Management
Goodhart’s Law states that when a measure becomes a target, it ceases to be a good measure. In license management, the common measure-as-target is license count. Organizations set their license count as the target and then optimize for hitting that number rather than for the underlying goal: ensuring engineers have access to the tools they need at a cost the organization can justify.
Usage data replaces count-based thinking with demand-based thinking. The target becomes: ensure access at peak demand with acceptable buffer, at minimum necessary cost. That target produces different decisions, because it is evaluated against actual utilization rather than a headcount estimate. It happens when you don’t have to assume usage and have complete software usage visibility over your software license pool.
A Glance at LAMUM
LAMUM provides the usage, concurrency, and denial data that replaces the assumption inventory above with actual evidence. The gap between what you assumed and what the data shows is usually the most compelling business case you will ever need to make. LAMUM is one unified license asset manager to help you see, measure, optimize, and reduce the cost of engineering software across every tool, team, and renewal cycle.
Get full software usage visibility in a single unified platform.
Explore LAMUM!
