Head of Marketing at 10Duke.
Connect with me on LinkedIn.
Most software companies have spent the last decade building out analytics stacks. Product usage tools, CRM dashboards, support ticket analysis, marketing attribution. All of it aimed at answering the same question: what are our customers actually doing, and what does it tell us about where the business should go next.
Meanwhile, sitting quietly wherever license checks happen, is a dataset that answers that question more directly than almost anything else in the stack: license usage data.
Every time a customer activates a seat, hits a usage limit, checks out a floating license, or gets denied access because they've exceeded their entitlement, that's a signal. Not a proxy for behavior, like a marketing click or a support ticket sentiment score, but the behavior itself. For desktop and on-premise software vendors in particular, whose products often don't sit inside a web analytics pipeline the way a SaaS app does, license usage data is frequently one of the few first-party records, and often the hardest to evade, of how the product is actually being used in the wild.
The trouble is that most vendors never see it that way, and for many the reason is structural. If licensing was built in-house, often as a validation routine bolted onto the app or an activation table added to a database years ago, it was designed to answer one question only: is this user allowed in or not. Nobody designed it to be queried, aggregated, or reported on, so the events happen, but they disappear the moment the check completes. The signal is real, but it's effectively unreachable without licensing infrastructure that was built to capture and structure it in the first place. That's the real cost of treating licensing as plumbing rather than intelligence: not that the data doesn't exist, but that most systems make it impossible to get to. It's the same technical debt that quietly stifles growth in other ways, this is simply the data side of it.
A better revenue signal than most CRM data
Sales and customer success teams spend a lot of effort trying to guess which accounts are ready to expand and which are quietly at risk of churning. Usually that guesswork is built on lagging indicators: a renewal date approaching, a support ticket spike, a check-in call that didn't go well.
License usage data is a leading indicator. An account consistently bumping against its seat cap or usage ceiling is telling you, unprompted, that it's ready for an upsell conversation before anyone on the account team has noticed. An account whose active usage has quietly declined over two quarters is telling you it's at risk, long before the renewal conversation gets awkward. This is the kind of signal that's hard to fake because it's derived from actual product consumption rather than self-reported sentiment.
It’s a more honest product roadmap input than customer interviews
Every product team knows the gap between what customers say they want in an interview and what they actually do with the software. Feature requests are loud. Feature usage is quiet. License and entitlement data closes the gap, but only when it's granular enough to show which modules or capabilities are actually being checked out, not just whether someone was let into the application at all.
It shows you the features nobody uses despite being requested loudly in every sales call. It shows you the quiet workhorse feature nobody talks about but that shows up in nearly every session. For product teams trying to prioritize a roadmap against limited engineering capacity, this is closer to ground truth than almost any other input available, and it's already being generated as a byproduct of normal licensing operations rather than something that needs a new research budget to produce.
It turns support from reactive to anticipatory
Support teams are almost always working from the point a customer complains backward. Usage and entitlement data flips that. Failed activations, unusual denial patterns, a validation check suddenly firing far more often than it used to for one account: these are early warning signs of a problem before it becomes a ticket. A vendor that can see a customer's licenses failing before that customer emails support has a materially different relationship with that account than one that finds out when the complaint lands.
The reframe worth making
For a long time, licensing was thought of as a gate: it lets the right people in and keeps the wrong people out. That framing undersells what's actually happening every time a license check runs. Each of those events is a small, structured, honest record of real usage, and most companies generate them constantly without ever looking at what they add up to.
As more ISVs shift toward consumption-based and hybrid pricing models (especially in the new era of AI tokens), this becomes less optional. You can't price on usage you aren't measuring, and you can't have an informed conversation about expansion, retention, or roadmap priority without the data that shows what customers are actually doing with what they've already bought. The companies that get ahead here will be the ones that finally looked at the data they were already sitting on.
If you're curious what this looks like once the data is actually surfaced, that's exactly the gap 10Duke Insights was built to close, turning license usage into revenue, product, and support intelligence you can act on.
