Guide
Reducing licensing cost almost always means calling the vendor. Negotiation is one lever of three, and rarely the one with the most money behind it.
When a business decides to reduce software licensing costs, it almost always means one thing: someone is going to call the vendor.
Negotiation is a real lever and it is worth pulling. It is also the first of three, and in our experience it is rarely where most of the money is.
The other two require no vendor conversation at all. They require knowing who is actually using what — which sounds simple, and is the reason they do not happen.
The familiar one. Renewal comes round, you push on price, you may get a discount tied to term length or commitment.
What it can do: reduce unit price. What it cannot do: fix the fact that you are buying the wrong quantity, or the wrong tier. A discount applied to over-provisioned licensing is a smaller bill for the same waste.
Licenses accumulate. Somebody joins a team that was provisioned for twelve people and grew to eight. A project ends and its tooling stays live. An employee leaves and their license is not reclaimed, because reclaiming it is nobody's Tuesday priority.
Right-sizing means comparing the licenses you pay for against the licenses being used — not assigned, used. Establishing that takes three sources joined up: the active user list, the analytics the platforms produce, and the log files underneath them. The first tells you who exists. The second tells you who signs in. Only the third tells you whether somebody is using the features they are licensed for.
Most platforms sell in tiers, and most organizations provision everyone at the same tier because uniform provisioning is administratively simpler. Somewhere in there is a population of users on a premium license who use none of the premium features.
Correcting this means looking at feature usage rather than seat count, then moving people to the tier that matches what they actually do. It is fiddly. It is also money that does not require a vendor conversation, a contract change, or anyone losing a capability they were using.
One example, from a recent review
120 users provisioned with a Microsoft Teams Phone Standard license, where the actual telephony usage belonged to an IT support team of ten. Nobody had made a bad decision. Somebody had made one decision, for everybody, at a point when working out who needed it would have taken longer than provisioning it.
Finding it took an afternoon with usage analytics from the Teams administrator. It did not take a tool, a consultant or a vendor. It took somebody asking a specific question and then reading the answer.
120
licenses provisioned
10
people actually using them
~10%
of the total saving from right-sizing alone
Everything above assumes you can cut freely. In healthcare, financial services or pharmaceuticals, you cannot.
Some licensing carries security, compliance or regulatory function — audit logging, retention, encryption, access controls, the evidence you need when someone asks you to demonstrate a control. Those features often sit in exactly the higher tiers that a naive right-sizing exercise would target.
A license that looks idle because nobody logs into it may be doing its job precisely by existing.
So the rule we work to: nothing that carries a control gets cut or downgraded, regardless of usage data. This makes the exercise slower and the savings smaller than an unconstrained version would produce. It also means the savings survive the next audit, which is the only kind worth having.
The constraint above has a softer version that catches people more often.
We have reviewed estates where two overlapping collaboration platforms were both licensed, and one of them was barely touched. On a usage report it looked like an obvious cut. It was not. The teams using it needed it for exactly the situations where ease of access mattered most, and the frequency of use told you nothing about what depended on it.
Asking about criticality rather than frequency is what surfaces this, and it tends to surface other things too. In that review, the conversation about why the tool was kept turned up a configuration that did not meet the regulatory requirements of the environment it was running in — a finding that had nothing to do with cost and would not have appeared in any usage report.
The practical rule: a usage figure tells you how often something is opened. It does not tell you what stops working if you take it away. Only the people who use it can tell you that, and the cost of asking is one short conversation.
Not ignorance. Three practical obstacles.
The data is not assembled. Usage information exists, spread across admin consoles, platform reports and log files nobody has joined up. Someone has to do that before any decision is possible.
It is nobody's job. Negotiation has an owner and a date — the renewal. Right-sizing has neither. It is important and never urgent, and it loses to everything urgent.
It needs a view of the whole estate. Individual teams see their own licensing. The duplication is visible from above, and in most organizations nobody is looking from there.
That order matters. Negotiating before you know what you need means negotiating a good price for the wrong thing.
The single biggest reason licensing savings do not last is that the exercise is treated as a project. Estates drift back. People join, projects end, somebody provisions a team at the default tier because that is still the simplest thing to do.
What we run instead is a rhythm. At six months, a check: license consumption against active users. It is a data point rather than a project, and it takes very little effort because the comparison was already built once.
Then, ahead of the final quarter of the year, the full analysis runs again — all three levers, in the same order. Doing it there is deliberate. It puts the answer in hand before renewal conversations start, which means lever one is pulled with a quantity and a tier that are already correct.
The point
Licensing cost is not one number to be argued down. It is a price, a quantity and a tier, and most organizations only ever argue about the price.
The other two are analysis rather than negotiation, which makes them less satisfying and usually more valuable.
Written by
Optimization Lead, Modern Signal
Modern Signal's optimization practice, running licensing and cost reviews across regulated estates.
Where this came from
A healthcare provider
A healthcare provider's technology estate had grown by addition, with every business unit solving its own problem. We took over running it, then rebuilt what duplication had made expensive.
Read the case study: Entered once, not three times — 70% faster RCM workflows and 35% off licensing~70%
Faster revenue cycle workflows
Tell us what it looks like today and we will tell you what we would look at first.
Nothing here is required except an email address and enough detail to reply usefully. The more you tell us, the better the first answer.
