Unused SaaS licenses: audit and reclaim that spend

A 2026 market synthesis measured what no CFO wants read aloud: a large company uses, on average, about 54% of the SaaS licenses it pays for. The rest is provisioned, invoiced, and parked. The same reading prices the leftover at close to USD 9.8 million per company per year. This is not visible waste. It is silent, because the invoice arrives, the line matches the contract, and finance approves the recurrence without seeing usage on the other side of the counter. No one lies and no one hides. The money vanishes because no one confronts what is paid against what is running.

Worse than the idle license is the direction of the spend. The software portfolio grows 34% to 37% a year, with about nine brand-new applications entering per month and double that in companies above ten thousand employees. Roughly 87% of applications and 85% of spending are decided outside IT, in the direct balance between the business unit and the vendor, and about 26 applications carry duplicated spend across channels. This article treats a specific mechanism that ties those ends together: audit the current base of already-paid subscriptions, find the idle license and the duplicate purchase, and reuse that budget without clearing a path to a new vendor every cycle.

What an unused SaaS license is and why it is the base's invisible cost

An unused SaaS license is not the tool a company decided not to have. It is the one it bought, renewed, and keeps paying, but that the team stopped using, used halfway, or provisioned in quantity above real need. The distinction matters because it changes the kind of decision. Canceling a tool with no owner is cutting a cost no one defends. Recognizing that a tool used by forty people is contracted for one thousand is reprovisioning against real usage, and that does not require giving up any critical function.

The invisible cost lives in the recurrence. A paid and idle application does not show up as an error in any accounting report, because the spend is budgeted, legitimized, and predictable. It hides in the gap between provisioned and usage, a figure finance almost never receives together with the invoice. The invoice says how much is paid. The vendor knows how many licenses are active. No one, on the default path, puts those two columns side by side per line item. It is that void, not a one-off slip, that turns a large company's SaaS base into inventory of paid, dormant capacity.

The austerity paradox is that no large company contracts that idleness on purpose, yet every incentive produces it. The purchase is born from a unit's need, grows past the first project, and the license is left over when the need changes shape or team. While no one is formally the owner of the question "are we using what we pay for?", the default answer will be yes, paid, however much reality says otherwise. The audit of the base begins when someone becomes the owner of that question over everything already contracted.

Where software spending veers off governance

Three deviations explain nearly all the idle licensing in a large company, and all three come from the same place: the purchase happens far from whoever governs. Understanding where the money escapes is the prerequisite for auditing the base and reusing the budget, because each deviation calls for a different correction.

The first deviation is the decision outside IT and procurement. The 2026 numbers point out that 85% of software spending is decided by business units, and 87% of applications come in through that path. The unit wants to solve an operational problem, finds a tool that solves it, closes the subscription with the vendor, and puts the cost on its own account. Legitimate for whoever needs the result. Fatal for the portfolio, because that purchase never passes a list of what already exists, so the new application arrives without anyone asking whether a contracted one does the same job. Part of the duplication is born exactly here.

The second deviation is duplication across channels. One capability contracted in two or three places: an application bought direct from the vendor, the same function inside a corporate suite, and still a third point the unit provisioned without looking at the first two. The 2026 synthesis spreads the number of about 26 applications with duplicated spend across several channels. Duplicating is not a user's whim, it is the product of the absence of a single place that shows the whole base. Without portfolio visibility, each unit repeats the others' errors out of ignorance, not stubbornness.

The third deviation is growth without an inventory. A portfolio that expands every year and is never consolidated becomes a debt that grows faster than any one-off savings effort compensates. The problem is not growth. It is growing without a ruler that warns when the new stacks on top of the already paid and without an owner for the whole account. Auditing the existing base is the only way to stop accumulating over blindness.

How to audit the idle SaaS subscriptions of the current base

Auditing the current base is a four-step process that the procurement team runs over what is already contracted, without depending on any new system to start. Each step confronts one piece of data that the default path never crosses.

1. Inventory what is paid against what is provisioned. The first step is to build the base list: every active subscription, the annual value, the number of contracted licenses, and the channel it was bought through. That is the provisioned column. On the default, many companies discover at this moment that they do not hold such a list ready, because the spend lives scattered across unit invoices and different vendors. The inventory is the price of entry, and it exists only when someone decides to consolidate.

2. Set the provisioned against real usage. For each inventory item, the next figure is actual usage: how many licenses really access the tool, how often, by which teams. It is this confrontation, and no other, that reveals the unused SaaS license. A subscription contracted for one thousand users and used by four hundred is not a price problem, it is a reprovisioning problem, and the value left over is exactly the disconnect between the two columns. The data comes from the tool's own administrative screens, which the administrator can read before any automation. Measuring usage at the time of the audit is what makes the decision happen against a real number, not against the memory of the contract.

3. Sweep for duplication across channels. With the base consolidated, the next step is to look for the same capability repeated in more than one contract: the dedicated tool that repeats a suite's function, the direct-bought application that replicates what another already covers. Each duplicated pair is a license candidate for cut or consolidation. The sweep looks at function, not at the application name, because it is the overlapping function that sustains repeated spend.

4. Check licenses purchased beyond usage. Finally, examine licenses provisioned above need, including the repositories that grew when the team was larger and were never reduced. Separating what legitimizes a seasonal peak from what is parked capacity avoids cuts that break the business and keeps the cut where it is safe.

The result of the four steps is a base reorganized into three groups: the in-use that gets reaffirmed, the idle that is ended or reprovisioned, and the duplicated that is consolidated. The audit decides nothing on its own. It produces, for the first time, the map on which a decision becomes possible.

The table below compares the default path with the governed process, for a one-page read.

How to reuse and reclaim the freed software budget

The second half is reusing the budget the audit released, and here the sequence guards against the error that undoes everything: climbing to a new vendor every time a cost is cut. Spending idle license money on another immediate novelty does not fix the base, it only trades one idle license for another in formation. The discipline has three moves, and the three follow the intention of reuse rather than the intention of new purchase.

The first move is to reaffirm only what is in use. Each subscription in the in-use group earns a conscious term, with volume adjusted to real usage and the value reaffirmed because it was decided, not because it was always there. Reaffirming is an act of decision that turns a spend recurring by omission into a spend maintained by judgment. Reaffirming becomes deciding. That is the step the SaaS procurement policy guide for LATAM companies complements on the rules side, by designing the regime in which the reused budget re-enters without turning into fresh disorder.

The second move is to end or renegotiate the idle. A paid and parked license can breathe in two ways: ended, when the capability is no longer needed, or renegotiated to the real volume, when the tool stays useful at a smaller scale. In both cases the value is released, and in the ending it is necessary to make sure the exit does not drop a critical service for lack of continuity. Part of that is settled by the contract design, above all when the subscription renews on a cycle with an exit notice period, which belongs to the renewal cycle governance, a distinct mechanism that governs the deadline, not the audit of the base.

The third move is to reapply the released value within the scope the audit already prioritized. The budget that stops paying for idle and duplicated returns to the software account of what the company really uses and needs to grow, without a leap into a new scattered purchase. Reapplying is the opposite of reinvesting by reflex: the value that would renew by inertia or replicate in a duplicated channel instead funds the base's real capacity. It is this reallocation that carries the spend back to whoever decided to govern it.

Why software spending needs to return to the single governance point's radar

The audit steps and the destination of the value depend on a background condition the default path does not offer: a single place where the entire spend is visible by item and decided against. Without that point, the inventory is redone from zero each cycle, the duplication reappears the following quarter, and the spend decided outside IT slips back in through the back door. Governance does not fail for lack of willingness in the units. It fails because there is no place where the whole base can be listed, confronted, and judged.

That is the function of the Nexforce Marketplace seen from the buyer's side: the single point of procurement and of spend governance over the live base. For that reason, more than a choice of where to buy, it is the context in which the subscription comes to live and be governed, a discussion the piece on direct purchase or software marketplace locates precisely. Over the current base, the single point delivers six conditions that sustain the audit and the reuse.

Item-level portfolio visibility, with centralized accounting and governance that put each subscription and each value in one account, is the condition for finding what is provisioned beyond usage. Software access and usage control sets what is paid against what is in fact active, and that is where the question about the license no one uses is born, because the instrument crosses the two columns that the default keeps apart.

The renewal and budget alert brings the spend data onto the radar before it turns into inertia, so an underused license stops repeating in silence and starts to demand a decision. Centralized contracting and a negotiated structure reduce, on the live base, the cost of international solutions in the contracting and renegotiation layers, and they give the reuse a channel to reapply the released value in a contract that the default would leave scattered and duplicated.

The local invoice, the nota fiscal in reais, and the currency hedge take the international slice of the base out of dollar exposure, and the domestic NF lets a company in Lucro Real recover the 9.25% PIS/COFINS credit on that account. That credit holds while the PIS/COFINS charge lasts, in transition to the CBS and the IBS from 2027 (LC 214/2025). For a company in Lucro Presumido the benefit is not the credit. It is the nota fiscal in reais, the currency hedge, and the operational simplification. The sixth condition is the governance the default does not give: bring back to procurement's radar the spend decided outside IT so the power to reclaim budget is restored. Without a single point, the audit is a periodic exercise. With it, it becomes the routine that stops the base from regressing. That point is the ruler of the spend.

Frequently asked questions

What is an unused SaaS license?

An unused SaaS license is a software seat or subscription the company keeps paying that is not in real use, whether because the team stopped using it, used it halfway, or contracted it in quantity above need. The cost sits in the recurrence: renewing and paying for capacity that does not operate. Paid, though parked. It only becomes a decision when provisioned is set against usage.

How do I know which SaaS licenses the company does not use?

By setting what is provisioned against what is in real use. The path is to inventory every active subscription, check each tool's administrative screen for how many licenses actually access it, and cross that figure with the amount paid. The idle license appears in the gap between the two columns, and duplication appears when the same function lives in more than one contract. Real data, not assumption.

Is auditing the SaaS base an IT or a finance task?

It is a joint procurement, finance, and IT leadership decision, because it involves whoever measures usage, whoever approves the value, and whoever inventories the portfolio. The single governance point exists so those three views meet over the same base. No party decides alone over what is already paid. One base, three views.

Does reclaiming an idle license force cutting the tool?

Not necessarily. The idle license can be ended when the capability is no longer needed, reprovisioned when the tool remains useful at a smaller scale, or renegotiated against the drop in usage. The released value returns to the software budget without requiring the company to give up any critical function.

What do I do with the money freed by the audit?

Reapply it to the base's real, prioritized capacity instead of climbing to a new vendor every cycle. What is in use is reaffirmed, what is idle is ended or renegotiated, and the value is reapplied to what the company uses and needs to grow, within the procurement policy already designed. It is reuse, not a reflex new purchase. Reused, not multiplied.

References and further reading

The market data cited throughout the text as corroboration of the direction of spend comes from 2026 syntheses on corporate software usage and subscription management, treated here as neutral figures of the buyer mechanism and not as a tour of any discovery tool. For the regime in which the reused budget re-enters under rule, consult the piece on SaaS procurement policy for companies in LATAM. For where the subscription comes to live and be governed through a single point, see the piece on direct purchase or software marketplace. As to the credit cited, note that PIS/COFINS transitions to the CBS and the IBS from 2027, under LC 214/2025.

Take the map of idle licenses to the next procurement meeting

The audit of the base does not start with an imposed system. It starts with the step that costs least and produces the map: consolidate in one list how many SaaS subscriptions renew today and, for each one, separate provisioned from what is in real use. Each line of the map points to the deviation it carries, the paid and idle license, the duplicated application, the spend that entered outside IT, and the value the audit releases without opening a new vendor.

The trigger is the same for the whole base. What is in use is reaffirmed with adjusted volume. What is left over is handled as idle or duplicated, and the released value returns to the software budget to fund the capacity the company uses. The sequence repeats each cycle, and it is the repetition, not the one-off effort, that stops the base from regressing.

The Nexforce Marketplace serves this need on the buyer's side: the single point of procurement and of software spend governance that gives the buyer item-level portfolio visibility, software access and usage control, renewal and budget alerts, and centralized contracting with a negotiated structure. The international slice leaves the dollar with an invoice in reais, a nota fiscal in BRL, and a currency hedge, and the domestic NF lets whoever is in Lucro Real recover the 9.25% PIS/COFINS credit. Taking the map to the meeting and deciding by it which single point the company adopts is the concrete gesture that turns the license no one uses into the first item the buyer reuses.