ServiceNow Software Asset Management Professional KPIs and Metrics That Matter

The SAM Professional KPIs on this page are the measures that show whether software licensing is under control, such as compliance, cost per user, utilisation and renewal readiness. The page also covers how to define each measure so that a figure keeps its meaning across reporting periods.

License Compliance Metrics

Definition: percentage of installed software that is licensed. If you have 100 copies of an app running and are licensed for 80, compliance is measured. Example: Oracle Database compliance = (licensed count / installed count) * 100.

How calculated: compare count of licensed installations from SAM module with actual installed count from discovery or vendor reports. Discrepancy indicates unlicensed use or over-license.

Typical misuse: conflating license seats with actual usage. You may be licensed for 100 seats but only 60 are used. This is over-licence, not under-licence. These are different problems with different solutions.

Review cadence: monthly or quarterly. More frequent review catches compliance drifts early.

Software Cost Per User

Definition: total software spending divided by number of users. If you spend one million dollars on software and have 1,000 users, cost per user is one thousand dollars. Example: SaaS application cost per user = annual subscription cost / number of users.

How calculated: sum all software costs (licenses, maintenance, support) divided by active user count from the HR system or user database.

Typical misuse: ignoring seats that are allocated but not used. If you have 100 Slack seats but only 80 active users, cost per active user is higher than cost per seat.

Review cadence: annually for budgeting. Quarterly trending to spot cost creep.

License Utilisation Rate

Definition: percentage of licensed seats actually being used. If you have 100 seats and only 75 are active, utilisation is measured. Example: CAD software utilisation = (daily active users / licensed seats) * 100.

How calculated: pull active user counts from application logs or usage monitoring. Compare to licensed seat count from SAM. Low utilisation indicates over-licensing.

Typical misuse: assuming inactive users will return to activity. If a user has not logged in for a set period, they are likely not returning. Optimise license allocation for active users.

Review cadence: monthly. Quarterly for trend analysis.

Software Spend to IT Budget Ratio

Definition: software spending as a percentage of total IT budget. If IT budget is ten million dollars and software spending is three million dollars, ratio is measured. Example: SaaS as % of IT budget = annual SaaS spending / total IT budget.

How calculated: sum all software costs (on-premise licenses, SaaS, cloud, open source support) divided by total IT operational budget.

Typical misuse: including only software licenses, not maintenance and support. Total cost of ownership is higher than the license price.

Review cadence: annually for budgeting. Quarterly for trend tracking.

Contract Compliance and Renewal Readiness

Definition: percentage of software contracts where you are compliant with terms (license count, usage scope, support level) and up-to-date on renewals. Example: contract compliance = (compliant contracts / total contracts) * 100.

How calculated: audit each software contract for compliance. Are you using the software within license scope? Are you paying for support levels you use? When do contracts renew? Track in SAM module.

Typical misuse: ignoring contract terms and assuming compliance. Many organisations are out of compliance with their license terms but do not know it. Audit regularly.

Review cadence: quarterly or before contract renewals.

Total Cost of Ownership Reduction

Definition: reduction in software spending year-over-year from optimisation. If you spent three million dollars on software last year and two point seven million dollars this year, reduction is measured. Example: TCO reduction = ((spending last year – spending this year) / spending last year) * 100.

How calculated: track total software spending each year. Reductions come from: eliminating unused licenses, renegotiating contracts, moving to open source alternatives, or consolidating vendors.

Typical misuse: claiming reductions without corresponding licence elimination. If you reduced spending by cutting support contracts, you have reduced cost but increased risk. Be honest about trade-offs.

Review cadence: annually for strategic review.

Open Source and License Risk

Definition: percentage of software portfolio where license type is known and compliant with company policy. Example: open source compliance = (software with known approved licenses / total software) * 100.

How calculated: scan installed software for open source components and their licenses. GPL, BSD, Apache, and other license types have different compliance implications. Track which software components have risky licenses.

Typical misuse: assuming open source is free and risk-free. GPL software obligates you to share source code if you distribute it. Not all open source is appropriate for all use cases.

Review cadence: quarterly. More often if you develop software or consume new dependencies.

Each SAM figure has to match the question it was built for

A reporting analyst who inherits a SAM dashboard often inherits a set of figures that look authoritative and cannot be explained. The compliance percentage moved last month, and nobody remembers whether the denominator changed. The utilization chart has a dip that coincides with a discovery schedule change rather than with any real drop in use. These are not calculation errors in the narrow sense. They are definition errors: the metric was specified in words that allowed two different calculations, and the platform implemented one of them without saying which.

The discipline that prevents this is a metric specification that travels with every figure. For each measure, the specification states the question it answers, the exact inputs, the denominator, the time window, the exclusions, and the owner who may change any of those. This part is about building and maintaining that specification for SAM Professional reporting. It does not repeat the definitions of the common measures, but assumes you have chosen them and now need to keep them trustworthy.

A specification sheet for each measure

The table below shows the fields a specification should carry. Filling every field for every metric is tedious, and that tedium is the point: a field left blank is often a question nobody has answered yet.

Measures that move when the data moves and not the estate

Several common SAM figures are sensitive to their own inputs in ways that are easy to misread. A utilization figure falls when an identity feed starts including accounts that were never removed, even though real use has not changed. A compliance figure improves when a normalization rule begins matching a product it previously missed, which shows more licenses in use without any new purchase. A cost figure rises when an invoice is added to the wrong product, not when spending has grown. Each of those inputs arrives through a feed, and the source-of-truth decisions for SAM data determine which change is a real one.

The working rule is to test every change in a headline figure against the input data before explaining it as a business trend. Compare the input counts across the two periods first. Only when the inputs are stable should the change be attributed to the estate. This takes little time and prevents a common embarrassment in a SAM review, which is a leadership conversation built on a data artifact.

Reporting to leadership without overstating precision

Leadership wants a small number of figures that say whether the position is improving, and the governance model should say who signs each one off before it goes in the pack. The temptation is to present every metric with a precise value and a confident trend arrow. A more defensible approach presents three things for each headline measure: the current value, the direction against the previous reconciled period, and a short note on the largest input change that affected it. That note is often the most useful sentence in the whole pack, because it tells the reader which part of the change is data and which part is real.

When a figure is known to be provisional, say so in the report and give the date the reconciliation will be complete. A provisional number that is labeled as provisional is far less damaging than one that is revised a month later without explanation.

A short checklist before any SAM figure is published

1. Confirm the specification sheet for the measure is current and names the owner, using the same recurring routine that runs the monthly close.
2. Compare input counts with the previous period and note any scope, feed or rule change.
3. Check that exclusions match the platform configuration rather than the report filter alone.
4. Write the one-line explanation of the largest movement before presenting the trend.
5. Record the publication in the change log with the reconciliation run it used.

Measures that belong on a working dashboard but not in a board pack

Not every figure that helps the SAM team belongs in front of executives. Exception queue size, the number of unmatched installs and the age of open entitlement corrections are operational measures. They drive the work, and they change daily. Presenting them to leadership invites the wrong questions, and it can make a healthy process look like a failing one. Keep them on the working dashboard with their own owners, and promote only the reconciled, specified measures to the leadership pack.

Common questions about SAM reporting measures

Why does the same license count differ between the platform and the vendor tool?

Each source applies its own rules for installation, edition, device and user. The difference is not a fault until it is unexplained. Keep a short reconciliation table that shows the category of each difference, and review it when either tool changes.

Should we report monthly, or wait for a full reconciliation cycle?

Report on the cycle that the reconciliation actually completes. If the monthly run is provisional until a later check, say that on the figure itself rather than publishing the early number as final.

How do we handle a metric the business keeps asking for but we cannot calculate reliably?

Say what it would require, which inputs are missing and who owns them, and offer the closest reliable measure in its place. Publishing an unreliable version of a requested figure creates a dispute later that costs more than the delay.

Who signs off a revision to a metric definition?

The named owner signs the revision, which is logged and either restates or flags the history it touches. A definition edited without that restatement leaves trend lines that cannot be compared across periods.