How to use AppSense to track software usage for license reclamation

Licence reclamation works best when it starts with evidence of entitlement and actual use, not with a list of employees who have not opened an application. AppSense can expose launch, update, service and error patterns when the relevant event logging is enabled and healthy. The practical goal is to separate genuine redundancy from valuable access that is seasonal, shared or difficult to observe. A good analysis produces three outcomes: retain access with a clear business purpose, reclaim seats that are demonstrably surplus, and investigate ambiguous patterns before making any change.

Many Australian organisations now combine Sydney and Melbourne offices with distributed staff in Brisbane, the Northern Territory and regional Western Australia, alongside contractors and hybrid workers who alternate between office devices and virtual desktops. In that environment, one inactivity rule will create false positives. The method below turns AppSense software usage analytics into a staged operating process, from data preparation and segmentation through to validation, communication and licence recovery. Privacy controls and change management remain part of the process because cost savings should not weaken trust or operational continuity.

Establish a defensible licence baseline

Begin with a reconciled source of truth. Record each product, licence metric, subscription term, quantity, assigned user or pool, cost centre and renewal date, then compare that inventory with AppSense event coverage. Named-user, concurrent-use, device and volume licences need different tests. A rarely used named seat may be reclaimable, while a shared service may require enough capacity at peak demand.

Define meaningful use before reviewing individuals. Installation, update checks and background services create events even when there is no productive interaction. For business software, a user-initiated launch with successful completion may be the stronger signal; for security or monitoring agents, continued service activity may be the real signal. Agree on role-sensitive thresholds, such as no qualifying event across a complete seasonal or contractual period.

Assign owners and measures. Track the proportion of paid seats with qualifying use, candidate accuracy, realised rather than theoretical savings, and service-desk impact. Give IT, finance or procurement, privacy and service-desk leads defined responsibilities. Pilot one controlled cohort before organisation-wide action, particularly where vendor terms, minimum commitments or audit rights restrict recovery.

Configure trustworthy event collection

Confirm that collection can answer the question. Useful fields generally identify the product, action, endpoint, authenticated user or device, timestamp, version, and success or failure. The detail varies by product and deployment, so document what the local AppSense configuration records instead of assuming every client contributes equivalent telemetry. Check agent health, reporting intervals, server connectivity and clock synchronisation before treating a gap as non-use.

Event volume can also distort interpretation. This capacity forecasting guide shows how log analytics can reveal reporting and capacity trends; apply the same discipline to establish whether declining event counts mean fewer applications are being used or agents have stopped reporting. Compare expected endpoint populations with reporting populations at several points in the day, and investigate unexplained drops before calculating reclaimable seats.

Keep telemetry proportionate. Under the Privacy Act 1988 and the Australian Privacy Principles, covered organisations should collect only reasonably necessary data, use it for the stated purpose, protect it from unauthorised access and retain it no longer than required. Assess any eligible notifiable data breach under the applicable rules, and follow relevant state or territory privacy obligations. Prefer aggregate reports and avoid document content, keystrokes or unrelated personal activity.

Create a representative usage baseline

Build a baseline long enough to reveal the organisation's rhythm. A 30-day window may miss month-end processing, school holidays or the Australian financial year from July to June; a 60- to 90-day window can be more representative where schedules are seasonal. Record launches alongside business dates, then compare departments rather than forcing every team into one peak. For event-based products, zero recorded activity is more meaningful than a missing feed.

Common entries become useful when connected to the service context. Use event log troubleshooting to compare ambiguous events with configuration issues, support tickets, failed launches and known agent behaviour. Separate a person who rejected a sign-in from a person who never attempted to open the application. Exclude installer activity, update checks and scheduled services unless those actions represent the product's intended use.

Hybrid work complicates the denominator. Shared hot-desk machines, kiosks, virtual desktops and contractor laptops can produce several device identities for one person, while one shared workstation can produce many. Maintain a documented mapping between people, accounts, licences and devices, and flag shared-service accounts separately. Never add device counts together and present the total as a count of active users.

Segment usage by entitlement and role

Segment usage before drawing conclusions. Compare active days, event frequency, last qualifying use, trend and concentration across departments, job roles, sites and device types. Melbourne, Sydney and Brisbane teams may follow different project calendars, while field and telehealth staff may work through shared or remote endpoints. Anonymised cohort views expose these differences without giving managers a list of allegedly inactive employees.

Match the pattern to the licence model. A named-user agreement often supports reclaiming an individual assignment, while concurrent licensing depends on measured peak demand and may tolerate dormant named accounts. Device licences and cloud subscriptions can behave differently again. Look for stable daily users, periodic specialists, seasonal users, dormant accounts, duplicated entitlements and products replaced by a standard tool, but do not collapse these categories into one inactivity label.

Connect usage with entitlement ownership, not identity alone. Service-desk ownership, procurement records, cost centres and contractual commitments show why access exists. Review assignments when roles change, and treat professional development access as an enabler rather than weak performance. Programmes with periodic participation, including Encourage Her Network, can appear inactive between sessions even though the underlying access remains valuable.

Score licence-reclaim candidates

Create a transparent candidate score rather than a silent blacklist. Start with rules such as no qualifying event for a defined period, duplicate seats for the same person, unused pooled capacity or a product replaced by an approved alternative. Weight business criticality and role exceptions separately so finance can see both the potential saving and the reason for caution. Keep the scoring logic consistent when the review repeats.

Estimate commercial value in Australian dollars, but distinguish potential from recoverable value. Multiply the relevant subscription or maintenance cost by a defensible reclaimable percentage, then deduct shared seats, contractual minimums, migration costs and vendor restrictions on mid-term transfers. Some Australian agreements are billed through local resellers and may use different cancellation terms from the publisher, so procurement should verify every material assumption before an approval is recorded.

Create an auditable candidate register with the entitlement, qualifying activity, last relevant event, score, exception, proposed action and approver. Exclude infrastructure, security, backup and monitoring agents unless their business function has genuinely changed. Where evidence is weak, assign an investigation owner and a review date rather than forcing the record into either a retain or reclaim decision.

Validate findings before changing access

Validate a sample from every meaningful cohort before changing access. Compare AppSense findings with application inventories, identity records, manager knowledge, service-desk tickets and approved business activities. Investigate failed launches, migrations, leave and long-running projects. A manual check may show that the application was used through an API, a mobile client or another approved route that the event definition missed.

Communicate with affected staff and their managers before removing a licence. Explain the usage evidence, the decision owner, the proposed action, any transition period and the help-desk route. Review relevant employment contracts, enterprise agreements, awards, internal policies and vendor terms; some workplaces have consultation obligations that apply even when the change concerns software rather than physical equipment. Keep discussions factual and avoid framing low usage as poor performance.

Run the change as a reversible pilot. Transfer one active user's seat to a candidate first where possible, or delay access changes for operational teams until a maintenance window. Monitor sign-in failures, support volume and workarounds during the trial. Essential accessibility tools, clinical or financial systems and emergency procedures need explicit exemptions. A short observation period after each change is more credible than immediate mass cancellation.

Turn findings into controlled licence reclaims

Execute only approved actions. Reassign or remove the entitlement in the identity platform, complete any required vendor or reseller claim, and preserve the evidence and approvals in the change record. Avoid deleting installers, deployment scripts or security definitions until dependencies have been checked. Some subscriptions can be reduced at renewal but cannot be refunded mid-term, so the realised saving may appear later than the usage decision.

Measure and repeat. Compare qualifying use, assigned seats, reclaimed seats and saved spend against the original baseline, and record false positives as well as savings. Schedule the next review for the same business cycle, update thresholds when a product changes, and send owners a short exception report. This creates a governance trail for finance and procurement and prevents a one-off clean-up from becoming permanent waste.

Start with one product and one clearly owned cohort. Download a relevant configuration or workflow from the AppSense Exchange tools library, compare it with your own event coverage, and record the thresholds and exceptions in a one-page decision sheet. Publish an anonymised result, share the reusable configuration with the community, and repeat the review monthly. The result is a controlled licence-reclaim programme in which every removal is supported by evidence and every essential user keeps the access they need.