Software Discovery, Inventory, and Reconciliation
Discovery and Inventory
Discovery finds software and hardware on the network; inventory normalizes what was found into a usable record. Candidates must distinguish agent-based discovery (installed on endpoints, reliable but limited to managed devices) from agentless discovery (network scans, SNMP, WMI, API integrations, which may miss offline or remote devices). A complete inventory combines discovery data with procurement records and HR/identity data to attribute usage to people and cost centers.
Normalization and the Software Catalog
Raw discovery produces executable names and versions. Normalization maps these to a recognized product, edition, and version in a software catalog. Without normalization, reconciliation is meaningless because the same product appears under many file names. Traps include treating a discovered executable as a licensed installation when it is a component, a driver, or a trial.
Reconciliation and the ELP
Reconciliation compares entitlements against consumption (installs or usage) to produce the Effective License Position. Positive ELP means surplus; negative ELP means a compliance gap. Consumption should be measured by the license metric: named users for user-based licenses, peak concurrency for concurrent licenses, and core counts for processor licenses.
Worked Example
Discovery shows 120 installs of a product licensed per named user; HR shows 95 active users; entitlements total 100 seats. Consumption by metric is 95, so the ELP is +5, even though raw installs exceed entitlements. Measuring by installs would wrongly report a 20-seat shortfall.
Traps
Beware stale discovery data, unmanaged devices, and counting installs when the metric is users.