Hardware asset management software evaluations fail when they focus on UI demos instead of data pipeline tests. Flexera’s 2024 State of ITAM Report found that 53% of IT teams report challenges gaining or maintaining complete visibility of their technology assets, and the buying process itself is a contributing factor.
Most buying guides conflate three structurally different things (device telemetry, configuration management, and financial asset records) and evaluate all three against a single tool as if they were interchangeable. IT teams end up with a system that looks comprehensive in a sales walkthrough but can’t reconcile live hardware states against static configuration management database (CMDB) records without manual exports.
This guide gives technical evaluators a working checklist organized around the three data layers any mature HAM deployment requires.
The three data layers
Every HAM evaluation should start with an architectural distinction. Three structurally different data layers exist, and a tool that excels at one won’t substitute for the others:
-
The telemetry and inventory layer captures continuous agent-reported endpoint state, including battery health, CPU load, OS version, BSOD history, and storage wear. This data is live, high-frequency, and hardware-centric, showing what is happening on a device right now.
-
The configuration management layer is the CMDB (configuration items, or CIs, with dependency mappings, change records, and relationships between services and the hardware that supports them). CMDBs are structured around configuration state, not raw telemetry, and they’re typically updated on a schedule rather than in real time.
-
The financial and contractual ledger tracks procurement records, depreciation schedules, cost center allocations, and disposal certificates. This is the domain of ITAM platforms, not device management tools. A depreciation schedule and a battery health reading serve entirely different audiences and governance processes.
The critical evaluation question is whether the tool you’re assessing clearly handles its designated layer and whether it can feed accurate data into the other layers without requiring manual exports. A device management platform that surfaces live telemetry is genuinely valuable, but treating it as a financial ledger replacement creates audit risk, and treating a financial ITAM suite as a source of live hardware state creates operational blind spots.
This guide’s evaluation criteria cover the five capabilities with the greatest impact across the first two layers:
- Discovery
- Asset identity and lineage
- Peripheral coverage
- Lifecycle and warranty tracking
- CMDB reconciliation and integration architecture
Financial ledger capabilities (depreciation, procurement invoices, disposal certificates) are out of scope by design.
Discovery architecture and endpoint coverage
The evaluation question for discovery isn’t “does this tool find devices?” Every tool finds devices under ideal conditions. What you need to know is what happens to an endpoint that has been off-network for a long time.
Three main discovery approaches exist, and each has a meaningful blind spot:
-
Agent-based discovery deploys a lightweight client to each endpoint, which sends telemetry outbound over HTTPS even when the device is off VPN. This is the most reliable method for zero-trust environments and remote work fleets, because it doesn’t depend on network adjacency. The agent collects hardware telemetry locally and transmits it when connectivity is available, so it stores and forwards data from an off-network device rather than dropping it. The coverage question concerns enrollment: what percentage of your fleet has the agent installed, and what is the process for devices provisioned before the agent rollout?
-
Agentless network scanning works well for on-premises devices with stable network presence. It fails entirely for off-VPN endpoints, guest devices, and any machine that was offline during the scan window. If your fleet includes remote workers or contractor devices, agentless scanning will systematically undercount it.
-
IdP sync and MDM integration pull device records from identity providers and mobile device management platforms. This gives you enrollment data and compliance status, but not hardware telemetry.
Production deployments typically require all three paths working together.
When evaluating any tool, ask specifically what the telemetry retention policy is for an endpoint that reconnects after an extended offline period. A tool that stores and forwards buffered readings is architected differently from one that drops data for any device that missed the collection window.
Asset identity, custody, and lineage records
The quality of a HAM record depends on the stability of the identifier anchoring it. MAC addresses are dynamic, changing when a NIC is replaced, and virtual machines can share or spoof them. IP addresses rotate. Hostnames get reassigned. A record built on any of these identifiers will drift.
The anchor that doesn’t drift is the chassis serial number or SMBIOS UUID. The factory burns these into the hardware, and they survive software updates, NIC replacements, and OS reinstalls unchanged. Every device record in a mature HAM system should trace back to one of these stable identifiers. You should verify during evaluation that the tool actually uses them rather than the MAC address that happens to be the most readily available identifier from a network scan.
Beyond the identifier, a trustworthy asset record requires a specific set of fields, including assigned owner, department, cost center, and physical location. These are baseline requirements. The field that separates an inventory snapshot from an auditable system is the reassignment history log.
When a laptop moves from one employee to another, preserve both the old and new assignments as timestamped entries rather than overwriting them. An overwrite-based system tells you who has the device today, but an append-only log provides the full chain of custody, supporting compliance audits, security investigations, and dispute resolution when a device goes missing.
During evaluation, ask specifically whether the assignment history field is append-only or whether reassigning a device overwrites the previous owner record. If the vendor can’t answer that question without escalating to engineering, treat it as a red flag.
Peripheral and non-telemetry asset tracking
Agent-based tools cover endpoints well and cover peripherals poorly, or not at all. Monitors, docking stations, keyboards, headsets, and scanners can’t run a telemetry agent, so most device management platforms can’t see them, yet they make up a significant fraction of the physical estate that needs lifecycle records.
The evaluation criterion is whether the tool provides a dedicated tracking module for non-agent assets, with serial number as the mandatory identifier and a defined set of lifecycle states, not a free-text note bolted onto the parent endpoint’s record. A purpose-built module typically covers common peripheral types (displays, docks, keyboards, mice, scanners) as individual records, carries each through states like In Use, In Stock, In Repair, and Decommissioned, and tracks deployment and retirement dates alongside warranty data.
The key question to ask in any evaluation is whether a docking station can exist as its own distinct inventory record with its own serial number, lifecycle state, and assignment history, or whether it only appears as an attribute of the endpoint it’s connected to. The latter approach means it disappears from your records the moment it’s moved.
Lifecycle status, warranty, and refresh planning
An asset record without lifecycle state is a photograph. It captures a moment but tells you little about where the asset sits in its operational journey or when it needs attention.
The fields that drive operational decisions are current lifecycle status, warranty start date, warranty end date, end-of-life flag, and scheduled refresh trigger. Warranty data is the field most likely to degrade. If the tool requires manual entry of warranty dates, those dates are likely to drift within weeks of initial population, because procurement records are often imprecise and rarely updated afterward.
The evaluation criterion is automatic warranty retrieval from OEM records. A tool that pulls warranty status directly from manufacturer databases eliminates manual entry and keeps data current as renewals and extensions are applied. During evaluation, ask whether warranty dates are manually entered, imported from procurement records, or retrieved automatically from the OEM. The answer tells you how much trust to place in the warranty data six months after go-live.
Draw a clear boundary between lifecycle state tracking and financial depreciation. Knowing that a device is In Use, In Stock, In Repair, or Decommissioned is an operational record. Computing book value, applying depreciation schedules, and generating disposal certificates are financial ledger functions. Most device management tools handle the former, and few handle the latter, so conflating them during evaluation leads to unmet expectations on both sides.
Reconciliation, deduplication, and CMDB integration
The hardest operational problem in hardware asset management is the one vendors most often gloss over in a demo. Keeping live-discovered hardware states consistent with static CMDB records, without creating duplicate CIs or leaving ghost assets, requires a specific integration architecture, not just a feature checkbox.
The failure mode looks like this: Discovery runs, finds a device, and creates a CI. A week later, discovery runs again. The device has a new IP address, and the tool creates a second CI. You now have two records for one physical device, and neither is authoritative. Multiply that across a 10,000-device fleet, and the CMDB becomes a source of confusion rather than truth.
A well-architected integration handles three things. First, it anchors CI identity on stable hardware identifiers (chassis serial number, SMBIOS UUID) rather than volatile network attributes, so the same physical device always maps to the same CI. Second, it passes incoming data through a reconciliation step that evaluates whether an inbound record matches an existing CI before creating a new one. Third, it defines rules that prevent the integration from overwriting CI attributes maintained by other authoritative sources, such as ServiceNow Discovery or an MDM platform.
When evaluating CMDB integration, ask “what prevents a duplicate CI from forming when a device changes IP addresses?” The answer reveals whether the integration architecture handles deduplication properly or relies on scheduled exports that create drift between sync windows.
What to test in a live vendor demo
Slide-deck walkthroughs won’t show you integration architecture. You need to run specific scenarios against live systems. Four tests will tell you more than any feature matrix.
Test 1: MAC address collision. Ingest two endpoints that share the same MAC address (common in VM environments or after NIC replacements) and verify whether the tool resolves identity via SMBIOS UUID or chassis serial number, or collapses both records into a single duplicate. A tool that collapses on MAC will lose hardware identity every time a NIC is swapped.
Test 2: offline endpoint telemetry. Simulate an endpoint that has been off-network for 21 days and reconnects. Confirm whether telemetry readings from the offline period are stored and forwarded or simply dropped. Dropped data means your HAM records will show systematic deltas for devices used by remote and traveling employees, which are the devices most likely to develop undetected hardware problems.
Test 3: peripheral as a distinct record. Connect a docking station to an enrolled endpoint and verify whether the dock is logged as a distinct inventory record with its own serial number and lifecycle state. If it appears only as an attribute of the endpoint, it will disappear from your records when it’s moved to a different desk or placed in storage.
Test 4: directory reassignment propagation. Trigger a user reassignment in the directory (move the device to a new owner in the IdP) and measure how long before the corresponding CI attribute updates in the CMDB. This latency shows the integration’s practical reliability for time-sensitive operations like employee offboarding.
If a vendor can’t demonstrate these four scenarios live, request the specific log output or API payload as proof. A vendor who can’t produce either hasn’t tested the integration at the depth you need.
Evaluation checklist
Use this table as a structured reference during vendor evaluations. Each criterion maps to a capability covered in the sections above.
| Capability | What to verify | Evidence to request |
|---|---|---|
| Discovery: agent-based | Agent beacons over HTTPS without VPN dependency | Agent connectivity docs and offline telemetry retention policy |
| Discovery: off-network coverage | Telemetry stored and forwarded after extended offline period | Demo: reconnect a 21-day-offline device and confirm data |
| Discovery: multi-path enrollment | Supports agent, IdP sync, and MDM integration as distinct enrollment paths | Enrollment architecture diagram and supported IdPs and MDMs |
| Asset identity: stable hardware ID | Records anchored on chassis serial number or SMBIOS UUID, not MAC address | Demo: ingest two devices sharing a MAC address |
| Asset identity: append-only history | Reassignment log is append-only, not overwrite | Schema documentation for the assignment history field |
| Asset identity: required fields | Owner, department, cost center, physical location present as structured fields | Data model or field-level schema |
| Peripheral tracking: dedicated module | Non-agent assets (monitors, docks, keyboards) tracked as distinct records with serial numbers | Demo: connect a dock and verify it appears as a separate CI |
| Peripheral tracking: lifecycle states | Peripherals carry In Use / In Stock / In Repair / Decommissioned states | Peripheral record detail view |
| Peripheral tracking: bulk onboarding | CSV import supported for large peripheral estates | Import template and row limit per file |
| Lifecycle: defined status vocabulary | Current lifecycle state is a structured field, not a free-text note | Status field options in the data model |
| Warranty: automatic retrieval | Warranty dates pulled from OEM records, not entered manually | Warranty data source documentation |
| Warranty: end-of-life flags | EoL flag and refresh trigger date exist as structured fields | Data model and sample record with EoL flag set |
| CMDB integration: identity resolution | Dedup logic prevents duplicate CIs on device identity change | Demo: change a device IP and confirm no duplicate CI is created |
| CMDB integration: reconciliation rules | Integration cannot overwrite CI fields owned by other authoritative sources | Reconciliation rule configuration documentation |
| CMDB integration: sync modes | Both real-time and scheduled sync options documented, with known latency | Integration architecture diagram and sync interval documentation |
| CMDB integration: in-context telemetry | Live telemetry (battery, warranty, BSOD) reachable from within CI forms | Demo: open a CI and confirm telemetry is retrievable without switching tools |
The four demo tests in the previous section are the fastest signal of whether a tool operates as an active reconciliation engine or a passive database. A tool that passes all four is architected to handle real enterprise conditions, not clean-room demos.
Frequently asked questions
What is hardware asset management software?
Hardware asset management software tracks physical IT assets programmatically across their full lifecycle, from discovery and enrollment through identity, assignment, and decommissioning. A complete HAM system maintains stable hardware identifiers, structured assignment history, lifecycle status, and warranty data, connecting that record set to configuration management and financial systems through defined integration points. The goal is an authoritative, auditable record of every physical asset in the organization, one that stays accurate without manual reconciliation.
What is the difference between hardware asset management and device telemetry?
Device telemetry captures real-time endpoint vitals (CPU load, battery health, storage wear, OS state, crash history) continuously from an agent running on the device. Hardware asset management maintains the device’s static identity record, including ownership, department, lifecycle state, warranty expiration, and the full history of every reassignment since provisioning. Telemetry tells you what is happening to a device right now, while HAM software tells you what that device is, who is responsible for it, and where it sits in its operational lifecycle. Both layers are necessary.
How does hardware asset management software integrate with a CMDB?
A well-architected integration pushes hardware inventory data to the CMDB via REST webhooks or a middleware layer that transforms and buffers payloads. When data arrives in the CMDB, a reconciliation engine (such as ServiceNow’s native IRE) evaluates each inbound record against existing CIs using stable identifiers like serial number and hostname, preventing duplicate CIs from forming when network attributes change. Reconciliation rules then govern which CI fields the integration is permitted to write, protecting attributes owned by other authoritative sources. Scheduled-only sync without near-real-time webhook events creates drift, leaving the CMDB reflecting hardware state as of the last batch run rather than current state, which undermines change management and incident response accuracy.
Final thoughts on hardware asset management software
The three-layer model is the evaluation frame that matters: telemetry and inventory, configuration management, and the financial ledger are structurally different problems, and no single tool should be expected to replace all three. Evaluate device management platforms against the first two layers, and treat financial ITAM capability as a separate purchase decision entirely.
The live-demo tests are the fastest way to tell an active reconciliation engine from a passive database. A vendor who can run the MAC-collision, offline-telemetry, peripheral-as-CI, and directory-propagation scenarios on request has built the integration. A vendor who can’t is likely relying on presentation rather than tested capability.
For teams evaluating device management platforms against this framework, WXP’s developer documentation and APIs cover its device management, physical-assets, and CMDB integration capabilities in more technical detail.