Six Asset Inventory Myths, Checked Against Vendor Documentation

Each myth below is a belief we hear stated as fact in buying conversations. Each check uses only what vendors publish on their own pages, reviewed in September 2026, and the data behind our ranking.

Myth 1: "The CMDB is already our asset inventory."

The CMDB is the system of record for configuration items, and IT workflows depend on it. ServiceNow describes its CMDB as a "single, trusted system of record for all configuration item (CI) data", populated by ServiceNow Discovery and by Service Graph Connectors. What a CMDB does not do on its own is check its records against every security tool that sees the same assets. A laptop can be in the CMDB with an owner and still be missing its EDR agent; a cloud workload can exist in the cloud console and not in the CMDB at all. On the pages we reviewed, ServiceNow does not document identities as an asset class. The fix is not to replace the CMDB but to reconcile it, which our note on how to reconcile a CMDB walks through.

Myth 2: "The EDR console lists every device."

It lists every device where the EDR agent runs, which is a different list. Lansweeper's analysis of several million endpoints across 30,000 organizations, published on 7 September 2026, found that 96% had some protection while 68.1% ran full EDR or next-generation antivirus. A device with no agent does not appear in the EDR console at all, so the console cannot tell you it is missing. Our EDR note shows how to build the device list from other sources first.

Myth 3: "A network scan finds everything."

A scan finds what answers on the networks it can reach, when it runs. It does not see a laptop on a home network, a SaaS app or a service account. runZero, which scores 90 on deployment effort for its agentless scanning, does not cover SaaS apps on the pages we reviewed, and documents identities only as directory data that enriches device records. In our coverage grid, SaaS apps are documented fully by 3 of seven tools and identities by 2. Scanning is the best way to find unknown devices; it is not a complete inventory.

Myth 4: "More integrations means a better inventory."

Integration counts measure different things. Axonius counts more than 1,200 adapters, JupiterOne more than 200 integrations, Armis more than 200 pre-built integrations, Oomnitza more than 1,500 data surfaces and Device42 more than 30 integrations. An adapter, an integration and a data surface are not the same unit. What matters is whether the tools you run are covered and whether their records are merged correctly. That is why our methodology weights correlation and deduplication at 18% next to source coverage at 22%: records from many sources are only useful once they are matched and their conflicts settled.

Myth 5: "Agentless means no setup."

Agentless means nothing is installed on the assets. Something still has to reach them. A scanner or a passive sensor has to sit where it can see each network segment, and an API integration needs access and credentials for each source. runZero states it needs no agents, authentication or appliances, which is why it has the top deployment score in our ranking. Armis says most customers see an inventory within minutes to hours. For Axonius, each adapter needs API access and credentials, so time to full coverage grows with the number of tools connected, which is why it scores 60 on deployment effort.

Myth 6: "Asset inventory tools never publish prices."

Three of the seven do. JupiterOne publishes annual tiers from $25,000. runZero publishes a free Community Edition for 100 or fewer assets and a Platform plan from $5,000. Lansweeper publishes Starter and Pro plans billed annually and a free tier up to 100 assets. Device42 publishes its pricing model but not amounts, and the other three quote through sales. Our pricing note sets out the figures and what their units mean.

What do the six have in common?

Each one treats a single source or a single number as the answer. An asset inventory is trustworthy when several sources are combined, matched and checked against each other, and when the gaps they reveal have owners. The University lessons on correlation, coverage gaps and ownership cover the method in more detail.