Vulnerability assessment
A vulnerability assessment is a structured review that identifies security weaknesses in systems, networks or applications and ranks them by risk so an organization can decide what to fix first.
The term also appears in physical security and disaster planning.
How it works
Start by defining the systems and data in scope. Use scanning tools and manual checks to find weaknesses, then validate results to remove false positives. Prioritize confirmed findings by severity, exposure and business impact. Assign fixes and retest them.
A vulnerability assessment finds and ranks weaknesses. A penetration test goes further by attempting to exploit selected weaknesses and demonstrate their consequences. Neither guarantees security.
Fixing nine findings still leaves three open
A fictional startup assesses its customer-facing application before an enterprise security review. Its scan produces 18 alerts, but manual review removes six duplicates or false positives.
| Result | Count |
|---|---|
| Initial alerts | 18 |
| Duplicates or false positives | 6 |
| Confirmed weaknesses: 18 − 6 | 12 |
| Fixed and verified through retesting | 9 |
| Still open: 12 − 9 | 3 |
The startup should report three unresolved findings, their risks and planned fixes, rather than simply say it "passed a scan." One exposed administrator account could matter more than several low-impact findings.
What the report should contain
NIST's Technical Guide to Information Security Testing and Assessment covers planning, conducting tests, analyzing findings and developing mitigation strategies. Ask for affected assets, supporting evidence, recommended fixes and retest results. A raw scanner export is not evidence that remediation worked.
Questions people ask
What is CVE vs CVSS?
CVE provides identifiers for specific vulnerabilities. CVSS is a method for scoring vulnerability severity from 0 to 10. Not every assessment finding has a CVE identifier, and a CVSS score does not capture your full business risk.
What are the four types of vulnerability?
There is no universal four-part classification. Assessments are often grouped by what they examine, such as networks, individual hosts, applications or cloud configurations. Those categories overlap; define the actual scope rather than rely on a label.
