Reading a vulnerability report: severity, likelihood, and what to fix first
How to read a security scan or penetration test report as a business owner, what the severity labels mean, and how to decide the order of fixes.
The short answer
A vulnerability report, whether from an automated scan or a penetration test, lists findings, each with a severity and usually a likelihood or exploitability. Severity measures potential impact: what could happen if the flaw were used. Likelihood measures how easily it could be used: whether it is reachable from the internet without a login, whether an exploit is public, whether anything else already blocks it. The order of fixes follows both together. Critical and high findings reachable without login are fixed first, within days. Findings behind a login, or that need unusual conditions, are scheduled. Low and informational findings, of which automated scans produce many, are explained and handled in routine maintenance. A good report separates confirmed findings from potential ones and false positives; a good partner explains each finding in a sentence, provides a fix plan with dates, retests afterwards and writes a summary you could show a client.
Reading the labels
| Label | Usually means | Typical examples | Typical response |
|---|---|---|---|
| Critical | Takeover or full data exposure, easily exploited | Injection on a public endpoint, exposed admin with default credentials, leaked keys | Fix or mitigate within a day or two |
| High | Serious impact, exploitable with some conditions | Access control flaws, stored cross-site scripting, outdated component with a public exploit | Fix within days |
| Medium | Real but limited impact or harder to exploit | Missing security headers on sensitive pages, weak session settings, verbose errors | Fix within weeks, in the next release |
| Low | Minor impact | Version disclosure, minor misconfiguration | Routine maintenance |
| Informational | Not a vulnerability; an observation | Technology fingerprint, best-practice notes | Note and move on |
| False positive | The scanner was wrong | A pattern match that does not apply | Document why and exclude |
Turning the report into a plan
- Sort by severity, then within each level by whether the finding is reachable without login.
- Confirm each critical and high finding; scanners overstate, testers usually do not.
- Fix or mitigate the top group immediately; a mitigation, such as a firewall rule or disabling a feature, can buy time.
- Schedule the middle group into the next release with dates.
- Explain the rest in a sentence each; accept what is genuinely not worth fixing, with the reason written down.
- Retest and get the report updated with each finding’s status.
- Keep the report and the retest as the record for clients, insurers and yourselves.
What a good report contains
For each finding: where it is, what it is, how it was confirmed, the evidence, the severity and exploitability, and a specific recommendation. A summary that a non-technical reader can follow. A separation of confirmed, potential and false positive. A retest section. And the scope: what was tested and what was not, because a clean report of a narrow scope proves little about the rest.
What this means for you
Read a vulnerability report by severity and exploitability together, fix the critical and high findings reachable from the internet within days, schedule the rest with dates, explain or accept the low and informational items, and insist on a retest that records each finding’s status. The report is not a verdict on your site; it is a work list, and reading it properly is what turns it into a safer site rather than a filed document.
Frequently asked questions
The report has thirty findings. Should we panic?
No. Look at how many are critical or high, how many of those are reachable without logging in, and how many are confirmed rather than potential. A scan of a typical site produces many low and informational items, such as version disclosures or missing headers, alongside a handful that matter. The handful is the work; the rest is scheduled or explained.
What is the difference between severity and risk?
Severity is what could happen if the finding were exploited: data exposure, takeover, defacement. Risk combines that with how likely exploitation is, given whether the flaw is reachable, whether an exploit is public, and whether anything else already blocks it. A severe flaw behind a login used by two trusted people is lower risk than a moderate one on the public contact form. Fix order follows risk.
How do we know the fixes worked?
By a retest: the same checks run again after the fixes, with the report showing each finding as resolved, accepted with a reason, or still open. A report without a retest is a list of problems; a report with one is a record of a site made safer. Ask for the retest as part of the engagement, not as an extra.
Sources
- FIRST: Common Vulnerability Scoring System (CVSS) specification (accessed 2026-09-12)