Vulnerability scan vs penetration test: which one your customer is really asking for
· 6 min read · Pentest reports, Security questionnaires
A customer asks for "a pentest report". Another asks for "evidence of regular vulnerability scanning". A third asks for "the results of your most recent security assessment". These are not the same request, and answering the wrong one costs you either money or a second round of emails.
The short version
A vulnerability scan looks for known weaknesses and reports what it sees. It checks versions, headers, certificates, open ports and configuration against a large set of known problems. It is broad, fast and repeatable, and it does not try to prove that anything can be exploited.
A penetration test tries to break in. It sends real attack payloads to see whether a weakness can actually be used, and it reports what it managed to do. It goes deeper on fewer things, and it produces evidence of exploitability rather than a list of possibilities.
Put another way: a scan tells you that a door looks unlocked. A pentest tries the handle.
Why the difference matters to the reader
A scanner export often runs to hundreds of entries. Most are not reachable, not exploitable, or already mitigated somewhere else in your stack. A security reviewer who receives one cannot tell your real exposure from noise, and neither can you.
A pentest report should be shorter and more useful. It should say what was tried, what worked, and what the proof is. That is why buyers ask for one when the stakes are higher, such as an enterprise contract, a SOC 2 audit, or a due-diligence review.
How to tell which one they want
Read the wording of the request closely.
| The questionnaire says | It probably wants |
|---|---|
| "Penetration test", "pentest", "ethical hacking" | A penetration test report |
| "Third-party penetration test", "independent test" | A penetration test by someone other than you |
| "Human-led", "manual penetration test" | A human tester. An automated test will not satisfy this |
| "Vulnerability scanning", "regular scans", "scan cadence" | Evidence of scanning that recurs, with dates |
| "Vulnerability management programme" | A process: scanning, triage, fix and re-test, over time |
| "Most recent security assessment" | Ambiguous. Ask, or send a pentest report, which covers both |
Scroll the table sideways to see every column.
Two cases need care.
"Human-led" is a hard requirement. If the contract or questionnaire specifies a human tester, an automated test, ours included, does not meet it. Do not try to argue it does.
"Regular" means a cadence, not a single report. One excellent pentest from eight months ago does not show regular scanning. Recurring tests with dated results do.
Where SecQon sits
SecQon does both, at different depths, and labels which is which.
- Light and Standard scans are non-intrusive. They send no attack payloads. They cover TLS and certificates, security headers and cookie flags, DNS and email spoofing (SPF, DMARC, CAA, DNSSEC), exposed services and ports, exposed version-control and build files, CORS, and template checks for known-vulnerable software. Validation at these depths re-tests that a finding is still observable. It does not try to exploit it.
- Deep scans add bounded proof where the class of issue allows it. Reflected markup injection is rendered in a headless browser, SQL injection is shown by a boolean differential that reads no data, and SSRF is shown by your server calling a collector we run.
- The penetration test adds an active scan with real payloads from a narrow allowlist: path traversal, file inclusion, OS-command injection, server-side template injection, XXE, XPath injection, open redirect, CRLF, SQL injection and reflected XSS. It also tests API operations that answer without the authentication their own schema declares, and measures how much of a fixed set of attack signatures your WAF blocked.
Every finding is labelled by what was established: demonstrated, confirmed, likely or unconfirmed. A reader can see the difference between a scan observation and a proven exploit.
The penetration test is automated, not human-led, and the report says so. It does not test business logic, race conditions or stored XSS, and it does not chain findings into attack paths.
If you are still unsure
Ask the customer. A one-line reply such as "Our most recent external test is an automated penetration test with evidence for each finding. Does that meet your requirement, or do you need a human-led test?" settles it, and the security team will respect the precision.
If an automated pentest is enough, the one-off pentest report runs one full Deep and penetration-test scan of one verified asset. If they want evidence of regular scanning, continuous monitoring re-tests daily and keeps the dated record.