The short answer
A vulnerability scan is software that checks your systems against a list of known problems. A penetration test is people trying to break in, the way a real attacker would, and proving what they could reach.
What a vulnerability scan does
A scanner looks at your systems and compares what it finds to a long list of known problems, like old software or a weak setting. It is fast, cheap and easy to run often.
But a scanner does not know what your app is for. It cannot tell that one customer should never see another's data, so it cannot test for it. It also flags problems that are not real, and someone has to sort through them.
What a penetration test does
In a pentest, skilled testers use tools to start, then work by hand. They check what the tools found, then go further:
- They chain small problems together into a real attack.
- They test the rules of your own app, like who can see or change what.
- They prove each finding, so you know it is real.
- They tell you what to fix first, in plain English.
Side by side
| Vulnerability scan | Penetration test | |
|---|---|---|
| Who does it | Software | People, with tools |
| How long it takes | Minutes to hours | Days to weeks |
| How often | Monthly, or all the time | Yearly, and after big changes |
| Finds known problems | Yes | Yes |
| Finds flaws in your app's rules | No | Yes |
| Proves each problem is real | No | Yes |
You need both
They do different jobs. Run scans often, to catch known problems as they appear. Get a pentest at least once a year, and after big changes, to find what scans miss.
Some rules ask for both. PCI DSS, for example, asks for scans every three months and a pentest every year.
How to tell what you are buying
Some products sold as a "pentest" are really a scan with a nicer report. Ask three questions before you buy:
- Who does the testing, a person or a tool?
- Will you test while logged in, as each kind of user?
- Will you test again after we fix things?
If the answers are vague, it is probably a scan.