Vulnerability Management

What is Penetration Testing?

Penetration testing is an authorised simulated attack on your systems, carried out by security professionals, to find and prove exploitable weaknesses before a real attacker does.

2 min read

A penetration test is not a scan. A scanner compares your software against a list of known issues and produces a report full of maybes. A penetration test puts a human in front of your system who tries to break in, and who then proves the finding by exploiting it.

That distinction matters commercially. Scanner output is cheap and mostly noise. A tester chaining three individually-low findings into full account takeover is the thing that changes what your engineering team does next week.

What a test actually involves

Testing follows a repeatable shape, whatever the target:

  1. Scoping. What is in scope, what is explicitly out, when testing happens, and who to call if something breaks. This gets written down and signed - see the note on authorisation below.
  2. Reconnaissance. Mapping the attack surface: subdomains, exposed services, technology versions, and anything about the organisation that helps an attacker.
  3. Discovery. Finding candidate weaknesses, both automatically and by hand. This is where owasp-top-10 categories get systematically checked on a web target.
  4. Exploitation. Proving the weakness is real. A finding nobody demonstrated is a theory.
  5. Post-exploitation. Establishing what an attacker could reach from that foothold - lateral-movement, data access, persistence.
  6. Reporting. An executive summary that a non-technical reader can act on, and technical detail a developer can reproduce.

Black box, grey box, white box

These describe how much you tell the tester. Black box means no information, simulating an outside attacker - realistic, but a lot of the budget goes on discovery. White box means full source and credentials, which finds far more per day. Grey box - usually credentials for each user role, no source - is the sensible default for most applications, because the interesting bugs like broken-access-control live behind login.

Authorisation is not optional

Testing a system without written permission is a criminal offence in most jurisdictions, India included. A legitimate engagement always has a signed Statement of Work and Rules of Engagement before anyone touches anything. Any firm willing to skip that is telling you something.

How often

Annually is the common baseline, but frequency should follow change. A product shipping weekly has an attack surface that looks nothing like last year's. Test after significant architectural change, before a major launch, and when a customer or framework requires it. A retest after remediation matters as much as the original test - a fix that was never verified is an assumption.

Next Step

Want this checked on your own systems?

We run the assessments this was written from. Tell us your stack and we will scope it - no commitment.

Mutual NDA before scoping · Reply within 4 business hours