A penetration test is an authorised, time-boxed simulated attack on systems you own, carried out by people, to find weaknesses and prove which ones are genuinely exploitable. It differs from a scan in one decisive way: a tester chains small flaws together and demonstrates the impact, rather than listing possibilities. The deliverable is a report with evidence, severity ratings and fixes, followed by a retest.
Why this matters for your website
Most real breaches of small sites are not exotic. They are an outdated plugin, an exposed admin panel, a forgotten staging subdomain, or a checkout flow that lets a customer view another customer's order by changing an ID in the URL. Automated tools catch the first three reliably. The fourth, a business-logic flaw, usually needs a human.
You may also be asked for a test report by an enterprise customer, an insurer or an auditor. Knowing what a penetration test actually includes helps you buy the right thing, scope it sensibly and judge whether the report you receive is worth what you paid. For a direct comparison of the two options, see penetration testing vs vulnerability scanning.
Before the test: scope, rules and written authorisation
Nothing legitimate starts without paperwork. Three documents define the engagement:
- Scope. The exact domains, IP ranges, applications and user roles in play, plus anything explicitly excluded, such as third-party payment providers or your email host.
- Rules of engagement. Test window, whether production or staging is used, rate limits, whether denial-of-service and social engineering are permitted, and an emergency contact for both sides.
- Authorisation letter. Written permission from the asset owner. Without it, the same activity is a criminal offence in most jurisdictions, and your hosting provider may suspend the account.
Expect to supply test accounts, a staging copy if production is fragile, and a note of any fragile integrations. Good scoping is where most of the value is won or lost.
During the test: recon, exploitation, post-exploitation
A structured tester follows a published methodology rather than improvising. The common ones are the OWASP Web Security Testing Guide for web applications, the Penetration Testing Execution Standard for broader engagements, and NIST SP 800-115 for a technical assessment framework.
- Reconnaissance. Mapping subdomains, technologies, endpoints, email records and exposed files to build an attack surface picture.
- Vulnerability discovery. Automated tooling plus manual review of authentication, access control, input handling and session management.
- Exploitation. Proving a finding is real, for example extracting a non-sensitive record through an injection flaw rather than merely flagging it.
- Post-exploitation. Establishing what the foothold reaches: other accounts, the database, internal hosts, cloud credentials. This is where the business impact is measured.
- Cleanup. Removing test accounts, uploaded files and any artefacts left behind.
Everything stays inside scope. If a tester discovers something out of scope that is clearly dangerous, the correct response is to stop and call you, not to continue.
After the test: the report, severity and the retest
A usable report has an executive summary in business language, a findings section with reproduction steps and screenshots, severity ratings, and remediation advice per finding. Severities are usually scored with CVSS, but ask the tester to adjust for your context: a critical flaw on a marketing microsite is not the same risk as the same flaw on checkout.
Agree the retest in the contract. You fix, the tester verifies, and the report is reissued with findings marked as resolved. A report without a retest clause tends to age badly.
Types of penetration test
Two choices define the engagement: how much information the tester gets, and what is being tested.
- Black box. No internal information. Realistic, but time is spent on discovery rather than depth.
- Grey box. Credentials and basic documentation supplied. The usual choice for web applications, and normally the best value.
- White box. Full access to source code, architecture and configuration. Deepest coverage, highest cost.
- Web application test. Your site, admin area, APIs and authenticated user journeys.
- External network test. Internet-facing hosts, open ports and exposed services.
- Internal network test. Simulates an attacker already inside the office or a compromised laptop.
- Social engineering. Phishing and pretexting against staff, only with explicit written consent.
For most WordPress, WooCommerce and small SaaS operators, a grey-box web application test is the relevant one. Costs vary widely by scope; see how much a penetration test costs.
Who performs them, and what the credentials mean
Credentials are a shortcut for judging competence, not a guarantee. The ones you will see most often:
- OSCP — a hands-on individual certification from OffSec with a practical exam; a reasonable floor for a technical tester.
- CREST — an accreditation body that certifies both individuals and companies against assessed standards; common in UK and EU procurement.
- Methodology references — a proposal that names OWASP WSTG or PTES and lists the test cases it covers is more informative than a logo.
Ask for a redacted sample report before you sign. It tells you more than any certificate.
What a penetration test is not
- Not a vulnerability scan. A scan is automated, repeatable and continuous; a pentest is manual, deep and point-in-time.
- Not a guarantee. It reflects the scope, the time budget and the code as it existed that week.
- Not a one-time fix. Every plugin update, feature release and new subdomain changes the attack surface.
- Not a compliance certificate. It is evidence that supports an audit, not the audit itself.
How to check your own site
Before commissioning a manual test, remove the findings a machine can find for you. Paying a specialist day rate to discover a missing Content-Security-Policy header or an unpatched plugin is poor value.
Start with the free tools, no account required: the Security Headers Checker, the SPF & DMARC Checker and the Website Security Check.
Scantinel (scantinel.ai) is an automated website security assessment platform that runs OSINT, network and OWASP Top 10 web vulnerability tests against ownership-verified assets and delivers actionable, business-owner-legible reports. The Essential scan covers DNS, TLS, security headers, cookie flags and known-vulnerable CMS plugins in minutes. The Full scan adds port and service fingerprinting with CVE correlation, a headless-browser crawl, an active OWASP Top 10 scan and WordPress or WooCommerce deep checks. None of this is a penetration test, and we do not call it one; it is the groundwork that makes a later pentest worth paying for. You can start with an automated security assessment and book human testing for what remains.
Key takeaways
- A penetration test is an authorised, scoped, human-led simulated attack that proves impact, not just possibility.
- No engagement should start without a signed scope, rules of engagement and an authorisation letter.
- Grey-box web application testing is the usual fit for WooCommerce stores and small SaaS products.
- Judge testers by methodology coverage and a sample report, not by certification logos alone.
- Automate the routine findings continuously; reserve manual testing for logic flaws and access control.
Frequently asked questions
Is penetration testing legal?
Yes, when you have written authorisation from the owner of every asset in scope. Testing systems you do not own, or that your provider forbids, can breach computer misuse laws and your hosting terms. If your site sits on shared hosting or a managed platform, notify the provider before the test window.
How often should a small business do a penetration test?
A common pattern is annually, plus after any significant change such as a new checkout flow, a payment integration or a platform migration. Between tests, run automated scanning on a schedule so new plugin vulnerabilities and configuration drift surface quickly. Contractual or regulatory obligations may set a stricter cadence.
What is the difference between a pentest and a security audit?
A penetration test is adversarial and technical: it attempts to break in and demonstrates what succeeded. A security audit is a review against a standard or policy, checking that controls exist and are documented. Auditors often ask for a recent penetration test report as evidence, which is why the two are frequently bought together.
Published February 2026 by the Scantinel Security Research Team. Security guidance changes; verify details against primary sources such as OWASP, the NVD and WPScan.
Comments
No comments yet.