Cybersecurity / Guide
Penetration Testing for SMEs: What It Costs and How It Works
Nov 11, 20269 min read
A pentest is a planned, authorised attack on your own systems, not a product you buy off a shelf. This guide covers what a pentest actually is versus a vulnerability scan, the three ways to scope one, what it costs, how a proper engagement runs from scoping call to retest, and how to recognise a trustworthy provider.

A pentest is not a product you buy off a shelf — it's a planned, supervised attack on your own systems, carried out by someone you've hired to do it. For a Dutch SME, the question usually isn't whether a pentest is a good idea; it's what one actually involves, what it costs, and how to tell a competent provider from one selling you a glorified scan under a more expensive name.
This guide walks through what a penetration test is and isn't, the three ways a test can be scoped, what it costs and how often to repeat it, how a proper engagement actually runs from first conversation to retest, and what to check before you sign anything.
What a pentest actually is
A pentest, short for penetration test, checks whether your IT security actually holds up. A hired specialist attacks your digital systems in a planned, authorised way, then tells you which weaknesses they found and how to fix them. It's useful for any business that handles sensitive data or depends on IT systems to operate, which by now is most of them.
The Dutch national cybersecurity centre NCSC, in an expert blog on the topic, points to a widely used definition from Cyberveilig Nederland's own cybersecurity dictionary: a pentest is a manual check that tries to penetrate as deep as possible into a digital system, to find weak points and understand the consequences if they were exploited. The tester uses each weakness found to dig further in. The goal is not to find as many weaknesses as possible — that's what a vulnerability scan does — but to establish whether a system has exploitable weaknesses at all.
That distinction matters more than it sounds. A vulnerability scan looks broadly across a system for known flaws and misconfigurations without actually using them to get anywhere. A pentest does the opposite: it exploits a smaller set of findings to gain a foothold, then explores what else is reachable from there. That's what gives it more depth than a scan, and why, in the NCSC's own words, it can never be a fully automated checklist no matter what some providers claim. Automated tools handle the repetitive groundwork; a tester with real attack experience interprets how a system responds and decides where to push.
Nor is a pentest trying to catalogue every vulnerability. A competent tester looks for a combination of weaknesses and misconfigurations that chains into a working attack path, one that actually reaches a system or confidential data, rather than listing every minor flaw regardless of whether it leads anywhere.
Black box, grey box or white box
Before a test starts, you agree with the tester on exactly what may be attacked and how much the tester already knows going in. That foreknowledge defines three common test types:
- Black box. The tester starts knowing nothing about your company, the same position a cybercriminal would be in on a first attempt.
- Grey box. The tester gets access to part of your systems, simulating an attacker who's already found a way inside your network.
- White box. The tester gets full information and access, including passwords and technical details about the software involved — useful when the goal is to thoroughly test one specific application or package rather than simulate an outside attacker.
None of these is objectively "better." The right choice depends on what you actually want answered: whether an outsider can get in at all, or how much damage someone could do once they're already past the front door.
What a pentest costs for an SME
Price and duration depend entirely on scope. A small, focused pentest costs around €1,000 and takes about a day, checking your system against vulnerabilities that commonly turn up. A broader engagement, covering more systems or a deeper white-box review, can run into the thousands or tens of thousands of euros and take several weeks. In most cases, your systems stay in normal use while they're being tested. A quote built around an automated scan alone, sold as a pentest, will miss exactly the chained, exploitable weaknesses a real one is meant to find — which is one reason to compare providers on more than price.
How a proper engagement actually runs
Before you contact a provider, decide what you actually want tested and why. As the system owner, you set the research question, the scope — how deep the tester may dig — and the timing. Don't schedule a test for a brand-new application the day before it goes live: fixing what a pentest turns up takes time, and reputable providers tend to have full calendars, so book ahead. Avoid changing anything in the systems under test while testing is underway; a changed component needs retesting, and an unrelated outage during the window can be mistaken for a test effect.
Choosing a provider starts with a scoping conversation, ideally with more than one candidate, explaining clearly what needs testing. Sign a non-disclosure agreement before sharing anything confidential. Ask about methodology, experience, certifications and training, and ask to see a sample report — the one tangible thing you'll actually receive at the end. A provider that asks a lot of questions about your goal, your scope and how your systems are set up is usually trying to deliver something tailored rather than a templated run-through. Compare quotes on quality and fit, not price alone.
During the test, agree in advance on the days and times testing happens and the IP addresses it will run from, so it can be told apart from a genuine attack. Settle the legal side with a liability waiver before anything starts — covering any third-party or cloud systems inside the scope too — and agree explicitly on when findings get shared. A critical or high-impact finding is something you want to hear about immediately, not buried in a final report weeks later.
The report is the one deliverable a pentest actually produces, so it needs to earn that. A good one states who tested what and when, including anything that couldn't be tested, backs every finding with evidence and an impact estimate, and gives a concrete fix for each one, plus a summary a non-technical manager can actually use.
After the report, the real work starts. Discuss the findings with whoever owns or built the system, assign remediation actions, and track them until closed — a report sitting in an inbox protects nobody. A pentest is a snapshot: a flaw found once in a piece of code is often not the only place that mistake exists, so fix the pattern, not just the one instance. Once fixes are in, book a retest — vulnerabilities aren't always correctly fixed the first time, and the only way to know for sure is to check again.
How often to repeat it
A pentest only tells you how your security looked on the day it was run. Criminals keep finding new ways in, and your own environment keeps changing — a new laptop, a new piece of software, a new integration — each of which can reopen a door that was closed before. For that reason, repeat a pentest regularly, at minimum once a year, and more often if you work with sensitive data or systems that change frequently.
Is a pentest itself risky
Yes, and that's worth saying plainly. A system can behave unpredictably or slow down during testing. Agree in advance with the tester on which risks you're willing to accept and which you're not, and make sure the tester themselves is someone you trust: they'll have access to real weaknesses in your systems and, potentially, real sensitive data, and you need confidence they won't misuse either.
Choosing a trustworthy provider
Only engage a recognised security firm. A practical filter: look for a provider affiliated with the industry association Cyberveilig Nederland, or one holding the Netherlands' national quality mark for pentesting, in place since 1 April 2021. The Centrum voor Criminaliteitspreventie en Veiligheid (CCV) maintains a public overview of providers carrying that mark, which is a reasonable starting point for a shortlist.
Before testing begins, you and the provider draw up a liability waiver. This document records your consent to a planned digital attack and states that the pentester isn't responsible for the consequences of the test itself, such as a system going down during it. Like the pentest, the waiver should be tailored to what's being tested; for anything beyond a small, routine engagement, having an IT lawyer look it over is a reasonable precaution.
Where this fits a wider security picture
A pentest is one check among several, not a complete security programme on its own. If your business falls under NIS2, our guide to NIS2 for SMBs covers the broader duty-of-care expectations that regular security testing feeds into. And one of the simplest fixes a pentest commonly turns up, enabling multi-factor authentication, is covered in detail in our guide to setting up MFA in Microsoft 365.
FAQ
What's the difference between a pentest and a vulnerability scan? A vulnerability scan looks broadly for known flaws and misconfigurations without exploiting them. A pentest takes a smaller set of findings and actually exploits them to get into a system, then explores what else is reachable from there — which is why it finds issues a scan misses, and why it costs more.
How much does a pentest cost for a small business? A small, focused test costs around €1,000 and takes about a day. A broader engagement covering more systems, or a thorough white-box review, can run into the thousands to tens of thousands of euros and take several weeks, depending on scope.
How often should we get a pentest done? At least once a year, and more often if you handle sensitive data or your systems change frequently. A pentest only reflects your security on the day it ran; new vulnerabilities and new changes to your own environment can reopen gaps afterward.
Is a pentest risky for our own systems? It can be — a system may behave unpredictably or slow down during the test. Agree with the tester beforehand on acceptable risk, and only work with a provider you trust with access to real weaknesses in your systems.
How do we recognise a trustworthy pentest provider? Look for affiliation with Cyberveilig Nederland or the Netherlands' national pentest quality mark (tracked by the CCV), ask for a sample report before committing, and compare providers on methodology and fit rather than price alone.
Do we need a legal agreement before a pentest starts? Yes. You and the provider sign a liability waiver covering the planned attack and the provider's responsibility for its consequences, including any third-party or cloud systems that fall inside the test's scope.
Get help acting on the results
A pentest is only worth what you do with the report. Our Cybersecurity & Identity team can help you scope a test properly, make sense of the findings, and fix what it turns up — from quick wins like MFA to deeper access and configuration issues. Once remediation becomes ongoing work rather than a one-off project, our Managed IT & Support team can take that on too.
- Pentest
- Cybersecurity
- Penetration Testing
- SME
Related Articles
Cybersecurity
AI-Powered Cybersecurity: How Machine Learning Is Transforming Threat Detection
Cyberattacks are growing in volume and sophistication. AI-powered security tools are shifting the balance back to defenders. This article examines how Dutch organisations are using ML-driven SIEM, EDR, and threat intelligence to stay ahead.
Cybersecurity
NIS2 for SMBs: What the Law Means and What You Must Arrange
The Dutch Cyberbeveiligingswet (NIS2) lands in 2026 and brings duty of care, a 24-hour reporting duty and personal director liability. Here is who is in scope, what you must arrange, and a practical step-by-step plan for SMBs.
Cybersecurity
Preventing Ransomware in the SMB: 8 Measures That Actually Work
SMBs are now the primary target for ransomware precisely because attackers assume their defences are weak. The good news: a handful of well-implemented measures stop the vast majority of attacks. Here are the 8 that deliver the most protection per euro.