Fast. Simple. No meetings. No delays.
Home / Blog / Denmark / A Web Application Security Checklist for Business Owners

A Web Application Security Checklist for Business Owners

A non-technical-friendly checklist for reviewing the security of a business web application: access, updates, backups, logging and testing.

Cover graphic for the article: A Web Application Security Checklist for Business Owners

A web application security checklist for a business owner is a short list of questions your developers or supplier should answer with evidence: who can access what, how updates are applied, whether backups have been restored, what is logged, which outside services hold keys, and when an independent review last happened. Use it to ask, not to accept promises.

Why a web application security checklist beats a promise

Vendors say they take security seriously. That statement costs nothing and tells you nothing. A checklist shifts the conversation from reassurance to evidence: show me the setting, the log, the restore test, the date. You do not need to understand the technical details to notice when an answer is vague, delayed or missing.

Run through this list with whoever builds or hosts your application, mark each item as yes, partly or no, and agree owners and dates for every item that is not a clear yes. Many security problems come from basic gaps rather than clever attacks, and the well-known OWASP project catalogues the most common web application weaknesses; your developers should know it. For the developer-level view, our OWASP-based checklist for PHP applications goes deeper.

Access control and passwords

  • Everyone has their own account. No shared logins for staff, suppliers or developers.
  • Multi-factor authentication is on for administrators, hosting accounts, the code repository, the domain registrar and email.
  • Roles follow the job. People see and change only what they need. Ask for a list of who has administrator rights and why.
  • Leavers lose access the same day. This includes former employees and former suppliers. Check that someone owns this process.
  • Passwords are stored hashed, never in plain text, and users can reset them safely. Ask the developers to confirm, not you to inspect.
  • Failed logins are limited, so someone cannot try thousands of guesses.
  • Permissions are enforced on the server. Hiding a button in the screen is not access control; the system must refuse the request itself.

Updates and dependency hygiene

Modern applications are assembled from many third-party libraries, plus a framework, a language runtime, a server and often a content system. Each one receives security fixes over time. A dependency vulnerability is a known weakness in one of those components; once it is public, it can be exploited by anyone who finds a system still running the old version.

  • There is a named person or team responsible for updates, and a regular rhythm for applying them.
  • Someone can show you the list of components in use and their versions.
  • Automated tools check dependencies for known vulnerabilities and report them.
  • The platform itself, such as the operating system, runtime and database, is still supported by its maker.
  • Updates are tried on a staging copy first and have a way back if something breaks.

If your software is on an unsupported platform, that is a business risk to plan around rather than a footnote; our article on WordPress security hardening applies the same ideas to one popular platform.

Backups and recovery tests

A backup you have never restored is a hope, not a backup. Ask for the following:

  1. What is backed up: the database, uploaded files and configuration, not only the code.
  2. How often, and how long copies are kept.
  3. Where copies are stored. At least one copy should be separate from the main hosting account, so one compromised account cannot erase everything.
  4. Whether backups are encrypted and who can access them.
  5. When a full restore was last tested, who did it and how long it took.

Also write down how much data loss and downtime your business could tolerate. Those two answers decide how often you need backups and how fast you need recovery, and they are business decisions, not technical ones.

Logging and monitoring

When something goes wrong, logs decide whether you learn what happened or only that something did. Confirm that the application records sign-ins, failed sign-ins, changes to permissions, and important business actions such as exports or deletions. Logs should include time and user, and must not include passwords or full card numbers.

  • Logs are kept somewhere that an attacker with access to the app cannot easily erase.
  • Someone is alerted about errors, repeated failed logins and unusual activity, rather than reading logs only after an incident.
  • Uptime is monitored from outside, so you hear about an outage before customers do.
  • There is a simple incident plan: who decides, who contacts customers, and who contacts the authorities when personal data is involved. Check current obligations with an adviser or the official source, as rules change.

Third-party services and keys

Most business applications depend on payment providers, email services, analytics, maps and other external APIs, each connected by a secret key. Those keys are as sensitive as passwords.

  • Keys are stored in protected configuration, never in the source code or in a shared document.
  • Each environment (test and production) has its own keys, and test keys cannot touch real data.
  • Keys have the minimum rights they need and can be rotated. Ask when this was last practised.
  • You know the full list of external services, who owns the accounts, and what data each one receives. This list also supports your privacy documentation; see GDPR by design.
  • Accounts for hosting, domains and key services are registered to your company, not to an individual developer.

If your application exposes an API, it needs its own review of authentication and limits; start with secure API development for SaaS products.

When to commission a security review

A checklist reveals gaps in process. It does not prove the application is free of flaws. An independent review, such as a code review or penetration test by a qualified person who did not build the system, adds a different perspective. Consider one when:

  • You are about to launch, or have made major changes.
  • The application handles sensitive personal, financial or health-related data.
  • A customer, insurer or partner asks for one.
  • You have taken over software from a previous supplier and do not know its condition.
  • There has been an incident or a near miss.

Define the scope in writing, agree what happens to the findings, and make sure someone has time to fix what is found. Ask the reviewer for a report you can understand, with priorities.

Putting it together

Pick the three items where your answer was no, assign each an owner and a date, and revisit the list at regular intervals and after any significant change. For national guidance, consult your country's official cyber security authority for its current recommendations. If you would like help working through this with your own application, see how we approach software development for Danish companies.

Frequently asked questions

How often should a security review happen?

There is no single interval that suits everyone. Review before launch, after major changes, after incidents, and when you take over software from another supplier. For applications with sensitive data, many businesses schedule a recurring independent review. Run the owner-level checklist more often, and after every significant change.

What is a dependency vulnerability?

It is a known security weakness in a library, framework or other component your application uses. Once the weakness is public, attackers can look for systems still running the affected version. Tools can scan your component list for known issues, and the fix is usually to update to a patched version.

Are backups enough?

No. Backups help you recover from data loss, but they do not prevent a breach, and an untested backup may not restore at all. You also need access control, updates, monitoring and a tested recovery process. Keep at least one copy separate from the main hosting account and practise restoring it.

Related articles

🚀 Start Now