ujans
Blog
Cyber security

Web application security: the OWASP Top 10 explained plainly for businesses

Your website, customer portal or admin panel is the first place attackers look. Knowing the most common vulnerability types lets you ask the developer the right questions.

ujans team··5 min read

Websites and web applications (customer portals, booking systems, e-commerce, admin panels) are constantly scanned and attacked because they face the internet. OWASP (the Open Worldwide Application Security Project) gathers the most critical security risks in web applications in the "OWASP Top 10" list and updates it periodically. This article summarises the list's risk categories plainly for non-technical managers and explains which safeguards to ask for. For details of the current list see OWASP's publications; category names and order change over time.

The big picture: the main risk families

1. Broken access control

Users reaching data or actions they are not authorised for. Example: a customer changing the number in the address to see someone else's order; an ordinary user entering the admin page.

Safeguard: Server-side authorisation checks on every action, "deny" by default, least privilege, role-based access. This topic should be central to the development process.

2. Cryptographic failures

Sensitive data not encrypted or weakly encrypted: unencrypted connections, passwords stored in plain text, old algorithms.

Safeguard: HTTPS on all traffic, passwords stored with strong salted hashing algorithms, not keeping card and personal data unnecessarily, secure key management. For data protection see the data protection guide and the KVKK compliance checklist.

3. Injection

Data from the user being executed as a command or query (database query, command line, template). The best-known type is SQL injection. Cross-site scripting (XSS) is also handled in this family: a malicious script running in other users' browsers.

Safeguard: Parameterised queries, input validation, output encoding, safe libraries.

4. Insecure design

The problem comes not from code but from design: no threat modelling, gaps in business logic (unlimited attempts, abuse of discount codes), abuse scenarios not considered.

Safeguard: The question "what would a malicious user do?" at the design stage, security requirements, workflow abuse tests.

5. Security misconfiguration

Default passwords, unnecessary open services, detailed error messages, exposed admin interfaces, wrongly permissioned cloud storage, missing security headers.

Safeguard: Hardened standard configurations, environment separation (test and live), disabling unnecessary features, automated configuration checks, review of cloud permissions.

6. Vulnerable and outdated components

Libraries, frameworks, plugins and server software that are not updated. Exploiting publicly known flaws is the easiest attack route.

Safeguard: Component inventory, regular updates, dependency scanning, avoiding software past end of support. For the process see the patch management guide and website maintenance.

7. Identification and authentication failures

Weak password policy, openness to brute-force attempts, missing multi-factor authentication, session management errors, weak password reset.

Safeguard: MFA, attempt limiting, secure session cookies, a strong password policy and breached-password checks, a secure reset flow.

8. Software and data integrity failures

Updates, dependencies or data flows not verified; pulling code or packages from untrusted sources (supply chain risk).

Safeguard: Signature verification, trusted sources, build pipeline security, review of changes.

9. Logging and monitoring failures

An attack going unnoticed: login attempts, authorisation errors and critical actions not recorded, no alerts.

Safeguard: Meaningful logging, central collection, alert rules, an incident response process; see the incident response plan and endpoint security.

10. Server-side request forgery and exception handling

The server being persuaded to make requests to addresses controlled by the attacker; unsafe behaviour on errors, error screens that leak information.

Safeguard: Allow lists for outbound requests, generic messages and safe default behaviour on errors.

Note: category names, order and grouping vary between versions; the headings above cover the core risk families.

What should you do as a business?

When buying or commissioning

  1. Write security requirements into the contract: Responsibilities for secure development, testing and updates.
  2. Ask the software team: How are input validation, authorisation checks, password storage, updates and logging done?
  3. Keep source code and access ownership with you; see how to manage a software project.
  4. Technology choice: Ready systems (CMS, plugins) need update discipline; in custom development secure design matters; see WordPress or custom development?.

When operating

  1. Regular updates and maintenance; see website maintenance.
  2. Admin panel protection: MFA, IP restriction, unique accounts, a password manager.
  3. Backup and recovery: see backup strategy and the disaster recovery plan.
  4. Network layer protection: A firewall and, if needed, a web application firewall (WAF); see choosing a firewall.
  5. Monitoring: Alerts for abnormal logins, errors and traffic spikes.

Verification

  • Security testing: Regular vulnerability scanning and periodic penetration testing; see penetration testing.
  • Code review and dependency scanning.
  • Retesting after major changes.

Privacy and compliance

If a web application processes personal data, technical measures are part of the legal obligation. Notification processes may also be needed when an incident occurs; see the KVKK compliance checklist and the incident response plan.

Sector notes

Common mistakes

  • Leaving security to the end,
  • Postponing updates,
  • Weak passwords and MFA-less access to the admin panel,
  • Leaking detail in error messages,
  • Leaving the test environment open to the internet and unprotected,
  • Not changing default passwords,
  • Testing once and never repeating,
  • Not clarifying source code and access ownership.

Quick-start checklist

  1. HTTPS and current versions.
  2. MFA on admin accounts.
  3. Removal of unneeded plugins and accounts.
  4. Regular backup and a restore test.
  5. Review of input validation and authorisation checks.
  6. Logging and alerting.
  7. At least one independent security test a year.

ujans offers secure software development, security testing, monitoring and maintenance together for your web applications.

Looking for help with this?

Tell us what you need and our team will get back to you within one business day.

Get a quote