A client portal security checklist is the list of controls a portal must have before clients trust it with their documents, orders and personal data: strong sign-in with multi-factor authentication, authorisation checked on every record and every file, encryption, logging, secure development and a plan for incidents. For European companies it also covers the GDPR’s requirement for appropriate security, data location, processor agreements and the rights of the people whose data the portal holds. Below is the checklist in eight areas, with what each control means in practice, who should own it and how often to review it.
Last updated: 19 November 2026.
What security does a client portal need?
A client portal needs security in proportion to what it holds and what it lets people do. A portal that shows delivery dates needs less than one that holds tax returns, medical records or contracts, but every portal shares three duties: make sure each user is who they claim to be, make sure they only see and change what belongs to them, and be able to show afterwards what happened.
The General Data Protection Regulation puts this into law for personal data. Its Article 32 asks for security “appropriate to the risk”, taking into account the state of the art, the cost and the nature of the processing. The checklist below translates that principle into concrete controls.
At Odduel, a creative and technology agency in Barcelona and São Paulo, we build portals through our application development service and treat this checklist as part of the specification, not a test at the end.
In short: require multi-factor authentication, check permissions on the server for every record and file, encrypt data in transit and at rest, log access, keep dependencies patched, test before launch, host where transfers are lawful, and rehearse what you will do if something goes wrong.
What should a client portal security checklist cover?
The checklist has eight areas. The table gives an overview; the sections below set out the controls in each.
| Area | Key controls | Typical owner |
|---|---|---|
| Identity and sign-in | MFA, secure passwords, invitations, session limits | Product owner with developers |
| Authorisation | Server-side checks per record and per file, roles inside client accounts | Developers |
| Data protection | Encryption, minimisation, retention, data location | Data protection lead |
| Files and documents | Scanning, signed links, type and size limits | Developers |
| Application security | Secure coding, dependency updates, security headers | Developers |
| Infrastructure | Hardened hosting, backups, access rules, secrets management | Operations |
| Monitoring and logging | Audit logs, alerts, log retention | Operations |
| Governance and response | Processor agreements, incident plan, testing, reviews | Management and data protection lead |
1. Identity and sign-in
- Multi-factor authentication offered to every user and required for staff and for client roles with access to sensitive data.
- Invitation-only accounts: no open registration; staff or client administrators invite users.
- Password rules that favour length over complexity, check against known breached passwords and allow password managers to paste.
- Throttling of failed sign-ins and password resets, with alerts on unusual patterns.
- Secure password reset by time-limited, single-use links.
- Session limits: sessions expire after inactivity, and users can sign out of all devices.
- Single sign-on for larger clients who want to use their own identity provider.
- Off-boarding: client administrators can remove users at once, and staff accounts are removed on the day someone leaves.

Is MFA required for a client portal?
Multi-factor authentication is not named as a specific obligation in the GDPR, but it is widely regarded as part of the state of the art for protecting accounts that reach personal data. For organisations covered by the EU’s NIS2 Directive, multi-factor authentication appears among the cybersecurity measures they are expected to use where appropriate. Clients increasingly ask for it in their own security questionnaires too.
In practice, require it for every staff account and for client users who can reach financial, health, legal or other sensitive data, and offer it to everyone else. Authenticator apps and passkeys are stronger than codes sent by text message.
2. Authorisation: every record, every file
Most serious portal breaches are not broken passwords but broken authorisation: a signed-in user changes a number in an address and sees another client’s invoice. This is why authorisation deserves its own section.
- Server-side checks on every request: the server confirms that the signed-in user may see or change this specific record, never relying on what the interface hides.
- Per-file authorisation: document downloads go through the application, which checks permission each time, or through short-lived signed links issued after that check.
- No guessable identifiers exposed where they could be used to enumerate other clients’ records.
- Roles inside client accounts: for example administrator, finance and viewer, each with clearly defined rights; our article on B2B customer portals shows how roles differ by sector.
- Least privilege for staff: support staff see what they need, not every client by default.
- Automated tests that try to read and change other clients’ data and must fail.
The OWASP Application Security Verification Standard is a useful reference for writing these requirements in testable form.
3. Data protection and privacy
- Encryption in transit with modern TLS on every page and API, and HTTP Strict Transport Security.
- Encryption at rest for databases, file storage and backups.
- Data minimisation: collect only what each portal task needs.
- Retention periods defined for each type of data, with automatic deletion.
- Data location: hosting and backups in the EU or in countries and arrangements that meet the GDPR’s rules on transfers.
- Records of processing updated to include the portal.
- A data protection impact assessment where the portal processes sensitive data at scale.
- Rights requests: a documented way to give clients access to, correct or delete their personal data.
4. Files and documents
- Malware scanning of every uploaded file before others can download it.
- Type and size limits on uploads, checked on the server.
- Separate storage for uploads, never executed or served from the application’s own domain without control.
- Version history for important documents, so a mistaken upload can be reversed.
- Watermarking or download logging for highly confidential documents, where clients expect it.
5. Application security
- Secure development practices: code review, protection against injection and cross-site scripting, safe handling of input.
- Dependency management: libraries and frameworks updated on a schedule, with alerts for known vulnerabilities.
- Security headers: content security policy, frame protection, referrer policy and similar.
- Secrets kept out of code: API keys and passwords in a secrets manager or protected environment settings.
- Rate limits on APIs and forms.
- Separate environments for development, testing and production, with no real client data in test environments unless properly protected.
6. Infrastructure and hosting
- Hardened servers or managed platforms, with only the needed services exposed.
- Regular patching of operating systems and runtimes.
- Backups taken automatically, stored separately, encrypted and restored in a test at least a few times a year.
- Access rules limiting administrative access, ideally through a VPN or restricted addresses.
- Protection against denial of service proportionate to the portal’s importance.
7. Monitoring and logging
- Audit logs of sign-ins, permission changes, downloads and edits, with user, time and source.
- Alerts for repeated failed sign-ins, unusual download volumes and administrative changes.
- Protected logs that ordinary administrators cannot alter.
- Log retention long enough to investigate incidents, and short enough to respect data minimisation.
- Client-facing activity history, so clients can see who in their own team did what.
8. Governance, testing and incident response
- Processor agreements with hosting, email, analytics and support tool providers.
- Security testing before launch, including an independent penetration test for portals with sensitive data.
- An incident response plan: who decides, who investigates, who informs clients and the supervisory authority, and within what time. Under Article 33 of the GDPR, a notifiable personal data breach must be reported to the supervisory authority within 72 hours of becoming aware of it.
- A vulnerability disclosure contact, such as a security.txt file and an address that someone reads.
- Staff training on phishing and on handling client data.
- Accessibility checked alongside security, since secure flows such as MFA must also be usable by everyone; our article on app accessibility and the European Accessibility Act covers the requirements.
Who should own client portal security inside the company?
Security fails when everyone assumes someone else owns it. Name owners for each area:
- An accountable executive, usually the managing director or operations lead, who approves the risk level and budget.
- A product owner for the portal, who keeps the checklist in the backlog and decides trade-offs.
- A data protection lead or the data protection officer where one is required, who owns processing records, assessments and rights requests.
- The development team, internal or external, responsible for application and authorisation controls.
- Operations or the hosting provider, responsible for infrastructure, backups and monitoring.
If you work with an external developer, write the checklist into the contract and the acceptance criteria. Our guide to client portal development explains how security fits into the wider build.
How often should the checklist be reviewed?
Review it on three rhythms:
- Continuously: dependency alerts, monitoring and failed sign-in alerts are handled as they arrive.
- At every significant release: new features are checked against authorisation and data protection controls before they go live.
- At least once a year: a full review of the checklist, user access, processor agreements, backups and the incident plan, plus a fresh penetration test for portals with sensitive data.
Also review it after any incident, any change of hosting or major supplier, and any change in the kind of data the portal holds.
Questions people ask
What security does a client portal need?
A client portal needs strong sign-in with multi-factor authentication, server-side authorisation on every record and file, encryption in transit and at rest, audit logging, secure development and patching, tested backups, and an incident plan. European companies also need lawful data location, processor agreements and a way to handle data protection rights requests under the GDPR.
Is multi-factor authentication required for a client portal?
The GDPR does not name it explicitly, but it asks for security appropriate to the risk and the state of the art, and multi-factor authentication is widely seen as part of that. Organisations covered by the NIS2 Directive are expected to use it where appropriate. Require it for staff and sensitive client roles, and offer it to everyone.
What is the most common client portal security flaw?
Broken authorisation is among the most common and serious flaws: a signed-in user can reach another client’s records or files by changing an identifier in a request. The fix is to check permission on the server for every record and every file download, and to run automated tests that try to cross between client accounts.
Who should own client portal security?
An accountable executive approves the risk level and budget, a product owner keeps the checklist in the backlog, a data protection lead owns processing records and rights requests, developers own application controls and operations owns hosting, backups and monitoring. When an external developer builds the portal, the checklist belongs in the contract.
How often should client portal security be reviewed?
Handle alerts continuously, check every significant release against the authorisation and data protection controls, and run a full review at least once a year with access checks, backup restores, processor agreements and a fresh penetration test for sensitive portals. Review again after incidents, supplier changes or a change in the data held.
Copy the eight areas into your backlog, mark each control as done, partial or missing, and give every gap an owner and a date. If you need a portal built or hardened to this standard, see our application development service or tell us about your portal.


