In this guide: Website security audit in 30 minutes: check accounts and access, software updates, backups, SSL and security headers, and malware scanning — Week 3 recap.
This guide organizes a preliminary security review for a WordPress website. It is not evidence of a completed audit, a penetration test or a guarantee that the site is free of vulnerabilities.
Define an authorized scope, collect evidence and record unresolved questions. The time required depends on the environment and depth of review.
See How to Set Up Automatic Backups.
Strict-Transport-Security (HSTS)X-Content-Type-Options: nosniffReferrer-PolicyX-Frame-Options or CSP frame-ancestorsContent-Security-Policy (advanced — test carefully)DISALLOW_FILE_EDIT).site:yourdomain.com for spam pages./wp-content/uploads/ for PHP files.See How to Scan Your Website for Malware.
Do not assign a security rating from the number of unchecked boxes. One serious access or recovery problem can matter more than several cosmetic gaps. Record each issue, evidence, possible consequence, responsible owner and next verification step.
Schedule it: repeat this audit monthly or quarterly. Security isn’t a one-time project — it’s a habit.
Next week’s theme: growth and monetization — how to turn your website into a business.
Further reading: OWASP Top 10 web security risks
List the website, accounts, integrations and environments included in the review. Identify who owns them and who authorizes the work. A public website can depend on hosting, domain registration, mail services and third-party applications that require different access and expertise.
The OWASP Web Security Testing Guide is a primary reference for structured application testing. A preliminary owner checklist covers a narrower scope. Do not present it as equivalent to an authorized professional assessment of every endpoint, integration and business process.
Keep observation separate from changes. Reviewing a setting does not automatically authorize disabling it, replacing software or testing disruptive behaviour. Record proposed remediation with its owner and verification plan instead of turning every observation into an immediate production edit.
Identify the people and service accounts that can affect the site, domain or recovery route. Record why access is needed and who approves it. Shared credentials and undocumented former-contractor access make ownership difficult to understand even when the public pages appear normal.
Review permissions according to the actual task and data involved. Our WordPress user roles guide distinguishes role labels from capabilities and the need for controlled checks. A hidden menu is not proof that the underlying action is restricted.
Document authentication and recovery separately. A strong login arrangement does not justify excessive permissions, and removing a person must not accidentally remove required content. Use an approved offboarding process and keep credentials out of the review report.
Record the components, versions and responsible maintainers rather than relying on a green dashboard alone. Identify unsupported, unmaintained or unfamiliar components for investigation. A version number can help locate relevant guidance, but does not establish the whole risk picture by itself.
WordPress’s hardening guidance describes a layered approach. Keep the review proportionate and avoid claims that any single plugin or setting guarantees protection. Application code, credentials, hosting and operations all affect the environment.
Use the WordPress update workflow for compatibility and recovery planning. Do not update every component blindly during an audit merely to clear a warning. The change needs its own approved scope, evidence and acceptance checks.
Check what the backup actually includes, its timestamp, storage arrangement and who can restore it. The existence of a backup job is not the same as a demonstrated recovery of the required files and database. Keep a suitable restoration rehearsal separate from production activity.
For live transactions, consider data created after the backup. A broad database restore can remove newer orders, users and content. A security review should record how the business would preserve or reconcile that activity rather than assuming that restoring the whole site is always safe.
The WordPress backup reference and our backup comparison guide provide supporting questions. Define recovery needs from the actual business rather than a universal retention number.
Check representative public and authenticated journeys for the intended secure connection and obvious mixed-content issues. Record the actual result and conditions. A valid certificate supports transport protection; it does not prove that application permissions, plugins or business logic are secure.
Header changes need careful planning. MDN’s HSTS reference explains browser enforcement and subdomain scope. Do not enable broad subdomain or preload behaviour simply to improve a scanner grade without understanding the operational consequences.
A restrictive content policy can also affect legitimate scripts and integrations. Treat scanner output as evidence to interpret, not a command to paste a universal header set into production. Assign the relevant developer or infrastructure owner to plan and verify changes.
A remote scan observes a narrower view than a server-side investigation. A clean result does not establish that every file, account or application path is safe. Likewise, an alert needs validation before it becomes a public claim that a site is compromised.
Record which tool was used, the scope, date and result. Separate a detected indicator from a confirmed incident. Avoid uploading private files, logs or credentials to an unrelated online service simply because it offers a convenient score.
Do not delete unfamiliar files during an initial review without understanding their purpose and preserving appropriate evidence. If an incident is suspected, follow the responsible response process. Our website malware guide provides useful conceptual boundaries.
Ask what the provider manages and what remains the customer’s responsibility. Account separation, resource management, application maintenance and recovery are different concerns. A hosting category or product name does not establish the exact control available.
Our server isolation guide explains why containers, permissions and resource limits should not be treated as interchangeable guarantees. Request plan-specific evidence where the boundary matters to the business.
Keep unconfirmed provider claims separate from independently observed results. A positive support answer can inform the review, but it is not a complete audit of the infrastructure. Define the additional assurance needed when the workload or data requires specialist assessment.
Inventory external applications, forms, webhooks and sending services that receive or change business data. Record the owner and purpose of each connection. An old integration can remain active after the person who installed it leaves the business.
Review the scope of access and the approved data flow. Do not assume that a trusted brand makes every connection appropriate. Keep secret values and private diagnostic records out of public screenshots and reports.
For mail, use the SPF, DKIM and DMARC guide and SMTP troubleshooting workflow. Authentication results, provider acceptance and inbox placement are distinct observations; one positive check does not certify the whole communication system.
Do not count unchecked boxes and label the site excellent or high risk without evaluating the actual findings. A single exposed privileged account can matter more than several informational observations. State the evidence, affected asset, possible consequence and confidence level.
Assign a responsible owner and next step. Some items need immediate escalation, some need an approved maintenance change and others need clarification. Keep uncertain findings open rather than forcing every row into a simplistic pass or fail.
Our project workflow guide helps organize accountability. A useful issue register should support decisions and follow-up, not create an impressive total that hides the most important unresolved item.
Summarize what was examined, what was observed and what was not tested. Separate completed observations from proposed remediation and verified fixes. A review is not complete merely because a checklist was filled out or an automated tool displayed a favourable score.
Revisit the findings after material changes and according to the risk and operating needs of the site. A recurring calendar can help, but a new integration or departing administrator may require review before the next scheduled date.
Keep an appropriate response and communication route for suspected incidents. Our incident communication guide explains how to state verified impact without unsupported assurances about customer data. This article does not diagnose the current website or certify it as secure.
Promotional recommendation: KAWEDMA offers website development and hosting services to consider for an agreed maintenance or implementation scope. Ask which checks, updates, backups and recovery tasks are included. A service recommendation is not evidence that a particular site passed a security assessment.
Affiliate disclosure: Hostreta may earn a commission if you purchase through this link. If managed WordPress infrastructure fits your requirements, compare hosting.com’s managed WordPress offering with alternatives and current plan terms. Verify the actual management and recovery boundaries rather than assuming that managed means every application risk is covered.
Consider a small service website with two editors, an external developer and a contact form. The business wants a preliminary review of access, maintenance and recovery. This is a fictional educational example, not a report that Hostreta examined a client’s infrastructure or found a vulnerability. The exercise begins with an approved scope and a responsible business representative.
The reviewer first creates an asset and ownership list. It includes the public site, domain account, hosting account, business mailbox and the integration receiving enquiries. The list records the relevant owner without copying passwords or personal recovery details into the report. If ownership is uncertain, that is an open issue rather than a reason to claim broader access.
The next step examines the editorial workflow. The reviewer asks which tasks each person performs and whether the selected access matches them. A former contractor account appears in the inventory, but the reviewer does not immediately delete it or its content. The business first confirms the account’s purpose, remaining responsibilities and the supported offboarding process.
The maintenance record reveals a component awaiting a compatibility check. That observation is not automatically labelled a confirmed exploit. The reviewer records the actual version, relevant vendor information and the owner responsible for a safe change. If reliable evidence indicates an urgent issue, the team escalates it through the appropriate response process rather than waiting for a cosmetic checklist review.
The recovery discussion identifies a recent backup, but no suitable restoration evidence. The report distinguishes those facts clearly. It does not say there are no backups, and it does not say recovery is proven. The next action is to define a safe rehearsal and check which files and database records are covered, while protecting current production activity.
A remote scanner then returns an informational header warning. The reviewer records the output and checks the relevant configuration context. They do not paste an aggressive content policy or broad HSTS setting into production simply to change the grade. The issue needs an owner, an explanation of the intended protection and a suitable verification plan for legitimate site behaviour.
The enquiry integration is assessed next. The team identifies the destination, approved data flow and responsible contact. Sample details are used for any controlled functional check. Private customer messages are not uploaded to a public tool as evidence. If a permission or retention question remains unresolved, the report preserves that uncertainty and assigns it to the appropriate owner.
When prioritizing the findings, the reviewer compares consequences rather than counting rows. Unclear privileged access may deserve earlier attention than an informational presentation warning. A missing recovery rehearsal also has a different significance from a confirmed loss of backup coverage. Each item needs an evidence-based explanation that another person can understand and challenge.
The business receives a short summary naming the scope, observations, proposed work and limitations. The report states that it was not a complete penetration test and does not certify the absence of compromise. It separates changes already verified from actions merely recommended. This prevents an attractive summary from becoming an unsupported assurance to customers or management.
Finally, the team schedules follow-up around the actual changes and responsibilities. A fixed monthly or quarterly rhythm may help organize work, but it should not postpone review of a new integration, a departing administrator or a suspected incident. The review process remains connected to the way the website operates rather than a calendar alone.
Keep the exercise record as a practical template. Replace the fictional details with authorized, verified observations when performing a real review, and retain only the information needed for the purpose. A transparent statement of what was not tested helps the next responsible person make better decisions than a simple pass badge or an unexplained total score.
Document the review date and accountable follow-up owner.
We may earn a commission on purchases made through these links, at no extra cost to you.
Don’t miss the chance to participate in the biggest giveaway of this year.