A website can look healthy while an important function has stopped working. The homepage may load even though enquiry emails are missing or a download link is broken. Maintenance should therefore be organised around what the business needs the website to do, rather than a quick visual check once in a while.
Identify the journeys that matter
List the actions a customer must be able to complete. For a service business, that might mean finding a relevant page and submitting an enquiry. For a digital shop, it includes selecting a product, completing the order and receiving access to the file. Give each action a simple test and an owner.
Keep an administrative record of the domain, hosting, email and backup services. Record who is responsible and how ownership can be recovered. Store passwords securely rather than adding them to an ordinary shared project document. Access should not depend entirely on one person’s memory.
Plan for recovery, not just backup creation
WordPress documentation explains that a complete backup involves both site files and the database. Before significant changes, retain an appropriate backup and test recovery in a separate environment. A successful backup notification is not the same as a demonstrated restore.
Decide how much recent work or order data the business can afford to lose and how long an outage would be tolerable. These questions inform backup frequency and retention. A frequently updated store and a small informational site will not necessarily need the same operating routine.
Make changes traceable
Review release notes and relevant compatibility information before an important update. Use a staging environment for larger changes when available. After updating, test the customer journeys you identified instead of assuming that the absence of an error message means everything works.
- Send a test enquiry and confirm its arrival.
- Check navigation, language choices and mobile layouts.
- Review key images and internal links.
- For a store, verify an appropriate test order and delivery flow.
Record what changed, when it changed and whether the checks passed. If something fails later, this small record helps narrow the investigation. Avoid changing many unrelated settings at once when diagnosing a problem.
Include editorial maintenance
Technical uptime does not make outdated information useful. Review contact details, service scope, product requirements and pricing descriptions. If a product has changed, update the screenshots and any articles that explain the old version. Check that the language versions still agree on essential facts.
Maintain a short issue log with symptoms, dates, affected pages and the outcome of each fix. This turns support into a repeatable process and reduces the chance of repeating an unsuccessful adjustment without new evidence.
Define responsibilities before an outage
If another professional maintains the site, agree on response expectations, backup restoration responsibilities and the way support requests are submitted. Distinguish routine maintenance from new feature development. A clear agreement makes it easier to prioritise incidents and understand the ongoing cost of keeping the website useful.