A website can appear healthy while its backups fail, its dependencies fall out of support or its main inquiry form stops reaching anyone. Maintenance is the work of finding those problems while they are still manageable. Start with the functions customers and staff depend on most.

Know what needs updates

A website may use a framework, a content system, plugins, hosted services and third-party integrations. Record which components your team maintains and which are managed by providers. An automatically updated service can still require you to review connected accounts or test changes in your own site.

The Canadian Centre for Cyber Security recommends patching operating systems and applications as part of baseline controls. Give someone responsibility for tracking updates, checking compatibility and deploying them. Separate urgent security fixes from routine improvements and agree who can approve work when someone is away.

Source: Canadian Centre for Cyber Security baseline controls

Test what customers actually do

Submit an inquiry, check who receives it and read the confirmation. Try a booking or test order when appropriate. Check that product or service information still agrees with the workflow behind it. A home page returning successfully tells you little about a form that stopped sending messages.

Document expected behaviour so anyone on the team can spot a failure. If an inquiry should reach a shared mailbox and generate an acknowledgement, verify both. Monitor the things you cannot easily see every day, such as domain and certificate renewals and integrations that depend on an external account.

Watch for performance drift

A perfectly usable site can become slower as staff add large images, promotional banners or scripts. Compare a small group of priority pages over time and after major edits. Test on a phone and note where people have to wait or retry.

Use the observed problem to guide the change. Resize a heavy image, question a rarely used embed, or review a slow service response. Repeat the same task after the fix. Add a simple image and embed review to publishing so the next update starts with better defaults.

Sources: web.dev Web Vitals · web.dev image performance guidance

Decide what to check and how often

Separate high-impact customer journeys from ordinary housekeeping. An online store might routinely review order placement, payment confirmation and incoming fulfillment records. A brochure site may prioritize inquiry delivery, availability and account renewals. Agree on what requires a quick response and what can wait for the next scheduled maintenance window.

A short record helps someone else take over. Include the page or service, what was checked, who ran the check and what happened. If a problem cannot be fixed immediately, record how customers can get help in the meantime and who will follow up. Keep repeatable checks close to the work so they become part of publishing and updates.

Make recovery a rehearsed task

Identify what you would need to restore: site code, uploaded files, content, orders or records, and access to services. Write down where backups are kept, who can use them and how far back you can recover. A restore drill in a suitable environment can reveal missing files or credentials while there is time to fix the process.

Discuss what the business would do during downtime. Decide which contact channels to keep available and who updates clients. The length of an outage that a portfolio can tolerate may be very different from the length an order system can tolerate. Match the maintenance plan to those operational priorities.

Run a maintenance check from the customer side

Take a business whose site sends quote requests to a shared inbox. After a form update, try it on a phone with both valid and incomplete details. Check that instructions are readable, the message arrives, and the team can reply. Record who gets alerted when delivery stops and where the change is documented.

Next restore a recent backup in a separate environment and compare an important page against the live version. Verify that files, configuration and data appear as expected. Use the gap list to update the backup process while people with the right access are available.

Source: Canadian Centre for Cyber Security baseline controls

Your Action Checklist

  • List critical website functions and schedule checks with named owners.
  • Run a restore drill and record what it recovers.

NSBDA can help set up a maintenance plan around your site’s most important customer tasks and recovery needs.

Explore Website Maintenance