Pause only the functions that need to stop
Keeping the website available while limiting the affected function is often the most stable option. A shop can pause purchases but retain products, hours and contact details. A booking business can stop new appointments while keeping services, location and phone visible. Customers get useful information and search engines can still access valid pages.
Identify the work before enabling site-wide maintenance. A brief update, database migration and several-week rebuild need different responses. Closing every route for technical convenience removes information from people who only wanted a phone number, address or service detail.
Prepare an alternative route. The temporary message should say what is available, when service is expected to return and how to contact you. A vague ‘back soon’ gives the customer no next step and makes an accidentally lingering notice harder to spot.
- Function that must pause
- Information that can stay public
- Tested alternative contact channel
- Owner and expected reopening time
Choose the HTTP response for the duration
For an urgent one or two-day interruption, the server can return an information page with HTTP 503: temporarily unavailable. Retry-After can indicate when to try again. Visitors get an explanation and search engines receive a different signal from a deleted page.
503 is not a solution for weeks. Prolonged responses can eventually remove URLs from the index, and Google cannot refresh their content during the interruption. For longer pauses, keep a useful homepage returning 200, with current information and links to working sections.
Returning the same maintenance screen with 200 at every URL creates another problem. The server says everything works, and a search engine may treat the temporary text as the new content of each service or article. Check the actual HTTP response, not just its appearance.
Check noindex, robots.txt and response headers
A noindex directive asks search engines to exclude a page from results. It can be intentional for a draft, private area or demo, but it is not a harmless pause for a business page that should return. The route remains unindexable until the directive is removed and processed again.
Blocking crawling in robots.txt is different from temporarily taking a page offline. A crawler that cannot enter cannot read a noindex inside either. Google's guidance is to keep robots.txt accessible during a pause and avoid a site-wide crawl block as a substitute for a temporary response.
Check HTML and headers: noindex may appear in a meta tag or an invisible X-Robots-Tag header. Review templates, plugins and hosting rules too. Clearing one CMS checkbox does not ensure another layer has stopped sending the directive.
- Meta robots on public pages
- X-Robots-Tag header
- robots.txt rules
- Hosting and maintenance-mode settings
Test service and article URLs as well as the homepage
Sample services, articles, contact pages and URLs receiving campaign or older-link traffic. Maintenance can affect the homepage but leave other routes exposed, or restore the homepage while important pages still return 503 or noindex.
Check the sitemap, canonicals and redirects. The sitemap should reflect public URLs you want to retain; canonicals should not accidentally point to the temporary screen. Contact links, phone numbers and forms must lead to real destinations.
Test signed out, on mobile and in each published language. An administrator's cookie may bypass maintenance and show a site the public still cannot access. Keep a short list of critical URLs and expected responses for the reopening check.
Reopen with checks you can repeat
Remove the temporary screen and confirm every critical route returns 200 with its real content. Remove inappropriate noindex and X-Robots-Tag directives, check robots.txt and open the sitemap. Then test contact, bookings, purchases and any paused function.
Search Console URL Inspection lets you compare Google's known version with the published page and run a live test. Once indexable, you can request crawling for a few priority URLs. For a larger group, update the sitemap instead of making individual requests.
Reopening is complete when an external visitor can find the information and finish the main action, and the owner has verified temporary signals are gone. Keep that evidence with the date and change made so the next pause begins with a procedure.
- Real content returning 200
- Temporary directives removed
- Accessible sitemap and critical URLs
- Main action tested from outside
