Pausing a website
The recommended method is to keep the site online with reduced functionality: disable the cart, show a banner, adjust the structured data. Taking the whole site down is described as an extreme measure, for a few days at most, whose recovery has neither a stated duration nor any way to be sped up.
The reflex, and what the documentation makes of it
A shop that cannot ship, a craftsman closing for three weeks, a service suspended while it is rebuilt: the first reflex is to take the site down. The documentation devotes a whole page to explaining why that is almost always the wrong answer.
The recommendation is to keep the site online with limited functionality rather than close it: disabling the cart changes nothing for the site's visibility in Search, and visitors can still find products and read reviews.
officialSource : Google Search Central, “Temporarily pause or disable a website”
The important half is the second: disabling the cart changes nothing for visibility. What blocks trade — the impossibility of ordering — has nothing to do with what keeps the site present in the results, namely its pages.
A banner or popup on every page makes the situation clear, and the data-nosnippet attribute keeps it out of search snippets; structured data must be adjusted to reflect real availability or a cancelled event.
officialSource : Google Search Central, “Temporarily pause or disable a website”
The detail about the exclusion attribute is the kind of precision that separates documentation from an article: without it, the “we are closed” banner ends up in the displayed snippet, and stays there longer than the closure.
What a complete shutdown costs
Removing a site completely from the index is described as a significant change that can take quite some time to recover from: no duration is stated, and there is no mechanism to speed that recovery up.
officialSource : Google Search Central, “Temporarily pause or disable a website”
Two rare pieces of information sit in that sentence: no duration is stated, and no mechanism exists to speed the return up. It is one of the few places in this corpus where the vendor explicitly writes that it can do nothing for you.
The page also lists the side effects, none of them technical: customers no longer know what is happening, first-hand information gives way to what third parties say, knowledge panels lose elements, and console verification eventually fails, which removes access to the data.
Annual closure: we noindex your site for three weeks so Google doesn't see an inactive shop, then we put it back.
The one defensible shutdown, and its conditions
If the site must urgently be taken down for one or two days, the documentation asks for an informational page returning a 503 status code; for a longer period it asks for an indexable home page returning 200 rather than a shutdown.
officialSource : Google Search Central, “Temporarily pause or disable a website”
One or two days, and an informational page rather than a blank one. The documentation even details the shape of that page: a header saying when to retry, static HTML, inline styles and embedded images to keep the load down, and above all useful text — the expected return date and a way to reach somebody.
During a shutdown, crawling must remain allowed: the robots.txt file must not return 503, which blocks all crawling, and a valid file disallowing everything may remove the site's content and its URLs.
officialSource : Google Search Central, “Temporarily pause or disable a website”
This is the mistake that turns a two-day shutdown into a multi-week incident. A server configured to return 503 for every address also returns 503 for robots.txt, and that file governs the rest.
Four means are explicitly advised against for closing a site: returning 403, 404 or 410, applying a noindex rule by tag or header, disallowing all crawling in robots.txt, and using the temporary removal tool, which would stop visitors from learning the situation.
officialSource : Google Search Central, “Temporarily pause or disable a website”
The last point deserves a remark: the documentation notes that a site removed from the results leaves room for others, and that resellers or affiliates offering the same products may keep appearing. Closing does not suspend the competition, it leaves it alone on the field.
While a page returns 503, titles, descriptions, metadata and structured data cannot be refreshed: what is displayed stays as it was at the last successful crawl.
officialSource : Google Search Central, “Temporarily pause or disable a website”
Reducing the site's functionality rather than taking it down during a closure
hypothesis
- Effort
- low
- Time to effect
- immediate
- Risk
- none
- Durability
- lasting
The effort is low: a banner, a disabled cart, adjusted structured data. The delay is immediate on the visitor’s side. The risk is nil, and that is exactly what separates this approach from a shutdown, whose recovery the documentation says has neither a known duration nor an accelerator. Durability is medium, the measure being provisional by nature.
What does not work, or no longer works
Taking the site down so as not to show an inactive shop. The documentation recommends the opposite, and disabling the cart changes nothing for visibility.
Then returning 404 or applying a noindex rule for the duration of the closure. Those means remove the URLs from Search, which is exactly the outcome you were trying to avoid.
Finally, letting the server return 503 for everything, robots.txt included. That file is what allows crawling to resume, and cutting it turns a pause into an absence.
Sources
- Temporarily pause or disable a website — Google Search Centralaccessed on
- How HTTP status codes, and network and DNS errors, affect Google Search — Google Search Centralaccessed on
Block revalidated on
Frequently asked questions
- Should you close your site over the holidays?
- The documentation recommends the opposite: keep the site online with reduced functionality. Disabling the cart changes nothing for visibility in Search, and visitors keep finding products, reading reviews and preparing a later purchase. A banner is enough to state the situation.
- How long can a site be down without harm?
- The documentation speaks of one or two days at most for a 503 shutdown, and calls complete closure an extreme measure. Beyond that it asks for an indexable home page returning 200. No recovery duration is stated for a site removed from the index.
- Must robots.txt keep answering during a shutdown?
- Yes, and it is the most common trap. A robots.txt returning 503 blocks all crawling. A valid file disallowing everything may remove the content and the URLs. During a shutdown that file must stay reachable and permissive.
- Can the removal tool be used for a closure?
- The documentation explicitly advises against it: visitors would no longer find the site to learn its status. It adds a side effect rarely anticipated — resellers or affiliates offering the same products may keep appearing in the results while you do not.
- Why is my description still out of date after reopening?
- Because a page returning 503 does not allow titles, descriptions, metadata or structured data to be refreshed. What is displayed dates from the last successful crawl. Updating resumes once pages are served normally again, at the pace of crawling.