Does Website Downtime Hurt SEO? What Actually Happens Hour by Hour

Website downtime and SEO: what Google actually does when your site returns errors, how long you have before rankings drop, and how to catch it early.

D

Sathi··5 min read

Your site just went down. Somewhere between the panicked Slack messages and the hosting dashboard, the SEO question surfaces: is this killing our rankings right now?

Short answer: website downtime hurts SEO on a timer, and the timer is more forgiving than most people fear - but far less forgiving than "it'll be fine." Here's what Google actually does when your site stops responding, hour by hour, according to Google's own documentation and statements.

What Google does when your site returns errors

Google's crawling behavior around downtime is documented, not mysterious. Per Google's HTTP status code documentation, when Googlebot hits 5xx server errors or timeouts, it slows its crawl rate, retries later, and only starts removing URLs from the index when the errors persist for days.

Google's John Mueller has been explicit about the timeline: nothing happens to indexing or ranking until a few days have passed - Googlebot retries over the next day or so, and a site that's back within roughly a day sees no lasting effect.

The website downtime timeline, hour by hour

DowntimeWhat happensLasting SEO damage
Minutes to a few hoursGooglebot retries; users bounceEffectively none
~1 dayCrawl rate drops; possible brief ranking flux that recoversMinimal, recovers on its own
2-4 daysURLs start dropping from the indexReal - recovery takes weeks
A week+Widespread deindexing; crawl frequency deeply reducedSevere - a slow, painful rebuild

Two nuances worth knowing:

The flux period. Even a single day of downtime can produce one to three weeks of ranking wobble before things settle back, per Mueller - alarming to watch, but temporary if the site is stable again.

Crawl-rate memory. Repeated short outages train Googlebot to visit less often. Analyses of server-error impact show affected pages receiving fewer crawls per day afterward - which means slower indexing of everything new you publish, a cost that never shows up in a rankings report.

The right way to go down: 503, not 200 or 404

If you have any control during an outage or planned maintenance, serve HTTP 503 (Service Unavailable) with a Retry-After header. A 503 tells crawlers "temporary - come back," and Google treats it exactly that way.

The genuinely destructive failure modes are the dishonest ones:

  • Serving 200 with an error page - Google indexes your "Something went wrong" page as your content
  • Serving 404/410 - Google takes you at your word that the pages are permanently gone and drops them fast
  • DNS failures - worse than server errors, because there's nothing to retry against

Mangools' downtime breakdown makes the same point from the other direction: the status code you serve while down matters nearly as much as the downtime itself.

The compounding costs nobody attributes to downtime

Rankings are the visible casualty. The quiet ones: users who bounce off a dead page and don't return, link-building outreach that lands while your site is unreachable, and - increasingly relevant in 2026 - AI crawlers hitting errors. ChatGPT fetching your pricing page mid-conversation doesn't retry tomorrow; it answers from a competitor's page right now. The same visit patterns we covered in our bot traffic breakdown apply during outages: the majority of your readers hit that error page, and most of them aren't patient humans.

Broken individual pages behave like miniature downtime for their URL, with similar deindexing mechanics - our post on broken links and their SEO cost covers that half of the problem.

Catching downtime before Google does

The whole game is noticing within minutes, not days - the damage curve is steep only after day one.

1. Monitor from outside your own infrastructure. If your monitoring lives on the server that's down, it goes down too. DidYouSEO's uptime and URL monitoring checks your key pages externally and alerts you on downtime, status-code changes, and content changes.

2. Monitor the pages that matter, not just the homepage. A dead checkout, docs section, or sitemap can 5xx for days while the homepage looks fine.

3. After recovery, check crawl stats. In Google Search Console's Crawl Stats report, confirm the error rate dropped and crawl volume recovers over the following week or two. If URLs were deindexed, request indexing for the most important ones rather than waiting.

FAQ

How long can my website be down before SEO is affected? Roughly a day is the documented grace period - Googlebot retries for about that long before anything happens to indexing or ranking. Past two or more days of persistent errors, URLs start dropping from the index.

Will my rankings come back after downtime? After short downtime (a day or less), yes - typically within one to three weeks of flux. After long outages that caused deindexing, recovery is slower: pages must be recrawled and re-ranked, which can take weeks and doesn't always restore the exact prior positions.

Does downtime at night matter less for SEO? No. Googlebot doesn't keep business hours, and neither do AI crawlers - crawl visits are spread around the clock. "Low traffic hours" only reduces the human cost, not the crawler cost.

Is a slow site treated like a down site? Not identically, but extreme slowness causes timeouts, which Google handles like server errors - and sustained slowness reduces crawl rate. Downtime and severe latency sit on the same spectrum.

See how your site scores - free

30+ SEO checks plus AI visibility. No credit card required.

Run a free audit