Does downtime hurt your Google ranking?

The panic is reasonable. You spent months getting people to find this thing, some of them arrive through Google, and for a stretch this morning the site showed a spinner or a blank white screen or a page saying something went wrong.

The fear is that Google walked past while the lights were off, drew a conclusion, and quietly demoted you. That is not really how it works. The crawler comes back.

How long can a site be down before Google drops it?

About two days. Google's own guidance for site owners is unusually specific here, which is lucky, because almost everything else about ranking is vague.

When Google's crawler asks for a page and the server answers with an error instead of a page, it retries those addresses for roughly two days — but returning those "not available right now" responses for longer than that will cause Google to drop the pages from its index.¹

Underneath that there's a softer effect that starts earlier. Google slows down how often it crawls your site when it hits a significant number of pages returning server errors, and that slowdown applies to the whole site — the working pages too. The good news is in the next sentence of the same document: once the errors stop, the crawl rate automatically starts increasing again.²

Gary Illyes, who has worked on search at Google for years, put a number on the harmless end. Asked about a site that intermittently returns the "temporarily unavailable" response, he said serving it for an extended period will cause a decrease in crawl rate — but ten to fifteen minutes every now and then is not "extended" by any means.³

John Mueller, who spent years answering exactly this question from panicking site owners, described the first phase the same way: if a page returns a server error or the site is unreachable, Google retries over the next day or so, and nothing happens — no drop in indexing or ranking — until a few days have passed.⁴

What does the damage look like hour by hour?

• Minutes. Nothing. Genuinely nothing.
• A few hours. Nothing you will be able to detect.
• A day. Crawling slows down. Rankings generally hold.
• Two days and beyond. Pages start falling out of search results.
• A week. You now have a real recovery job, not an outage.

Do rankings come back after an outage?

Yes, in most cases, and getting the app working again is the entire technical remedy. Mueller was asked directly about a site that had been down almost five days and lost a large chunk of its daily search traffic. His answer was that when things come back — assuming that's within the range of days — rankings bounce back too.⁴

He attached a warning worth taking seriously, because it's the opposite of reassuring and that's exactly why it's credible: don't assume a ranking drop after a temporary indexing drop will fix itself. That's something you need to address, not wait out.⁴

His reasoning was that if you're still seeing a drop after indexing recovers, it's probably not the outage. Google makes changes regularly and reevaluates sites over time, and while being offline wouldn't trigger that, it can happen around the same time anyway.⁴

Translated: if you were down on Tuesday and traffic is still flat the following Tuesday, stop blaming Tuesday. Something else moved, and you'll waste a month staring at the wrong week.

Why does Search Console look normal the day it happened?

Because the report hasn't caught up yet. Google's reporting for site owners is rarely live and often lags by roughly 48 hours, so a flat line today can simply mean the numbers haven't arrived.⁵

So if you open it the afternoon of an outage, you'll either see a perfectly calm graph and conclude you got away with it, or see nothing and conclude the tool is broken. Neither is true. Check it later in the week.

What kind of broken actually costs you rankings?

The quiet kind. A clean outage — server unreachable, error returned, everyone can see it — is something Google handles gracefully. The worse case is an app that is broken but looks fine to a machine.

Your front end loads. The page reports success. The content is a friendly little "Oops, something went wrong, try refreshing" box, because whoever wrote that component was being polite.

Google sees a page that loaded successfully and contains almost nothing. Pages that report success while showing an error or barely any content get treated as a soft 404 — Google's term for a page it decides is effectively missing.⁶

That's the failure mode that catches apps built with AI tools specifically, because those tools are good at generating graceful error screens and nobody ever checks what those screens report to a crawler. It's also the one you're least likely to notice, because the app looks alive when you open it on your phone.

Audit every error and fallback screen in this app. When the page can't load its data or hits an error, the server response should say the page failed rather than reporting success with an empty or "something went wrong" message. List each place where we currently show an error message on a page that still reports success, and fix them.

What should you actually be worried about?

Not the ranking. If your app was down for a morning, the loss is the people who arrived that morning and left. Signups that didn't happen. A customer who tried to log in at the worst possible moment. Those don't come back when the server does.

And the number that matters most isn't how long you were down — it's how long you were down before you knew. Everything above is survivable at hours and dangerous at days, and the difference is almost never how long the fix took. It's how long the app sat broken while you were asleep, at dinner, or building the next feature.

That gap is what Fomio is for. You paste a link to your app and it watches it from outside, the way a visitor would, and emails you when it stops working. It reports one of three things — Working, Having trouble or Down — and when something breaks it says in plain English what likely went wrong instead of handing you a log file. There's nothing to configure.

The free plan watches one app, checks it every five minutes, emails you when it breaks and keeps a week of history, with no card needed. The Builder plan watches three apps, checks every minute, keeps thirty days, and writes a prompt you can paste straight into the tool you built with — so the fix starts from something better than "it's broken." Current prices are on fomio.ai/pricing.

What should you do right now?

1. 1
   If you're mid-outage: fix the app

   That's it. That is the search work. Nothing else you do today moves the needle.
2. 2
   If it was yesterday: do nothing

   Don't resubmit your sitemap, don't mass-request indexing, don't rewrite content. Crawling recovers on its own once the errors stop, and touching things adds noise to a week where you want signal.
3. 3
   Check your error screens

   Make sure a page that failed says it failed, rather than reporting success with an apologetic message on it.
4. 4
   If it's a week later and traffic is still flat: look elsewhere

   It probably isn't the outage. Something else changed around the same time.
5. 5
   Close the gap between breaking and knowing

   If you found out from a user that the app had been down since yesterday, that's the thing to fix — and it isn't a code problem.