Why doesn't Google consider a single site's ranking drop as a system-wide incident?

What is Google's definition of an incident?

An incident refers to a technical malfunction in the search engine that strikes multiple sites at the same time. Typically: a massive indexing bug, a crawl system outage, or a display problem in the SERPs affecting an entire category of results.

Google only intervenes in these cases — because they stem from a problem with their own infrastructure. If you crash alone, in Google's eyes, it's not their problem to fix.

Why doesn't a single site losing its ranking count as an incident?

Because Google's logic is based on the following assumption: if a single site tanks, the cause is specific to that site. A noindex tag added by mistake, a structural change, a targeted algorithmic penalty, a perceived quality drop.

It's a defensive stance — but not illogical. A global incident is detected by its scale. An isolated drop requires an internal diagnosis before pointing fingers at the algorithm.

What are the implications for an SEO practitioner?

Concretely, this means you can no longer hide behind a supposed Google bug whenever a client loses positions. You must prove that other sites are affected simultaneously to qualify the problem as an incident.

Otherwise, the fault is presumed internal — and it's up to you to dig deeper: server logs, recent technical changes, accidental deindexing, lost backlinks, duplicate content, failed migration.

• An incident requires Google to take action on its infrastructure
• An isolated drop is presumed to be a problem specific to the site
• If you want to prove an incident, you must show that multiple unrelated sites are affected
• Google will never validate an incident based on a single case

Is this distinction really that clear in practice?

In theory, yes. In practice? It's sometimes murkier. We've seen targeted algorithmic penalties strike an entire vertical (example: affiliate sites, health sites) without Google communicating about it immediately.

Technically, it's not an incident since the algorithm is working as intended — but for affected sites, it looks a lot like a sectoral incident. The line between "algorithm evolution" and "bug impacting a segment" is sometimes thin.

Does Google use this definition to get itself off the hook?

Let's be honest: yes, partly. By making this distinction, Google protects itself against complaints from webmasters convinced a bug explains their drop — when they've simply been outpaced by competitors or penalized for good reason.

But this stance has a downside: certain real bugs that only affect a small number of sites can be overlooked because they don't reach the scale threshold needed to qualify as an incident. [To verify]: how many isolated cases are actually unrecognized bugs?

In what cases does this rule not apply?

When you observe a strong correlation between multiple unrelated sites (different industries, different hosting) that drop at exactly the same moment. There, you have grounds to document a potential incident.

Another case: silent indexing bugs. A site wrongly deindexed by a crawl bug can remain invisible if no one reports it. Google will never detect it as an incident if you don't report it with solid evidence.

What should you do concretely when facing a ranking drop?

First reflex: check indexation. A site:example.com search in Google, a quick look at Search Console to spot an accidentally added noindex tag, a misconfigured robots.txt, a canonical pointing to the wrong URL.

Next, compare with your direct competitors. If they dropped too, at the same time, on the same queries: you might have found an incident. Otherwise, it's a problem specific to you.

How do you verify if it's an incident or an internal problem?

Use position tracking tools on a sample of unrelated competitors. If everyone drops on a given query, it's probably an incident or an algorithm change. If you're alone, it's you.

Check SEO forums, Twitter, professional groups. Incidents are quickly spotted: multiple people report the same issue simultaneously. Others' silence is a clue: the problem is yours.

What mistakes should you absolutely avoid?

Not documenting your technical changes. A migration, a CMS change, a redesign — everything should be tracked. Without it, you can't correlate a drop to a specific action.

Never assume "Google bugged out" without evidence. This stance discredits you — and worse, it makes you waste precious time while the real problem goes unsolved.

• Check site indexation (site:, Search Console, noindex tags, robots.txt)
• Compare competitor positions on the same queries at the same time
• Analyze server logs to detect a drop in Googlebot activity
• Consult SEO communities to identify a potential global incident
• Document all recent technical changes (migration, redesign, redirects)
• Only qualify a problem as an incident if multiple unrelated sites are affected