How to Check if a Google Spam Update Hit Your Site

Most traffic dips that people blame on a spam update aren’t one. The check takes 15 minutes in Search Console, and it tells you whether to start auditing or stop panicking.

Google’s September 2026 spam update ran from September 24 to October 8, according to the Google Search Status Dashboard. Here is the exact method I used on my own site, with the numbers.

Step 1: Get the rollout dates

Open the Google Search Status Dashboard and note the start and end date of the update. Don’t compare against the days in between. Rankings move around while a rollout is in progress, and a number taken mid-rollout proves nothing either way.

Step 2: Compare two equal windows on either side

In Search Console, open Performance, set the date range to the 14 days ending the day before the update began, and use Compare to line it up against the 14 days starting the day after it ended. Fourteen days on each side covers every weekday twice, so a weekend dip doesn’t pose as a penalty.

For this update that meant September 10–23 against October 9–22. Here is what my site did:

Those figures come from a Search Console pull, not an estimate. Clicks and impressions rose together and position didn’t move, so this update did nothing to me. If your clicks and impressions both fall sharply, and average position gets worse, you have something to investigate.

Read the combination, because each one points somewhere different:

1. Impressions down, position worse: Google is showing you less. This is the pattern a spam action produces.
2. Impressions flat, clicks down: you kept your rankings and lost the click. That’s a snippet or AI Overview problem. My post on why best-ranking pages get zero clicks covers it.
3. Everything down by the same percentage across every page: suspect tracking, a sitewide technical change, or seasonality before you suspect Google.

Step 3: Split the change by page and by query

A sitewide total hides the story. Switch the compare view to Pages, sort by the click difference, and look at the ten biggest losers. Then do the same for Queries.

An update aimed at spam almost never takes a little off everything. It takes a lot off a specific group of pages. If your ten biggest losers share something, such as all being programmatic pages, all carrying paid guest-post links, or all being thin rewrites of other sites’ articles, you’ve found your suspect. If the losses are scattered across unrelated pages, the update probably isn’t the cause.

Step 4: Rule out the boring explanations

Before you blame Google, check four things:

• Did you change anything? A redesign, a migration, a robots.txt edit or a plugin update in the same window beats any algorithm theory.
• Is it seasonal? Compare to the same weeks last year if you have the data.
• Is your analytics clean? My GA4 for the same month reads as mixed quality, with 13 seconds of average engagement per user, so I don’t trust session counts from it for decisions. Search Console clicks come from Google’s own logs, which is why I use them here. If your GA4 numbers jumped or fell without a matching Search Console change, read how to tell if your GA4 traffic spike is bots.
• Is it a different kind of drop? My longer guide on diagnosing traffic drops walks through the other suspects.

Also open the Manual Actions report in Search Console. A spam update is automated and won’t appear there. A manual action will, and it comes with a message telling you what to fix.

Step 5: If it did hit you, fix the cause first

Don’t rewrite titles or buy links in a panic. Pull the list of losing pages and compare them against Google’s published spam policies: scaled content abuse, expired domain abuse, site reputation abuse (hosting third-party content to borrow your rankings), cloaking, link spam. Find the policy your pages break, then remove the pages or fix them properly. Merely tweaking them won’t get you out. Expect recovery to take weeks to months, because Google has to recrawl and re-evaluate. There’s no form to file for an automated update.

The prompt I run on the export

Export the Pages comparison to CSV from Search Console, then paste this into Claude or ChatGPT along with the file:

This CSV is a Google Search Console page comparison: clicks and impressions
for the 14 days before a Google spam update and the 14 days after it.

1. Total the clicks and impressions for each period and give me the percent
   change.
2. List the 10 pages that lost the most clicks. For each, say what the URL
   pattern suggests the page is (programmatic, tag page, guest post, product
   page, normal article).
3. Tell me if the biggest losers share a pattern, and which of these Google
   spam policies it would most plausibly violate: scaled content abuse,
   expired domain abuse, site reputation abuse, link spam, cloaking.
4. If they share no pattern, say so plainly. Don't invent one.

Check the answer against the pages themselves before you act on it. The model can only see URLs and numbers, so it will guess a pattern from slugs, and you need to open the pages and confirm.

If you run several sites, pulling these windows by hand gets old. I use SiteWide MCP, my Search Console and analytics connector, to ask Claude for the before-and-after comparison across all of them in one request. For keyword-level tracking across many clients, Semrush is the paid option I’d look at first.

Disclosure: the Semrush link is an affiliate link. I earn a commission if you buy through it, at no extra cost to you.