Google has finished rolling out its September 2026 spam update. If your search traffic changed around the rollout, start with a dated record of what changed, then examine the affected pages. A drop that coincides with an update is a reason to investigate; timing alone does not establish its cause.
What Google confirmed
The official Search Status Dashboard announcement records a start on September 24, 2026, at 09:15 US/Pacific and an end on October 8 at 01:00 US/Pacific. Google confirmed completion on October 8 and stated that the update applies globally and to all languages.
The announcement names the September spam update. It does not identify affected websites, publish a list of newly targeted practices, or announce an October core update. The checks below use existing guidance; they are not a list of confirmed targets of this particular rollout.
Google’s general spam-update guidance describes these updates as improvements to automated spam detection and advises sites seeing changes to review its spam policies.
Check for a manual action separately
Open the correct property in Search Console and inspect the Manual actions report. A manual action follows a human review of a policy violation and is reported there and in Search Console’s message center. An automated ranking change is a different process. An empty manual-action report does not prove that automated systems have had no effect.
If an action is listed, read its description and affected-page patterns. Follow the issue-specific instructions, fix all affected pages, and use Request Review when the documented issues have been resolved. A reconsideration request is for a manual action; it is not a general request to reverse an algorithmic traffic change.
Build a comparison you can repeat
Google’s traffic-drop troubleshooting guide covers several possible causes, including technical problems, changes in search demand and ranking changes. Look at the broader history before interpreting a short dip. An annual pattern or a site migration can change the explanation.
- Open Search Console’s Search results performance report. Keep the same search type and filters for both periods.
- Select clicks, impressions, CTR and average position. Use date comparison to examine equal-length periods.
- Review Pages and Queries to locate the largest changes. Check Countries and Devices when a loss appears concentrated in one audience.
- Export the comparison and save the selected dates and filters alongside it. Exclude preliminary data from a final assessment.
The official performance-report guide explains these metrics, dimensions and filters. CTR is clicks divided by impressions; a lower click count alone does not show whether impressions, click-through rate, or both changed.
Here is a calendar example for this update. These are our suggested measurement windows, not waiting periods prescribed by Google:
- Baseline: September 10–23, a 14-day period before the rollout.
- Rollout and transition: September 24–October 8, kept separate.
- Later comparison: October 9–22, another 14-day period containing two of each weekday.
As of October 11, the later window is incomplete. You can check broken pages immediately, but a short snapshot is not the finished comparison. Once complete data is available, compare the windows and note holidays, promotions, publishing changes or outages that could affect them. For a low-traffic site, extend both periods rather than drawing a large conclusion from a handful of clicks.
Make a small review sheet with one row per affected URL: earlier clicks, later clicks, impression change, main query, last site edit, suspected issue and evidence. Add an unaffected page for comparison. This keeps the investigation tied to particular pages instead of treating every fluctuation as a site-wide problem.
Rule out a publishing or technical mistake
Check whether affected URLs still load, whether Search Console reports indexing problems, and whether a recent edit introduced an unwanted noindex rule or an incorrect redirect. Use URL Inspection on representative pages rather than assuming their search visibility is unchanged.
If crawl and indexing controls are confusing, our robots.txt versus noindex guide explains their different purposes. Changing either control without understanding the current problem can create another issue.
Review existing spam-policy risks
Use Google’s current spam policies as the reference. Useful audit questions include:
- Scaled content: Were many low-value pages produced mainly to manipulate rankings?
- Doorway pages: Do similar query-targeted pages funnel readers to a destination without useful intermediate content?
- Keywords: Has repetition or a block of place names replaced a readable explanation?
- Links: Were links bought for ranking credit, exchanged excessively, or generated primarily to influence rankings?
- Functionality: Does a page intentionally promise a service while deceptively sending visitors elsewhere?
These are review questions, not proof of a violation. Shared templates, a short article, or an ordinary software bug alone do not establish the intent and behavior described by these policies.
A fictional page audit: find what the visitor actually gets
Imagine a fictional utility website publishes 60 pages titled “Image helper for [city].” Each page has the same introduction, a changed city name and a button pointing to the same editor. There are no local services, instructions or examples. The owner’s review sheet shows that these pages lost impressions while the main editor page remained relatively steady.
That pattern warrants checking the purpose and usefulness of the city pages against doorway and scaled-content policies. It does not prove that this update caused the loss.
Review three examples and ask what each page lets a visitor accomplish before clicking away. If the only difference is a place name, record that observation. Do not manufacture a local address or rewrite the same paragraph with synonyms. Plan the page structure around real reader tasks, and decide which content deserves a distinct page. Handle any later removal or redirect deliberately, with a record of the old URLs and their intended destinations.
For a retained guide, try its instructions yourself. Replace vague promises with the actual steps, supported limits and a useful example. Save the old copy and the corrected draft so the review can show what improved. This exercise produces evidence of better content; it cannot identify Google’s internal reason for a ranking change.
Compare the revision before publishing it
The Text Compare tool can help review the old and revised plain text. Put the same section into Original text and Changed text, keep paragraph boundaries comparable, then click Compare Texts. It displays added, removed and unchanged lines. We checked that behavior in the live browser on October 11, 2026.
Review removals carefully: a rewrite might delete a real limitation along with an exaggerated promise. The comparison shows text changes, not whether a page violates Google’s policies or will rank better. Our guide to comparing two drafts explains how line breaks and comparison options affect the result.
Keep the separate AI documentation change in context
Google also recorded an October 1 documentation update to its guidance on generative AI content, adding information from the Search Quality Raters guidelines. This is a separate documentation change, not an announcement about this spam rollout.
The current AI-content guide calls for checking accuracy and trustworthiness before publication, including titles, descriptions, structured data and image alt text. Useful AI assistance is not automatically spam. Producing many pages without added user value can raise scaled-content concerns.
Make corrections, then measure again
Keep a change log with the URL, issue, evidence, correction and date. Fix a confirmed broken redirect or misleading instruction when you find it. For broader editorial changes, retain the comparison data and avoid changing unrelated pages solely because they share the same publication date.
Google’s spam-update guidance says improvements may occur as automated systems recognize compliance over a period of months. It does not promise immediate recovery or restoration of previous positions. The incident announcement also does not label this rollout a link spam update, so avoid assuming a specific link-related mechanism.
A useful result from this review is a clear record: what changed in Search Console, which pages were checked, what problems were confirmed, and what you corrected. Keep uncertain explanations marked as uncertain until the evidence supports them.