Google now names four qualities of main content. How to audit a page against effort, originality, skill and accuracy

KEY TAKEAWAYS

1. Main content now explicitly includes titles, headings, tools, tabbed sections and user reviews. A text-only review misses part of what Google describes.
2. Effort covers the systems behind a page. A templated page with a real tool or dataset is not the same as a templated page with reworded text.
3. The rater threshold for the Lowest rating is all or almost all of the main content. Triage by template, not by single sentences.
4. Act because spam updates enforce the scaled content abuse policy, not because of a forecast core update whose timing nobody outside Google knows.

In the first week of October 2026, Google rewrote two pages that most content teams treat as background reading. The people-first content page gained a definition of main content, four attributes for judging it and a warning about invented authors. The generative AI page gained an instruction to fact-check AI output by hand, including metadata. Most coverage has reported the wording. This piece turns it into a page audit you can run.

What exactly changed between 1 and 5 October?

Three edits matter.

1. Main content is defined. Google now describes it as any part of a page that directly helps the page achieve its purpose. The list is broad: primary text and media, interactive features such as calculators and site search, user contributions such as reviews and forum replies when they serve the page's purpose, tabbed or expanded sections, and page titles and headings.
2. Four attributes are named. Effort, originality, talent or skill, and accuracy. Effort is the human work in the content or in the systems that produce it. Content generated at scale without manual oversight or curation counts as little to no effort, and crediting a source does not stand in for original work. Talent or skill does not always mean credentials; Google notes that not every topic needs specialist expertise. Accuracy is a sliding scale. Topics that affect health, money or safety must match established expert consensus.
3. Invented authors are called deception. AI-generated headshots, made-up names and false credentials are listed as examples, and Google calls any such deception a signal of a low-quality page.

Separately, on 1 October the generative AI page added that models predict likely word sequences rather than retrieve facts, so outputs can be wrong. It now says manual fact-checking of all AI-generated content before publishing is critical, and that this review also covers title elements, meta descriptions, structured data and image alt text.

One detail matters for monitoring. Google's documentation changelog logged the generative AI edit on 1 October. Search Engine Journal noted that the people-first edit was not listed there, and it was still absent when we checked on 13 October. That page now shows a last-updated date of 5 October. If you only watch the changelog, you miss changes like this one.

Is this a new ranking rule?

The evidence says no. The attributes describe what Google's quality raters assess. Google's own generative AI page repeats that raters' ratings do not directly influence ranking. The ideas are also not new. The rater guidelines updated in January 2025 already had a section on scaled content abuse (4.6.5) and one on main content made with little effort, originality or added value (4.6.6). Search Engine Journal points out that the guidelines already treated fake author profiles as deception.

Some well-known voices read the edits as a warning shot. Lily Ray, writing on 4 October, predicted a major update, while describing her view as an informed opinion rather than knowledge. Roger Montti at Search Engine Journal expects a core update before the holiday season. Google had announced none when he wrote on 5 October, and we found no announcement by 13 October.

Who is right? Both, on different questions. The criteria are a restatement, so there is nothing new to optimise for. But the restatement now sits in public guidance for site owners, and the enforcement path is real. By Ray's count, Google has run four spam updates in 2026: March, June, August and September. The spam policy that matters here, scaled content abuse, applies however the content was made. Act on the enforced policy, not on the forecast.

Does Google penalise AI-written pages?

Not for the method. Ahrefs scored the top 20 results for 100,000 random keywords, about 600,000 pages, with its own AI detector. It found 4.6% pure AI pages, 13.5% pure human and 81.9% mixed. The correlation between AI share and ranking position was 0.011, close to zero. Pure AI pages did appear in the top 20 but rarely held position one.

That study has a blind spot. It only sees pages that already rank. Glenn Gabe's case studies from the August 2026 spam update, which ran from 18 to 21 August, show the other side. One site with programmatic, fully AI-generated pages on a high-stakes topic lost rankings for more than 200,000 queries. Another had over 1.5 million indexed URLs, about 85% of them in a programmatic section, and saw a site-wide impact.

Put together, the data supports one reading. AI-assisted pages rank. Large sets of pages with little effort and nothing original are what get removed, and AI makes it cheap to build them. That matches the new effort definition almost word for word.

Why does the main content definition widen your audit?

Because titles and headings are now named as main content, and the AI fact-check explicitly covers metadata. A review that reads only the body copy misses the parts a searcher sees first. Google's title link guidance has long warned against boilerplate titles that change by only one detail. A templated heading pattern now falls under both documents.

The definition also works in your favour. Effort covers the systems behind a page, not only the prose. A location page built on a real price table, a working calculator or verified reviews has a source of effort and originality that a reworded paragraph does not. Before you rewrite text, check whether the page has a tool, dataset or first-hand detail that could carry it.

Which pages should you fix first?

Use the rater threshold. Search Engine Land's summary of the January 2025 guidelines quotes the Lowest rating as applying "if all or almost all of the MC" lacks effort or added value. A Low rating applies when some content is reused but shows at least minimal curation. Filler that crowds out useful material can also earn a Low.

So triage by template and by share of the page, not by sentence. One generic paragraph on an otherwise useful page is a Low-level problem. A template where almost everything is paraphrase is a Lowest-level problem and a scaled content risk.

A worked example (hypothetical)

A home services company has 400 city pages. All were drafted with an AI tool from one prompt. Each has a 250-word introduction, a list of services and a meta description, also AI-written. Sixty pages also show a price range table pulled from the company's own job records for that area. The other 340 swap only the city name.

Applying the method:

• The 340 swap-only pages fail on effort and originality across almost all of their main content. That is the Lowest pattern and the scaled content pattern. Merge them into a smaller set of regional pages built on real data, and redirect the old URLs.
• The 60 pages with local price tables have original main content. Keep them. Check that each table is current, and that the AI-written title and description state the same figures as the table.
• Every author box uses one invented name. Replace it with the real person who signs off the content, or remove it.

The result is about 60 strong pages plus a few regional hubs, not 400 weak ones. This is an illustration of the method, not a measured outcome.

The seven-step audit

1. List templates, not URLs. Group pages by the template or prompt that made them. Sample five pages from each group.
2. Mark the main content. For each sample, note what helps the page do its job: text, tool, table, reviews, tabs, title and headings.
3. Score the four attributes. For each element, ask: did a person or a real system produce it (effort)? Is it on other sites already (originality)? Does it work and read well (talent or skill)? Can every claim be traced to a source (accuracy)?
4. Apply the threshold. If all or almost all of the main content fails effort or originality, flag the whole template. If only parts fail, flag the parts.
5. Check metadata against the body. Every figure in a title, description, structured data field or alt text should appear in the body with the same value.
6. Check authors. Every byline should be a real person with a page that states verifiable background. Remove anything invented.
7. Decide and log. Merge, exclude, rewrite or keep, using the table above. Record the decision and the reviewer for each template.

If you already saw a drop during the September spam update, run this audit after the diagnosis in our September 2026 spam update guide. If the fix is a rewrite, our note on what counts as a real content update explains why a changed date alone does little.

What to watch next

Watch the two Google pages themselves, not only the changelog, since one edit did not appear there. Watch for a core update announcement, and read any drop against the table above before you change anything. Bylines still do not help ranking on their own; Google's Search Liaison said so in January 2024. Real authors matter because invented ones are deception.

Main content quality is one input into how search engines and AI answers judge a brand. We cover the wider picture in what digital authority means. If you want to track it as we build our tools, join the waitlist.

Questions

What did Google change in its helpful content documentation in October 2026?

Between 1 and 5 October 2026 Google added a definition of main content, four attributes that quality raters assess (effort, originality, talent or skill, accuracy) and a warning that fabricated author profiles are deception. A separate 1 October edit to the generative AI page says AI output should be fact-checked by hand, metadata included.

Is effort, originality, talent or skill, and accuracy a new ranking factor?

Google has not said so. The attributes describe what quality raters assess, and Google states that rater scores do not directly influence ranking. The same ideas have been in the rater guidelines since January 2025. What is new is that they now sit in documentation written for site owners.

Does Google penalise AI-generated content?

Not for the method. An Ahrefs study of about 600,000 top-20 pages found a 0.011 correlation between AI share and position. Google's spam policy targets many pages made without adding value, however they are produced. Sites hit in 2026 spam updates combined AI text with programmatic scale.

Which pages should I audit first?

Start with templates where all or almost all of the main content is generic or paraphrased, because that is where raters are told to give the Lowest rating. Then check pages on health, money or safety topics, where accuracy must match expert consensus. Review titles, headings, descriptions, structured data and alt text as part of each page.

Sources

1. Google Search Central, Creating helpful, reliable, people-first content developers.google.com
2. Google Search Central, Google Search's guidance on using generative AI content on your website developers.google.com
3. Google Search Central, Latest Google Search documentation updates developers.google.com
4. Google Search Central, Spam policies for Google web search developers.google.com
5. Google Search Central, Influencing your title links in search results developers.google.com
6. Ahrefs, AI-Generated Content Does Not Hurt Your Google Rankings (600,000 Pages Analyzed) ahrefs.com
7. GSQI, Google's August 2026 Spam Update case studies gsqi.com
8. Search Engine Journal, Google Tells Sites To Fact-Check AI Content Before Publishing searchenginejournal.com
9. Search Engine Journal, Google Adds Fake Author Warning To Helpful Content Guidance searchenginejournal.com
10. Search Engine Journal, What To Expect From Google's Next Search Ranking Update searchenginejournal.com
11. Search Engine Roundtable, Google Helpful Content Doc With New Main Content and EOT/SA Sections seroundtable.com
12. Search Engine Land, Google quality raters now assess whether content is AI-generated searchengineland.com
13. PPC Land, Google says fake author profiles signal a low-quality page ppc.land
14. Lily Ray, Prediction: the next massive Google update is around the corner lilyraynyc.substack.com

NEXT STEP

See where your brand stands.

Your Digital Authority Score, who AI recommends for your target questions, and the placements that close each gap. Your first report is free.

Get your free report →