Five SEO Myths Checked Against Google's Own Documentation

Each of these five claims gets repeated in sales pitches and SEO forums, usually without a source. Google has published its own position on all five, so we read the pages and quoted them word for word, with the date each page was published or last updated. We read every Google page quoted here on 8 October 2026, and each section ends with a check you can run on your own site in about a minute.

One limit runs through the whole article. Google's documentation covers Google Search, including AI Overviews, the AI-written summary Google shows above some results. ChatGPT and other AI assistants are separate systems with their own rules, so where a claim reaches past Google, we say so.

Does posting on social media improve your Google ranking?

The claim usually arrives as advice to share every new page on every channel, because Google supposedly ranks a page higher once it gets likes, shares and comments.

Google's SEO Starter Guide, last updated 10 December 2025, files social media under promoting your website. The first method on its list is "Social media promotion", below this sentence:

"Effectively promoting your new content will lead to faster discovery by those who are interested in the same subject, and also by search engines."

The ranking question was put to Google's John Mueller directly, in a Google hangout for site owners in August 2015: do social signals have an impact on organic rankings in Google? His answer, at 19:52 in the recording, was "Not directly, no." He went on: "To a large part, social networks also have a nofollow on the links that they kind of provide when they post." Nofollow is a tag in a link's code that tells Google not to pass ranking credit through that link.

Search Engine Roundtable transcribed the answer on 25 August 2015. It is an old answer, and we found no later Google statement that reverses it. Social media does not appear on the Starter Guide's list of things Google says you shouldn't focus on either.

That gives social media a clear job. A post puts your page in front of people. Some of them visit, some search your business name a week later, and a few mention or link to you from their own websites. Links like those are what Google's systems can use: the Starter Guide calls PageRank "one of the fundamental algorithms at Google" and says it "uses links". The post itself leaves your ranking where it was.

Post where your customers already spend their time, and judge it by the visits and enquiries it brings in. A posting package sold as SEO is charging you for a direct ranking effect that Google says doesn't exist.

AI assistants are where the limit above matters. ChatGPT can quote a social post directly in an answer, which has nothing to do with Google's ranking. We looked at that separately in does ChatGPT cite LinkedIn posts.

One-minute check. In Google Analytics, open Reports, then Acquisition, then Traffic acquisition. The Organic Social row shows the visits your posts sent you over the period. That number, and the enquiries behind it, is the fair test of your posting.

Can backlinks hurt a new website?

Backlinks are links from other websites to yours. Two versions of this claim go around: that a new site should avoid getting links because Google punishes young domains that gain them, and that a competitor can sink your site by pointing spam links at it.

Google's spam policies, last updated 28 August 2026, draw the line at intent:

"Link spam is the practice of creating links to or from a site primarily for the purpose of manipulating search rankings."

The first example on Google's list is "Buying or selling links for ranking purposes." Paid links are allowed once they are labelled: "It's not a violation of our policies to have such links as long as they are qualified with a rel="nofollow" or rel="sponsored" attribute value to the <a> tag." In plain words, a paid link is fine when the page's code tags it as paid.

Links you did not make are covered by Google's help page on disavowing links. Disavow is a tool inside Search Console, Google's free dashboard for site owners, that tells Google to ignore links you list. The page says:

"In most cases, Google can assess which links to trust without additional guidance, so most sites will not need to use this tool."

The same page says "Google works very hard to make sure that actions on third-party sites do not negatively affect a website," and names the case where incoming links do cause trouble: "you or an SEO that you've hired may have built bad links to your site via paid links or other link schemes".

So backlinks fall into two piles. Links you bought or built to move rankings are the risk, and Google's spam policies warn that sites breaking them "may rank lower in results or not appear in results at all." Links from sites you have never heard of, which every website collects, are mostly ignored. None of these pages mentions a penalty for being new and earning links.

When you pay for a placement, such as a sponsored article or a directory listing, ask the seller to tag the link as paid, and judge it by the visitors it sends. Turn down link packages sold by the number. If someone offers you a link from a site with an impressive authority score, our guide to what Domain Rating means explains why that number says less than it seems. Leave the disavow tool alone unless Google has flagged your site.

One-minute check. In Search Console, open Security & Manual actions, then Manual actions. A manual action is a penalty a Google reviewer applies by hand, and Google's help page for the report says that if your site has none, "you'll see a green check mark and an appropriate message." Then open Links and look at Top linking sites. If you never paid for or built any of those links, there is nothing to fix.

Can AI-written content rank on Google?

This one says Google detects AI-written pages and buries them, so anything drafted with AI is wasted.

Google answered this in February 2023, in a Search Central blog post by Danny Sullivan and Chris Nelson:

"Appropriate use of AI or automation is not against our guidelines."

The next sentence defines appropriate: "This means that it is not used to generate content primarily to manipulate search rankings, which is against our spam policies."

On 1 October 2026, Google added this line to its guidance on using generative AI content:

"It is critical to manually factcheck and review all AI-generated content for accuracy and trustworthiness before publishing."

We compared the live page with an archived copy from 27 September 2026. The September version told site owners to "focus on accuracy, quality, and relevance, especially when automatically generating the content," and said nothing about fact-checking. The current version, last updated 1 October 2026, adds the fact-check sentence and the reason for it: "generative AI outputs may contain inaccuracies (also known as hallucinations)." It applies the same review to the parts of a page that show up in search results, naming "metadata like <title> elements, meta description elements, structured data, and alternate texts for images". Google's documentation changelog describes the 1 October change as information taken from its Search Quality Raters guidelines.

Google's question, then, is whether the page is accurate and was made to help people. Who or what typed the first draft does not come into it. A page drafted with AI and then checked line by line meets the guidance. A page published unread skips the one step Google now calls critical.

So check every number, name, date and price against a source before the page goes live, and include the page title, the meta description and the image descriptions in that check. Then add what a model cannot know, such as your prices and what happened with your own customers in Singapore.

One-minute check. Open your most-visited page that was drafted with AI. Find its first number or factual claim and try to trace it to a source. If you cannot find one within a minute, that page has not had the review Google calls critical, and a reader checking it would come up empty too.

Does an AI humanizer protect you from a Google penalty?

A humanizer is a tool that rewrites AI text so that AI detectors score it as human. Sellers say running your drafts through one keeps an AI-content penalty away.

Google's spam policies, last updated 28 August 2026, define the practice that does get pages demoted, scaled content abuse:

"Scaled content abuse is when many pages are generated for the primary purpose of manipulating search rankings and not helping users."

The definition continues that this "is typically focused on creating large amounts of unoriginal content that provides little to no value to users, no matter how it's created." Google's examples include "Using generative AI tools or other similar tools to generate many pages without adding value for users", and scraping content to generate pages "including through automated transformations like synonymizing, translating, or other obfuscation techniques".

That takes away the reason to buy a humanizer for Google. The 2023 statement in the section above says appropriate use of AI is within Google's guidelines, so there is no AI penalty to hide from. The test that remains is purpose and value, applied "no matter how it's created", and a thin page stays exactly as thin after its words are swapped. Swapping words for synonyms is on Google's own list of obfuscation techniques.

A humanizer changes wording and checks no facts, and any rewrite that runs after your fact-check produces text nobody has checked. Put the time into the review Google describes instead, checking the facts and adding what only your business knows. If you like how a humanizer reads, run it first and fact-check last.

One-minute check. Put one paragraph beside its humanized version and circle every number, name and date in both. The tool verified none of them. If any of them changed between the two versions, the rewrite added an error you now have to catch.

Does FAQ schema get you into Google's AI Overviews?

FAQ markup, also called FAQ schema, is hidden code (FAQPage structured data) that labels the questions and answers on a page for search engines. The advice going around is to add it to every page so AI Overviews quote you.

Google's page on AI features and your website, last updated 10 December 2025, says:

"There are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary."

Further down the same page:

"You don't need to create new machine readable files, AI text files, or markup to appear in these features. There's also no special schema.org structured data that you need to add."

Google's newer guide to optimising for generative AI features, last updated 10 July 2026, lists "Overfocusing on structured data" among the things you can ignore, and says "Structured data isn't required for generative AI search". The same list says Google Search ignores llms.txt files, a claim we checked separately in does llms.txt do anything.

The FAQ result itself, the dropdown of questions that used to appear under a page in Google, was retired in three steps:

FAQ markup is still valid code and does no harm. Google's 2023 announcement said "there's no need to proactively remove it" and that "Structured data that's not being used does not cause problems for Search, but also has no visible effects in Google Search." It no longer earns a visible feature in Google, and Google says AI Overviews need no special markup. This article ends with an FAQ section that our site marks up automatically, and we keep that markup because it is valid and costs nothing.

What Google does list among its best practices on the AI features page is plainer: "Making sure that important content is available in textual form". Put the questions your customers ask on the page as headings, and answer each one in the first sentence underneath, in text a reader can see. Our guide to what questions your website should answer covers which questions to pick, and our free small business SEO guide walks through building and publishing the page step by step. How to show up in Google AI Overviews covers the rest of the AI Overview work. None of this settles how ChatGPT or other assistants treat markup, which Google's documentation does not cover.

One-minute check. Run your page through our free checker. It lists the structured data it finds and whether any heading on the page is phrased as a question. The headings are the part your reader sees, and for Google's AI Overviews the markup is optional either way.

Which Google pages should you check yourself?

Every quote above comes from one of these pages. Google edits them, so check the date at the bottom of each one before you rely on it.

Frequently asked questions

Do social media posts help SEO?

They help people find your pages, which is how Google's SEO Starter Guide treats them: as promotion. Asked in 2015 whether social signals affect organic rankings, Google's John Mueller said "Not directly, no." Judge your posting by the visits and enquiries it sends.

Does Google penalise AI-generated content?

Not for being AI-generated. Google's position since February 2023 is that "Appropriate use of AI or automation is not against our guidelines." Since 1 October 2026, its guidance also says it is critical to manually fact-check all AI-generated content for accuracy before publishing.

Should I disavow spammy backlinks?

Usually no. Google's disavow help page says most sites will not need the tool, and that it is meant for sites with a considerable number of spammy or artificial links that have caused, or will likely cause, a manual action. Check the Manual actions report in Search Console first.

Should I remove FAQ schema from my website?

There is no need. When Google limited FAQ rich results in 2023, it said there was "no need to proactively remove it", and that unused structured data "does not cause problems for Search". It no longer produces a visible result in Google, and Google says AI Overviews need no special markup.

Can a humanizer make AI content safe for Google?

A humanizer rewords text and checks no facts. Google's spam policies judge content by its purpose and value "no matter how it's created", and they list automated synonymizing as an obfuscation technique. Fact-checking is the step Google's guidance asks for.