Content Spinning: What It Is, How It Works and Why Google Penalizes It

Spinning is one of the oldest shortcuts in SEO. Wikipedia's entry on article spinning classifies it as spamdexing, a black hat technique, and notes that it is also known as "Rogeting" (after Roget's Thesaurus). Large language models have given content spinning a second life: a rewrite that took a spinner tool several passes now takes one prompt. This guide explains how spinning works, what Google says about it, how it differs from legitimate rewriting, and what to do instead.

What is content spinning?

Content spinning takes one source text and produces variants that keep the meaning and structure but change the surface: words, word order, sentence length. The source may be your own article or, more often, someone else's. The goal is apparent originality: a version different enough to pass a plagiarism checker or a duplicate content filter, without the work of writing something new.

There are two ways to do it:

• Automatic spinning: software replaces words using a synonym dictionary, sometimes reorders clauses, and outputs hundreds of variants in a few clicks.
• Manual spinning: a writer rewords the text by hand, sentence by sentence, without adding information. The result reads better, but it is still the same article in disguise.

Spinning is related to paraphrasing, with one difference that matters: paraphrasing restates an idea in your own words, inside your own work, usually with attribution; spinning restates a whole text to pass it off as new.

How does article spinning work?

Automated spinners rely on spintax (spin syntax), a markup where each interchangeable element is written between braces with its options separated by vertical bars. A sentence such as {Quick|Fast|Rapid} {tips|advice} to {grow|increase} your {traffic|audience} can produce 3 x 2 x 2 x 2 = 24 versions. Options can be nested at the word, sentence and paragraph level, so a single annotated article can generate thousands of combinations.

Researchers at the University of California San Diego reverse-engineered one of the most popular tools for their paper DSpin: Detecting Automatically Spun Content on the Web (NDSS 2014). They found that the heart of these tools is a hand-built synonym dictionary, and that spinners leave some words untouched (the paper calls them "immutables"). Comparing those immutables between two pages reveals that they come from the same source, which is how spun content can be detected at scale. In a set of 427,881 pages from heavily abused wiki sites, they found that 52% of the English content pages were automatically spun articles.

Automation also explains why spun text reads badly. A synonym engine does not understand context: Wikipedia cites the example of "Great Britain" becoming "Good Britain". Idioms break, technical terms are replaced by near-synonyms that change the meaning, and the text keeps the exact paragraph structure of the source.

Why was content spinning used in SEO?

Spinning answered two needs of 2000s and early 2010s SEO:

1. Volume of pages: one article could be turned into dozens of pages targeting slight keyword variations, often combined with keyword stuffing.
2. Volume of backlinks: spun versions were posted on article directories, wikis and blog networks, each with a link to the site being promoted. Search engines filter exact copies, so each copy had to look different.

Google's Panda update (2011) targeted content farms and low-quality pages, then Penguin (2012) targeted manipulative links. Both made the classic spinning model unprofitable. Generative AI revived it in a new form: instead of a synonym engine, an AI model rewrites a competitor's article in seconds, with better grammar but the same lack of originality.

Does Google penalize spun content?

Yes. Google's spam policies for Google web search describe spinning in two places:

• Scraping: the policy lists copying content from other sites, modifying it only slightly "(for example, by substituting synonyms or using automated techniques)" and republishing it.
• Scaled content abuse: "when many pages are generated for the primary purpose of manipulating search rankings and not helping users". Among the examples: scraping content to generate many pages "through automated transformations like synonymizing, translating, or other obfuscation techniques".

The scaled content abuse policy was announced with the March 2024 core update. Google's Search Central post of March 5, 2024 explains that it applies "no matter whether content is produced through automation, human efforts, or some combination", so manual spinning is covered too. On the Google blog, Google later reported that after the rollout completed on April 19, 2024, search results showed 45% less low-quality, unoriginal content.

A site with spun content faces three risks: the pages are filtered and never rank, the site receives a manual action (visible in the Manual actions report of Google Search Console), or the low quality of part of the site weighs on how Google sees the rest. Google's helpful content guidance asks whether a page adds original information beyond the obvious, and spun content fails that test by design. Using AI is not the problem in itself, as explained in our article on whether Google penalizes AI content: the problem is unoriginal content produced at scale.

Content spinning vs paraphrasing, AI rewriting and repurposing

Several practices change the words of an existing text. What separates them is whether the result adds value for the reader.

The test is simple: if you removed the source, would your page still say something the source does not? If not, it is spinning, whatever the tool.

How do you spot spun content?

Whether you are reviewing content bought from a freelancer, a guest post submitted to your blog or your own old pages, these signs point to spinning:

• Odd synonyms: "search engine improvement" for "search engine optimization", broken idioms, technical terms replaced by vague words.
• Mirror structure: the same headings, the same order of arguments and the same examples as a page that ranks, only reworded.
• No information gain: no data, example, opinion or experience that is not already in the source. Our glossary entry on information gain explains why this counts.
• Partial matches in a plagiarism checker: sentences that are not identical but share long sequences of words with another page.
• Batches of near-identical pages on your own site (city pages, product pages) that differ by a few words, which also creates duplicate content.

What should you do if your site has spun content?

1. List the affected pages

Export your pages with their traffic and impressions from Google Search Console. Flag pages that were bought in bulk, generated by a spinner or rewritten from competitors, and check the Manual actions report.

2. Decide page by page

Rewrite the pages that target a real need, merge near-duplicates into one strong page with a 301 redirect, and delete or noindex the pages that serve no one. Removing weak pages is a normal part of quality work, sometimes called content pruning.

3. Rewrite with original information

Start from the reader's question, not from the source text: add your own data, screenshots, prices, steps you have tested and a clear answer at the top. Our method to humanize AI text applies here too.

4. Request a review if there is a manual action

Once the problem is fixed across the site, submit a reconsideration request from the Manual actions report, explaining what was removed or rewritten.

What are the safe alternatives to content spinning?

• Repurpose your own content: turn one strong article into a checklist, a video script, a LinkedIn post or a newsletter. Our guide to content repurposing shows how to turn one article into ten.
• Refresh what you already have: update figures, add sections the top results cover and remove outdated advice. See our method for a content refresh.
• Cover the topic from another angle: if a competitor's article is the reference, write the one it does not cover (a comparison, a case, a template), instead of a reworded copy.
• Consolidate instead of multiplying: one complete page beats five thin pages that compete with each other.
• Use AI to draft, not to disguise: AI is useful for outlines and first drafts built on your sources, keywords and internal links, followed by a review. This is the approach of the Sorank AI blog article generator, which writes each article from keyword research and cited sources, with automatic or draft publishing for review.

Does spun content get cited by AI assistants?

Rarely. ChatGPT, Gemini and Perplexity search the web, then pick the passages that best answer the question. When ten pages say the same thing in different words, the assistant tends to cite the clearest or most authoritative one, which is usually the original. A spun page adds nothing the assistant cannot already get from its source, so it has little reason to be chosen. Our study of the sources ChatGPT cites shows which types of pages AI assistants rely on.

What earns citations is the opposite of spinning: original data, precise definitions, comparison tables and direct answers that only your page provides. To see whether those efforts pay off, you can track which questions mention your brand, and who is cited instead of you, with AI mention tracking.

Is content spinning worth the risk?

No. Spinning saves writing time but produces pages Google is designed to filter and AI assistants have no reason to cite, and it can drag down the rest of a site. The only durable way to scale content is to add something the source does not have. Spinning sits firmly on the black hat SEO side; publishing regularly does not have to. To publish original, sourced articles every month without spinning, see how Sorank works, with a 3-day free trial.