The parasite SEO label hides several different claims under one phrase. Google does not use the term. Its March 2024 announcement set out three named policies — expired domain abuse, scaled content abuse and site reputation abuse — and each turns on a different condition. Reading the label as a single tactic usually means skipping the part that matters most: the stated purpose of a page, and the oversight behind it. This guide unpacks that, then hands the question back to you as something measurable. If you want to draft a protocol while reading, the study design tool records a research question, host page, observation window, comparison and confounders, and exports the result as JSON. Related guides cover discovery, indexing and ranking, interpretable comparisons, and reproducible evidence logs.
Three policies, not one label
Google's blog post of 5 March 2024 describes the March 2024 core update as designed to show less content made to attract clicks and more content people find useful, and notes the complex rollout could take up to a month (Google Search Central blog). The same post introduces the three new spam policies. Expired domain abuse is buying an expired domain and repurposing it primarily to manipulate Search rankings with content of little or no value; Google adds that using an old domain for a new, original site designed to serve people first is fine. Scaled content abuse is generating many pages primarily to manipulate rankings rather than help users, whether produced by automation, humans or a combination. Site reputation abuse is third-party pages published with little or no first-party oversight or involvement, where the purpose is to manipulate rankings by leaning on the host site's ranking signals. That policy took effect starting 5 May 2024.
The label "parasite SEO" collapses these. A hypothetical example: a retired medical domain relaunched to host casino pages, and a news site hosting unattributed duplicate affiliate copy, could both be described with the same shorthand. Under Google's framing they are different policies with different tests. Any study that does not separate them is measuring a word, not a mechanism.
What the label conceals: purpose and oversight
Two conditions recur across the definitions. First, purpose — content made primarily to manipulate rankings. Second, oversight — whether the host site is genuinely involved. Google states that not all third-party content is a violation, only that which is hosted without close oversight and intended to manipulate rankings, and gives native advertising and advertorial as content that typically need not be blocked from Search. The spam policies documentation sets out illustrative examples on both sides: a third-party page of payday-loan reviews on an education site, or a low-quality "best casino" advertising page on a medical site, read one way; a news or press-release site, forums and comment sections, editorial columns, properly marked affiliate links and third-party ad units read another. Page-level and site-level guidance is described in the Russian and Japanese versions of the same policy documentation (Google spam policies). Those versions list illustrative factors — how content is presented relative to the host domain, quality differences from the main domain, authorship indications, and whether identical or near-identical content appears on multiple sites — and state that no single factor is necessary or sufficient, and that the list is not exhaustive.
That qualification is the most useful line in the whole set. It means a checklist cannot decide a case, and it also means a researcher should not treat any single visible feature as a proxy for the policy condition. Hypothetical: two pages with mismatched design and no named author may sit on opposite sides, because one host commissioned and edited the section while the other republished a feed. The visible features are the same; the oversight is not.
Third-party scores are not Google signals
A large share of what circulates under this label is a number attached to a host domain — an authority or reputation score. Google's FAQ in the same blog post states that some third-party services provide "reputation" or "authority" scores for sites, and these do not correspond to any of Google's own signals nor come from Google. It also states that its core ranking systems are primarily designed to work at the page level, with some site-wide signals also considered, without enumerating them. Host Notes publishes no host scores and no benchmark tables, because there is no supplied primary source that would support one. If a protocol you are drafting depends on a score, record where the score came from as a confounder rather than as an outcome.
What Bing's guidance adds about measurement
Bing's Webmaster Guidelines describe how Bing discovers, crawls, indexes, evaluates and surfaces content, and state that failure to follow them may result in reduced visibility, reduced eligibility for grounding experiences, or delisting from the Bing index (Bing Webmaster Guidelines). For experiment design, three statements matter. robots.txt controls crawl access, not indexing, and NOINDEX should be used when a URL should not appear in Bing search, Copilot experiences or grounding API results — so those two are separate stages to observe separately. A decline in clicks does not always indicate a loss of visibility, and monitoring impressions, indexing status and grounding eligibility is advised. And Bing states SEO and GEO do not guarantee rankings, grounding or citations, which is a limit on what any single observation can prove.
Bing also states that duplicate URLs dilute signals and reduce its confidence in selecting a URL for grounding or citations, framed as an explanation of consolidation without published numeric thresholds. Microsoft's IndexNow blog post by a named product manager describes notifying search engines about content changes, with incremental notification suggestions for reviews and engagement that the author presents as examples, not ranking rules (IndexNow post). Separately, Bing Webmaster Tools' Top Insights announcement lists example recommendations such as submitting important pages via IndexNow and addressing pages with insufficient content or identical titles; the figures and URLs shown there are demonstration data and must not be reported as findings (Top Insights announcement).
Turn the label into a testable question
Once purpose and oversight are the conditions, the research question has to be about them rather than about the label. A workable protocol names the host page, the third-party section, the observation window, a comparison or control page that is not under question, the intervention or exposure, and the confounders you cannot remove — concurrent core updates among them, given Google's note that rankings fluctuate during a complex rollout. Write the question before you look. Record the evidence URL for every claim you plan to make, and state plainly which stage you are observing, because crawl access, indexing and ranking can move independently. Keep the log public and dated so a reader can re-run the reasoning.
Host Notes will not show findings until they are measured. Nothing here is a host score, a benchmark or a result. What the label hides, once you strip it back, is a set of policy conditions that Google has defined in its own words and declined to reduce to a formula — which is precisely why they belong in a recorded protocol rather than in a slogan.
