Site reputation abuse sounds like someone else's problem. Coupon subfolders on national newspapers, casino reviews parked on a university domain, the sort of thing that gets written up on SEO news sites and then forgotten. But the policy is written about page types, not company sizes, and plenty of ordinary UK businesses are hosting content that fits the description without ever having planned to.

This article takes the auditor's seat. How to crawl your own domain, list every URL you did not commission, decide which ones are actually a problem, and fix the ones that are. Also, when not to panic: most ranking drops blamed on this policy turn out to be something much more boring.

What Google actually means by site reputation abuse

The definition, in Google's own spam policies documentation, is third-party content published on a host site mainly to take advantage of that host's established ranking signals. Not "content about the wrong topic". Not "content written by an outsider". Content that is there because of where it sits, rather than what it is.

The policy went live in March 2024 and enforcement started that May. The clarification that matters most came on 5 November 2024, when Google updated the wording to say that a violation stands regardless of first-party involvement or oversight. Read that again if you commission partner content. It removed the most popular defence in the industry overnight: "but our editor reviews everything before it goes up"is no longer an exemption. The test is why the page exists, not how carefully it was proofread.

The nickname parasite SEO stuck because it describes the mechanics honestly. There is a supply side, selling space on a domain that ranks, and a demand side, buying that space. They carry different risks. Sell it and you put your own domain in front of a manual action. Buy it and your commercial visibility lives on somebody else's property, which can be deindexed or ranked separately without warning. Rented traffic, no asset at the end of it.

Plenty is explicitly not covered, and it is worth being clear about that before you delete half your site:

• Normal advertising, including display and sponsored placements that are not there to rank.

• Wire service and syndicated news content.

• Genuine user-generated content: forums, community posts, customer reviews.

• Coupon or deals content produced by the site's own team, with the site's own oversight and judgement.

The difference between the last bullet and a violation is not the topic. It is who actually made the thing.

The page types that trip the policy most often

Four patterns account for most of it.

Rented subfolders and subdomains. An affiliate network or white-label operator takes /discount-codes/, /casino/, /loans/ or /supplements/ on a domain that has nothing to do with any of those things. The host gets a monthly fee and touches nothing. A UK home services company that lets an affiliate run a voucher subfolder can watch that section outrank dedicated deals sites for months, then see it disappear entirely, taking a visible chunk of reported organic sessions with it while the boiler and bathroom pages carry on exactly as before. That split is the tell: one section collapses, the rest of the site does not move.

Sponsored article hubs and paid "best of"listicles. Written by the advertiser or their outreach agency, published under your masthead, positioned to rank for a commercial term. If the advertiser wrote the ranking, the advertiser bought the ranking.

Bulk guest posting on unrelated themes. A B2B SaaS site taking roughly twenty outreach-agency posts a month on crypto, CBD and casino topics is the textbook case. Ask the simple question: could anyone in the business defend a single one of those articles to a customer? If nobody can vouch for the accuracy, nobody owns it.

Partner and affiliate directories where the host contributes the domain and the page template, and nothing else.

Where smaller sites get caught out without meaning to

Nobody sets out to do this. It arrives sideways.

A reseller or manufacturing partner asks to put their product pages on your domain "because you rank better than we do". That sentence is the policy, said out loud, in a friendly email. A site migration brings across a legacy /resources/ section from an acquired brand that nobody has opened since 2021. Job boards, event listings and press release feeds get pulled in automatically, indexed by default, and left there generating thousands of URLs that no human at your company has read.

And then there is the misdiagnosis trap, which we see more often than the real thing. Rankings fall, somebody reads about site reputation abuse, and a perfectly healthy blog gets torn apart looking for a spam problem that was never there. The actual causes, most of the time:

• Three of your own pages competing for the same query, so Google keeps swapping which one it shows and none of them settle.

• Duplicate titles and near-identical meta descriptions spilling out of an old category archive or tag system.

• A batch of vague, filler-heavy drafts with no first-hand detail, published quickly, ranking briefly, then sliding.

Diagnose before you demolish. A crawl and a query-level export from Search Console will tell you which of these you are looking at in an afternoon.

How to audit your own domain for third-party pages

The audit is a spreadsheet exercise, and it is not complicated. It is just tedious, which is why it rarely gets done.

1. Crawl everything. Run a full website SEO audit crawl with any standard crawler and export every indexable URL. Most UK SME domains land somewhere between 100 and 500 pages, which is a ten minute crawl. If your export comes back with 4,000 URLs and you were expecting 300, you have found something already: paginated feeds, listing archives or parameter URLs nobody knew were indexed.

2. Add an author-ownership column. Tag every URL as in-house, partner-supplied, or unknown. Do it by section first, then spot-check individual pages. The unknown column is the real output of this audit. Everything in it needs a name against it before you go any further.

3. Ask three questions of every third-party section. Did we commission the topic, or did it arrive pre-chosen? Would we still publish this if our domain had no ranking authority at all? Does one named person here own its accuracy? Two nos and a shrug means you have a problem. One no might just mean the content is weak, which is a different fix.

4. Cross-check indexation against query data. Export your Search Console queries and look for clicks on terms that have nothing to do with what you sell. A plumbing company picking up impressions for "best crypto wallet uk"does not need a second opinion.

5. Check the Manual Actions report. In Search Console, a manual action for this will be labelled and will name the affected part of the site. Read the Manual Actions report documentation so you know what you are looking at. Note whether the message covers one section or the whole property, because that changes the scale of the clean-up considerably. No manual action and a site-wide traffic slide? Then you are almost certainly dealing with an algorithmic issue or one of the boring causes above.

Fixing what you find, and what recovery realistically looks like

The remedy has to remove the ranking benefit. That is the whole logic of the fix, and it is where most advice goes wrong.

A canonical tag pointing at the partner's own domain does not resolve it: the page stays live on yours and can still be eligible to rank in some situations, because canonicals are a hint, not an instruction. A disavow file does nothing whatsoever, because nothing here is a link penalty. Forget both. What works is noindex on the offending pages, or removing them outright and returning a 410. If the commercial relationship matters to you, the partner can host the content on their own domain and you can link to it like any other external resource.

Some content is worth keeping. If the underlying topic genuinely belongs to your business and the problem is authorship, take real ownership: rewrite it in your voice, add detail only your team could supply, and file it under a properly scoped section rather than a rented one. That is not a cosmetic pass. If the result reads like the same article with your logo on it, you have not fixed anything.

After a manual action, remove or noindex first, then file a reconsideration request that shows your working: what you found, how many URLs, what you did with each group, and what has changed in your publishing process so it does not come back. Half-measures get rejected. "We cleaned up some of it"is the most common reason a request bounces, and every bounce costs you another review cycle. Google does not publish a fixed turnaround, and review times vary, so plan for weeks rather than days and do not republish anything in the meantime.

One development worth watching: through 2025 there were credible reports that in the EEA, Google may respond by ranking an affected section independently of the host domain rather than issuing a penalty. The section stays live, it simply stops borrowing the parent domain's strength. Check the current wording in Google's own policy and regional documentation before you rely on that, because it is the kind of detail that changes.

A publishing model that never needs this policing

If your content is only ever about things you actually do, this policy never touches you. That is the short version.

The harder version is discipline about what you publish and why. Reactive posts on unrelated hot topics sit in an uncomfortable middle ground: chasing a trending subject with no connection to your business is not a violation in itself, but it fails the same underlying test Google is applying. Would this page deserve to rank if it were not on your domain? If the honest answer is no, the page is weak whether or not anyone calls it spam.

Four habits keep you clear:

• Write from standing. Topics where someone in the business has first-hand experience, so the content earns its place instead of borrowing the domain's.

• Check what already exists before every brief. This is the single cheapest keyword cannibalisation fix there is: read your own archive, and update the existing page instead of publishing a third one on the same query.

• Grade drafts twice. One score for whether the piece can rank, one for whether a human expert would put their name to it, with integrity winning the tie. We build that into how we write every article, alongside a sceptical sub-editor pass before anything goes near your CMS.

• Ramp the cadence. Forty pages in a week on a domain that published four last year looks exactly like a rented subfolder. Build up, and keep schema, internal links and images consistent with the rest of the site.

We do not sell placements on other people's domains, and we do not host other people's content on ours. If you want to see what our own publishing setup includes, the full feature list is public, and unfamiliar terms from this article are explained in the content marketing glossary. If you would rather start with a read of your current site, the free content marketing snapshot is a sensible first step before any clean-up.

Frequently Asked Questions

What is site reputation abuse in Google's spam policies?

It is third-party content published on a host website mainly to take advantage of that host's existing ranking signals. Google added it to the spam policies in March 2024 and began enforcement in May that year. The judgement is about why the page is on your domain, not what subject it covers.

Does site reputation abuse apply to guest posts and sponsored articles on a small business site?

It can. Domain size is not part of the definition. A guest post on a topic you genuinely work in, commissioned by you and owned by someone in your business, is normal publishing. Twenty agency-supplied posts a month on subjects nobody at the company could defend to a customer is the pattern the policy targets.

Is a manual action for site reputation abuse site-wide or limited to one section?

The Manual Actions report in Search Console names the affected part of the site, so it can be scoped to a single directory or subdomain rather than the whole property. Check the message text before assuming the scale. A section-only action typically shows as one area of traffic collapsing while the rest of the site holds steady.

Will adding a noindex tag or a canonical fix third-party pages?

noindex works, because it removes the ranking benefit, which is the thing the policy is about. A canonical pointing at the partner's own site does not: canonicals are a hint, the page stays live on your domain, and it can still be eligible to rank. A disavow file is irrelevant here, since this is not a link penalty.

How long does it take to recover after removing the pages?

Google does not publish a fixed timeline, and reconsideration review times vary, so plan in weeks rather than days. Remove or noindex everything first, then file one request that shows the full scope of the clean-up. Partial fixes get rejected and put you back in the queue.