THE SHORT VERSION
— Site reputation abuse is not a content-quality rule — it is an intent-and-independence rule. Google cannot read intent, so it infers it from three observable proxies: who made the content, how well it fits the host’s topic, and whether it would survive on its own domain.
— The November 2024 clarification killed the “we had editorial oversight” defence. No degree of first-party sign-off changes the third-party nature of content that exists to borrow a host’s rankings.
— This guide gives you an original compliance instrument — the First-Party Substance Test — and runs it across the five placement types a link builder actually touches, plus the host-side exposure most agencies never audit.
— Enforcement is still led by manual actions you can see in Search Console and appeal. The June 2026 spam update did not target this policy — a confirmed fact worth knowing before you panic.
— A live EU antitrust investigation means the policy’s application to news domains may change in 2026. If you place content on European publishers, treat the current line as provisional.
Most write-ups of the site reputation abuse policy are written for publishers who got hit. This one is written for you — the person who places content on other people’s domains for a living, and who often advises the very brands that host third-party content. You sit on both sides of this policy at once, and that is exactly why it is your problem more than anyone’s.
Here is the reframe that makes everything else fall into place. Site reputation abuse looks like a spam rule about thin content, but it is really a rule about economic substance. Google is doing what a tax authority does with a sham transaction: it looks past the paperwork — the editorial sign-off, the “in partnership with” byline, the white-label contract — and asks what the arrangement is actually for. If the honest answer is “so this content can rank on a domain stronger than its own,” the paperwork does not save it. The fundamentals of how link equity and authority flow haven’t changed; what changed is that Google now polices the seam where borrowed authority meets independent content.
Google cannot see intent. So it infers intent from three things it can see: provenance, topical fit, and standalone value. Your entire compliance job is making those three proxies tell the truth.
That single sentence is the spine of this guide. Everything below — the definition, the test, the placement audit, the recovery playbook — is an application of it. Companion reading: our sister piece on why parasite SEO tactics are structurally failing in 2026 covers the placement-buyer’s side of this coin; this article covers compliance for anyone who publishes, or advises those who publish.
What the policy actually says in 2026 (and what it doesn’t)
Google introduced site reputation abuse in March 2024 as one of three new spam categories, and began enforcing it with manual actions on 5 May 2024. Within weeks, the visibility of high-authority publisher subsections — the recommendation and “best-of” verticals bolted onto major news brands — collapsed in tracked datasets. The pattern was unmistakable: the host kept ranking; the parasitic section did not.
The definition you should commit to memory is Google’s own, in paraphrase: site reputation abuse is publishing third-party pages on a host site chiefly because of that host’s already-earned ranking signals, so the content ranks better than it could on its own domain. Two words in that sentence do all the work — third-party and chiefly. Miss either and you will misjudge a placement.
The clarification that closed the loophole (19 November 2024)
For the first six months, publishers had a defence: “we had editorial oversight, so it’s first-party.” Google’s Chris Nelson ended that on 19 November 2024. Reviewing white-label deals, licensing agreements, partial-ownership arrangements and other complex structures, Google concluded that no amount of first-party involvement alters the fundamentally third-party, exploitative nature of content published to borrow a host’s rankings. The updated language dropped the oversight carve-out entirely. And Google added the line that should govern how you think about every placement: it does not take a site’s claims about how content was produced at face value.
Why this matters to you
The old link-building playbook — “get a sponsored slot on an authority domain, have their editor rubber-stamp it, enjoy the borrowed rankings” — is now a way to earn the host a manual action and get your placement de-ranked or de-indexed. Sign-off is not a shield. Substance is.
The safe harbour: what is NOT site reputation abuse
A compliance guide that only lists violations is useless, because it makes you afraid of legitimate work. Google is unusually explicit about the green zone, and you should be able to recite it. Having third-party content on a site is not a violation on its own. The following, per Google’s spam documentation, are not site reputation abuse:
- Wire-service and syndicated news republished by news publications — the reader, not the ranking, is the point.
- Advertorial and native-advertising pages whose purpose is to reach that site’s readers directly (properly labelled, and — critically — with monetised links marked so they don’t pass ranking signals).
- Editorial columns and opinion pieces from outside contributors commissioned into the site’s actual remit.
- User-generated content platforms — forums, reviews, comments — that are the site’s genuine purpose.
- Affiliate links inside genuinely first-party editorial that a user would value regardless of the commercial layer.
- Coupons and offers sourced directly from the merchants and businesses that serve consumers — not scraped, licensed-in directories dressed up as editorial.
Read that list next to the violation examples Google publishes — an education site running third-party payday-loan reviews; an established brand branching into a new commercial vertical staffed by freelancers because the content will rank better on the parent domain than on its own — and the dividing line resolves. It is never the format (guest post, sponsored, affiliate) that decides. It is the answer to “would this exist here if the host were a brand-new domain with no authority?”
Don’t confuse it with its two sister policies
Three spam categories arrived close together and get blurred constantly, which leads people to apply the wrong fix. Site reputation abuse is about whose content sits on whose domain. Scaled content abuse is different: it is you generating thousands of low-value pages on your own domain — usually via automation — to manipulate rankings by sheer volume. And link spam is different again: it is manipulation of the links between sites. The reason this matters operationally is that the remedies do not transfer. You cannot fix a site-reputation-abuse manual action by improving content quality alone, because quality was never the charge — independence and motive were. And you cannot fix it with a disavow file, because no external link is involved. Diagnose the actual policy before you spend a week fixing the wrong thing.
The distinction in one line
Scaled content abuse = too many pages on your own site. Link spam = manipulated links between sites. Site reputation abuse = borrowing someone else’s domain authority for content that doesn’t belong to them. Same era, three separate charges, three separate fixes.
The deliverable: the First-Party Substance Test
Because Google infers intent from observable proxies, you can pre-empt it by running the same inference yourself. The First-Party Substance Test is four gates. Gate 2 is the verdict; Gates 1, 3 and 4 are the evidence a reviewer (or an algorithm) reads to decide Gate 2, precisely because they cannot ask you and believe your answer.
| Gate | The question | What a “flag” looks like |
| 1. Provenance | Who actually created it — the host’s own editorial operation, or a separate entity? | Agency, white-label vendor, licensee, or the placing brand itself produced it. Freelancers count as first-party only if commissioned and directed by the host’s own staff. |
| 2. Motive | Strip the host’s authority away. Would this content still be published here, and routed here, if the host ranked no better than a fresh domain? | “No — it’s here for the rankings.” This is the abuse motive itself, and the only gate that is decisive. |
| 3. Fit | Does it sit inside the topical territory the host has actually earned trust in? | A bolt-on commercial vertical (loans on an education site, casino reviews on local news). Google’s independent-section systems wall these off; reviewers read them as parasitic. |
| 4. Standalone value | Would a user seek this page out on its own domain, ignoring where it currently sits? | The page’s only advantage is the host’s rankings. Nothing about it would earn attention standing alone. |
Now translate the gates into a verdict. The logic is deliberately asymmetric, because Google’s enforcement is asymmetric — it forgives genuine first-party work and it punishes ranking arbitrage, with a wide judgment zone in between.
Classification logic
COMPLIANT (green) — Passes Gate 1 outright (genuine first-party provenance and aligned), OR is third-party but clears Gates 2, 3 and 4 convincingly (the safe-harbour cases: wire news, merchant coupons, reader-facing advertorial with marked links).
EXPOSED (red) — Fails Gate 2 (motive is ranking arbitrage) and fails Gate 3 (bolt-on vertical). This is the publisher-recommendation-section pattern that was penalised from May 2024 onward. Do not place here for SEO value; price that value at zero.
GREY (amber) — First-party provenance but affiliate-heavy, or aligned and useful but externally produced. This is the “know it when I see it” zone where enforcement is a human judgment call. Your job here is to strengthen Gates 3 and 4 until the proxies clearly say “legitimate.”
The reason this beats a generic checklist is that it forces you to answer Gate 2 honestly. A checklist lets you tick “has editorial oversight” and feel safe. The Substance Test asks the one question oversight can’t paper over — and it happens to be the exact question Google’s reviewers are trained to ask.
Operationalise it as a one-line record per placement or section: the verdict (Compliant / Grey / Exposed), a single sentence answering Gate 2 in plain language, and the date. That record does double duty. It disciplines the decision at the moment you make it — writing “this only ranks here because of the host’s authority” in your own words tends to end the debate — and it becomes contemporaneous evidence of a functioning compliance process if you ever have to defend the domain in a reconsideration request. A governance log that predates a penalty is worth more to a reviewer than any amount of after-the-fact cleanup.
Running the test on the five placements a link builder touches
Abstract tests are worthless until they classify real work. Here are the five things link builders actually do that intersect this policy, each run through the four gates, with the specific move that shifts a Grey or Exposed placement toward Compliant.
1. Guest posts
A genuinely commissioned guest post that fits the host publication’s editorial remit and provides real value passes cleanly: provenance is a contributor writing for that audience (Gate 1 borderline but Gate 2 clears), fit is strong (Gate 3), value is standalone (Gate 4). The version that fails is the templated, agency-produced post placed on an unrelated authority domain purely for the link — that fails Gate 2 and usually Gate 3. The move: write to the host’s readers, not to Google; keep the topic inside the host’s demonstrated territory; and accept that if the post only makes sense as a link vehicle, its SEO value is now zero regardless of the domain’s rating.
2. Digital PR and syndication
This is the safest of the five, because its motive is genuinely reader-facing. A data study or expert commentary picked up and republished by outlets clears Gate 2 outright — the coverage exists because it is newsworthy, not to lend you the outlet’s rankings. Syndicated news is on Google’s explicit safe-harbour list. The move: keep earning coverage that would run even if the link were nofollowed. That is the definition of a placement no policy update can take away, and it is why digital PR keeps compounding while arbitrage tactics decay.
3. Sponsored content and advertorials
Here the policy and the existing advertising rules converge. A sponsored placement built to reach the publication’s readers is safe-harbour territory — provided it is labelled as advertising and its links are marked sponsored or nofollow so they don’t pass ranking signals. The moment a sponsored slot is engineered to rank organically on the host’s authority, it fails Gate 2 and re-enters site-reputation-abuse territory. The move: decide up front whether a placement is a brand/referral play (label it, mark the links, enjoy the audience) or an SEO play (then it must pass the Substance Test on its own merits, not on sponsorship).
4. White-label and managed content sections
The reviews vertical run by an external partner under the host’s brand is the archetypal Exposed case — the pattern that took down the marquee publisher subsections. Provenance is a separate entity (Gate 1 fails), motive is the host’s authority (Gate 2 fails), and it’s usually a bolt-on vertical (Gate 3 fails). Google specifically named white-label services and licensing deals as arrangements that do not escape the policy. The move: if you manage such a section, the only durable fix is to bring genuine first-party production in-house or to isolate the section technically and stop relying on the host’s rankings for it. There is no wording of a contract that converts Gate 2 from red to green.
5. Affiliate content
Affiliate links inside real first-party editorial are explicitly fine. But 2026 brought a sharper edge: some site sections produced entirely by in-house staff have still drawn manual actions because heavy affiliate monetisation was read as evidence of third-party, ranking-first intent. That is the enforcement reality — Google infers Gate 2 from the commercial density even when provenance is clean. The move: make the editorial the point and the affiliate layer incidental; ensure a reader gets full value with every commercial link stripped out. If the page collapses without the affiliate links, it will read as parasitic no matter who wrote it.
The one-line field rule
Before any placement, ask: “Would this content be published here, and routed here, if the host had zero domain authority?” If no, the placement’s SEO value is zero and its downside is a manual action on the host. Pursue it only for genuine audience or referral reasons.
The host-side exposure most agencies never audit
Here is the blind spot. You can run a flawless outreach programme and still watch a client’s rankings crater — because the client’s own domain hosts third-party content that trips the policy. Big brands routinely carry a “partners,” “marketplace,” “offers” or “experts” section that some other team, or an external vendor, quietly runs. A manual action there can suppress the whole domain’s ranking if the violation recurs — which means your hard-won links stop translating into visibility.
So a modern link-building engagement includes a hosted-content audit of the client’s own site. Treat it as an extension of the backlink and on-site profile audit you’d run against a competitor, turned inward. The workflow:
- Inventory every subsection, subdomain and path where content is produced by anyone other than the client’s core editorial team — partner hubs, licensed directories, white-label verticals, syndicated feeds.
- Run each through the First-Party Substance Test. Flag every Exposed and Grey section.
- For Exposed sections, decide per section: bring production genuinely in-house, remove it, or de-rank it deliberately (noindex plus a clear separation) so it stops leaning on the domain’s authority.
- Log the decision and the date. If a manual action ever lands, this log is the backbone of your reconsideration request.
This audit is also where the technical side of link building earns its keep — section-level noindex, canonical hygiene, and genuine architectural separation are the levers that make Google’s independent-section systems treat a risky vertical as standalone rather than as a parasite on the parent brand.
Manual-action anatomy and the SRA reconsideration playbook
Enforcement of this policy is still led by manual actions — a human reviewer’s decision — which is good news, because manual actions are visible and appealable in a way algorithmic suppression is not. Here is exactly how it plays out and how you get out of it.
How you find out
You are notified in the Manual Actions report and the Search Console message centre (the bell, top right). The notice states the scope: the action may hit specific pages, a subdomain, or a directory. Google’s own warning is explicit that repeated violations can escalate to further actions and can drag down the site’s overall ranking — the partial action is a warning shot you must not ignore.
Read the scope carefully, because it tells you how bad the exposure is. A partial match confined to one directory means Google has walled off a single parasitic section and left the rest of the domain alone — the cleanest situation to recover from. A match that names the whole site, or a partial action that keeps recurring across new sections, signals that Google sees a pattern of behaviour rather than a single lapse, and the reconsideration has to address the governance failure, not just the pages. Do not celebrate a partial action as “ontained” and move on; the recurrence clause means an unaddressed partial is a sitewide action waiting to happen.
The correction everyone gets wrong
noindex alone does not lift the penalty
Adding noindex to the offending pages is a necessary fix, but it does not clear the manual action by itself. Google is explicit: you must still reply in Search Console, explain that the content has been noindexed or removed, and submit a reconsideration request. Sitting on a noindex and waiting is how sites stay penalised for months.
The reconsideration request that actually gets approved
Reviewers reject requests that fix the symptom but ignore the system. A strong site-reputation-abuse reconsideration has four parts, in this order:
- Cause. Name the violating sections honestly and say what made them non-compliant — third-party production, ranking-first motive, bolt-on vertical. Do not minimise; Google already knows.
- Remediation. List exactly what you did to each section: removed, noindexed, brought in-house, or architecturally separated. Be specific and page-level.
- Evidence of changed editorial practice. This is the part most people skip and the part reviewers weight most. Attach the new editorial guidelines, the commissioning process, and a dated timeline showing the practice has genuinely changed — not just that the pages are gone.
- Prevention plan. Explain the control that stops recurrence: who signs off on third-party sections, what the Substance Test gate is, how often you re-audit. Skipping this measurably lowers approval odds.
Two operational traps. If you move the offending content to a new domain rather than removing it, nofollow the links from the old site to the new, and know that dumping it on another established site can simply import the violation there. And approval is not recovery: Google must re-crawl and re-assess, so rankings return gradually over roughly 30 to 60 days, not overnight. Monitor indexed-page count and impressions weekly through that window.
Anonymised case: a UK comparison brand’s partner section
A composite drawn from several 2025–26 engagements, details changed. A UK price-comparison brand had licensed a “deals” subdirectory to an external affiliate operator, who filled it with thin voucher pages that ranked on the parent domain’s authority. A partial manual action hit the subdirectory; comparison-term rankings on the main site wobbled as the recurrence warning bit. The fix was not cosmetic: the subdirectory was removed, a dated editorial-governance document was written, the affiliate arrangement was restructured so any surviving offers were merchant-sourced and first-party, and a quarterly Substance-Test audit was instituted. The reconsideration led with the governance change, not the deletions. The action was revoked; visibility on core terms recovered over about seven weeks. The lesson the brand took away was the one that matters: the penalty was never about the voucher pages’ quality — it was about why they existed.
The algorithmic dimension — and how to talk about it honestly
You will read confident claims that site reputation abuse is now “fully algorithmic.” Be precise here, because a compliance guide that overstates enforcement is as useless as one that understates it. What Google has actually confirmed is narrower and more interesting.
- Confirmed: Manual actions are the primary, documented enforcement mechanism for this specific policy, and they are appealable.
- Confirmed: Separately, Google runs “independent-section” systems that identify subsections starkly different from a site’s main content and treat them as standalone — so a bolt-on vertical stops inheriting the parent’s authority. That is a structural mechanism, and it explains why parasitic sections lose rankings even without a named penalty.
- Confirmed negative: The June 2026 spam update did not target the site reputation abuse policy. Barry Schwartz confirmed this directly with Google; the update was a general SpamBrain improvement, not an SRA-specific one. If your rankings moved in late June 2026, this policy is unlikely to be the cause.
- Inferred, not confirmed: Vendor claims that a specific 2025 update “completed” algorithmic SRA enforcement are third-party pattern analysis, not Google statements. Treat them as informed inference about the direction of travel, not as enforcement scope you can plan around.
There is a candid critique worth internalising: a respected Google-news SEO analyst documented sections penalised as site reputation abuse that were, on a strict reading, entirely first-party — produced by employees — but carried heavy affiliate links, which Google apparently treated as sufficient evidence of third-party, ranking-first intent. The honest conclusion is that enforcement here is partly a “know it when I see it” judgment, not a mechanical rule. That is precisely why the Substance Test optimises the proxies a reviewer reads, rather than pretending there is a bright line to game.
The 2026 regulatory overlay: why the line may move
This is the section your competitors do not have, and the reason a UK or European link builder cannot treat the current policy as settled. The site reputation abuse policy is now the subject of an antitrust fight.
On 13 November 2025 the European Commission opened a formal Digital Markets Act investigation into whether Google, acting under this policy, is unfairly demoting news media and other publishers. In May 2026, reporting indicated Google offered remedies — an adjusted application of the policy specifically to news domains, plus greater transparency about how the policy affects publisher pages. The Commission’s spokesperson signalled the offer was not strong enough. As of writing the remedies are not finalised.
The practical implication
If you place content on European news publishers, the compliance line for those domains is provisional. A future settlement could soften how the policy applies to news properties — or Google could tighten transparency in ways that change what “compliant” looks like. Build for the substance test, which is durable, and keep a watch on the DMA outcome rather than hard-coding today’s application into client strategy.
For teams working across European markets specifically, this is not a footnote — it is a live variable in placement strategy. A news-domain placement that is risky under the current application might sit differently after a DMA settlement, and vice versa. The link builder who tracks the regulatory file will price European news placements more accurately than the one working from a static 2024 mental model.
A word for UK-based teams specifically. Post-Brexit, the DMA does not bind Google’s conduct in the United Kingdom, and the Competition and Markets Authority runs its own digital-markets regime with its own timetable. That means UK news publishers could, in principle, sit under a different application of the policy than their EU counterparts if Google settles with Brussels but not London — or the reverse. In practice Google tends to apply global spam policies uniformly, so do not over-engineer around a split that may never materialise. But if you place on UK news domains, watch both the EU file and the CMA’s posture, and keep the substance test — which is jurisdiction-agnostic — as your primary control rather than betting on any single regulator’s outcome.
A compliance monitoring cadence you can actually run
Compliance decays. Sections get added, vendors get hired, someone launches a “partners” hub without telling SEO. A quarterly rhythm keeps exposure from accumulating silently.
| Cadence | What you do |
| Every placement | Run the four-gate Substance Test before committing. Record the verdict and the Gate 2 answer. If Exposed, price SEO value at zero. |
| Weekly | Check the client’s Search Console Manual Actions report and message centre. A partial action caught in week one is a manageable problem; caught in month three it may have escalated. |
| Quarterly | Re-inventory every third-party section on the client’s own domain and re-run the Substance Test. New partner hubs and licensed feeds are the usual surprises. Update the governance log. |
| Ongoing | Watch the DMA outcome and any Google spam-documentation refreshes. When the policy language moves, re-classify borderline European news placements. |
Most of this runs on tools you already have. Search Console carries the manual-action signal and the recovery loop; your standard link-building and audit toolset surfaces the section-level indexation and monetisation patterns that flag a Grey vertical before Google does. And when you need to argue a placement’s value to a client, the broader data on what actually moves rankings in 2026 makes the case that earned, reader-facing links outperform borrowed-authority arbitrage on every horizon that matters.
Fast answers to the questions clients actually ask
Is guest posting dead because of this policy?
No. Guest posting that fits the host’s remit and serves its readers is untouched — that is ordinary editorial contribution. What died is the agency-templated guest post placed on an unrelated authority domain purely to pass a link. If the post only makes sense as a link vehicle, it fails Gate 2 and its SEO value is now zero. The format is fine; the motive is what’s judged.
We had full editorial oversight of the content. Doesn’t that protect us?
Not since 19 November 2024. Google explicitly stated that no degree of first-party involvement or oversight changes the third-party nature of content published to exploit a host’s rankings, and that it does not take a site’s production claims at face value. Oversight was the loophole; it was closed. What protects you is genuine first-party substance and topical fit, not a sign-off record.
Does adding noindex remove the penalty?
It is a necessary step but not a sufficient one. Noindex alone will not lift a manual action. You must also reply in Search Console, explain that the content is noindexed or removed, and submit a reconsideration request documenting the fix and the governance change behind it. Waiting on a noindex without filing is how sites stay penalised for months.
How long does recovery take once the reconsideration is approved?
Approval is not the same as recovery. Google still has to re-crawl and re-assess the affected sections, so rankings typically return gradually over roughly 30 to 60 days rather than snapping back. Monitor indexed-page count and impressions weekly through that window, and expect the parent domain’s core terms to recover before every affected sub-page does.
We host third-party content but never do outreach. Are we exposed?
Potentially yes — this policy is about hosting, not outreach. If your domain carries a partner hub, marketplace, licensed directory or affiliate vertical produced by anyone outside your core editorial team, run each section through the Substance Test. The brands penalised in 2024–25 were overwhelmingly hosts, not link buyers. Hosting third-party content is the exposure; outreach is a separate discipline.
Will the policy change because of the EU investigation?
It might, for news domains specifically. The European Commission opened a formal DMA investigation in November 2025, and Google offered to adjust how the policy applies to news publishers and to increase transparency — an offer the Commission’s spokesperson called insufficient. Nothing is finalised. Build for the durable substance test and treat the application to European news domains as provisional until the file closes.
The bottom line
Site reputation abuse compliance collapses to one discipline: stop trying to make third-party content look first-party, and start making sure it is — or accept that its SEO value is zero and pursue it for audience alone. Google infers intent from provenance, fit and standalone value because it cannot ask you and believe the answer. Optimise those three proxies to tell the truth, run the Substance Test on every placement and every hosted section, keep a governance log, and watch the DMA file. Do that and the policy stops being a threat and becomes what it was designed to be — a moat that rewards the people doing genuine work and drains the ones renting authority.
For the strategic frame around all of this, our guide to building durable link-building strategy in 2026 places compliance inside the wider shift from borrowed authority to earned relevance. This policy is not an obstacle to that shift. It is the shift, enforced.
