Ch 18 — Spam enforcement & recovery¶
Part VI — Change events · The Technical SEO Reference
Playbook coupling:
seo-checklist.md's two ⛔ Never-Do lists (lines 230–249) enumerate the policies and mental model #2 defines the three rulebooks. This chapter supplies the enforcement machinery the playbook promises but can't hold: the complete manual-actions taxonomy, partial-vs-sitewide scoping, the reconsideration process, site reputation abuse mechanics (the 2024–2026 window's marquee policy), and the algorithmic-vs-manual distinction that decides every recovery strategy. Directly load-bearing for the gambling-affiliate portfolio: casino content is a named example in two policies.
The playbook's mental model holds: spam policies are named violations enforced two ways — spam updates (algorithmic; no notice, no review path; recovery = fix + wait for reprocessing) and manual actions (a human flagged you; visible in GSC; recovery = fix + reconsideration). If it's not in the Manual Actions report, you don't have one — and "we were hit by a manual action" without that report entry is always misdiagnosis. Quality problems (E-E-A-T, helpfulness) are a third thing entirely: no violation, no action, no review — just losing to better content (→ playbook Phase 3/7).
18.1 The named spam policies — the 2026 inventory¶
🟢 From the spam policies page (2026-05-15, fetched live), sixteen named policies: Cloaking · Doorway abuse · Expired domain abuse · Hacked content · Hidden text and link abuse · Keyword stuffing · Link spam · Machine-generated traffic · Malicious practices · Misleading functionality · Scaled content abuse · Scraping · Site reputation abuse · Sneaky redirects · Thin affiliation · User-generated spam.
🟢 The consequence, verbatim: "Sites that violate our policies may rank lower in results or not appear in results at all." Deindexing is the ceiling, not the default (playbook line 232 stands). Policy-specific mechanics live where they're used: sneaky redirects → Ch 16 §16.4; hacked content → Ch 17 §17.6; cloaking boundaries → Ch 6 §6.10 (dynamic rendering), Ch 16 §16.8 (A/B testing); doorway/scaled-content in the playbook's ⛔ list. ⚠️ Note thin affiliation as a named policy — affiliate pages that add no value beyond the merchant's feed are a spam matter, not merely a quality one (the vertical's standing exposure).
18.2 Manual actions — the complete taxonomy¶
🟢 Definition (Manual Actions report help, fetched live): "Google issues a manual action against a site when a human reviewer at Google has determined that pages on the site are not compliant with Google's spam policies."
The current full list of action types:
| Manual action | What it flags |
|---|---|
| Back button hijacking | Manipulating browser history navigation |
| Site abused with third-party spam | Spammy third-party content on your (often free-host/platform) property |
| User-generated spam | Spam in forums/comments/profiles |
| Spammy free host | A hosting service where a large fraction of sites are spam |
| Structured data issue | Markup outside the SD guidelines — penalty = rich-result ineligibility only (→ Ch 12 §12.11) |
| Unnatural links TO your site | Bought/manipulative inbound links |
| Unnatural links FROM your site | Selling/passing unqualified outbound links |
| Thin content with little or no added value | Doorways, thin affiliation, scraped/auto-generated filler |
| Cloaking and/or sneaky redirects | Different content to Google vs users; deceptive redirects |
| Major spam problems | Aggressive, pure spam |
| Cloaked images | Image swaps/obscuring |
| Hidden text and/or keyword stuffing | The classics |
| AMP content mismatch | AMP ≠ canonical content |
| Sneaky mobile redirects | Mobile-only spam redirects (also the hacked variant → Ch 17) |
| News and Discover policy violations | Multiple sub-types for those surfaces |
| Site reputation abuse | §18.3 |
🟢 Scope: an action hits a pattern — "This can be a subset of your site, or the entire site. Not all of the pages matching the pattern are necessarily affected." Partial matches list example patterns in the report; sitewide means everything.
⚠️ What is NOT here: no "hacked" action (Security Issues handles that → Ch 17 §17.6), no algorithmic-update entry, no CWV/experience action. And the report being empty is dispositive: quality-system losses never show here.
18.3 Site reputation abuse — the marquee policy of the window¶
🟢 Definition (live page): "a tactic where third-party content is published on a host site mainly because of that host's already-established ranking signals, which it has earned primarily from its first-party content."
- 🟢 The named IS-abuse examples — note the vertical relevance: "a medical site hosting a third-party page about 'best casinos'," educational sites hosting sponsored payday-loan reviews, news sites hosting white-label coupons, and ⚠️ "an established site using freelance content in new areas for ranking advantage" — first-party commissioning doesn't launder it.
- 🟢 The NOT-abuse list (the safe harbor clients need): wire/press-release services, syndicated news, genuine UGC/forums, editorial and opinion content, third-party content promoted for readers (not rankings), affiliate links properly
rel=sponsored/nofollow, merchant-sourced coupons. - 🟢 The Nov 2024 clarification (playbook line 235), now verified against the announcement post itself (raw-HTML fetch, 2026-08-04 — the SPA shell yields to curl): "We're making it clear that using third-party content on a site in an attempt to exploit the site's ranking signals is a violation of this policy — regardless of whether there is first-party involvement or oversight of the content." Google's stated basis: having reviewed "white-label services, licensing agreements, partial ownership agreements, and other complex business arrangements… no amount of first-party involvement alters the fundamental third-party nature of the content or the unfair, exploitative nature" of the tactic.
- ⚠️ Technical remediation shape: the policy targets sections exploiting host signals — containment is sectional: remove, or
noindexthe section, or move it off-host (subdomain/domain with its own standing — noting a GSC Domain property still groups subdomains → Ch 19). 🟢 ⚠️ noindex does not auto-clear the manual action — the post's own FAQ: "If I noindex the content, does that mean the manual action automatically gets removed? No. You still need to reply to the manual action in Search Console and explain that the content has been noindexed." Containment plus reconsideration, always (§18.4). - Architecture cross-ref: leased-section detection is structural (distinct templates, commercial density, off-topic clusters) → Ch 10; the 37-domain-network context in
reviews/shows the adjacent risk pattern from the satellite side.
18.4 Reconsideration — the process that actually works¶
🟢 From the report help (fetched live): fix all affected pages, ensure Google can crawl them, then Request Review. "A good request does three things: Explains the exact quality issue on your site. Describes the steps you've taken to fix the issue. Documents the outcome of your efforts."
- ⚪ Practitioner shape of a strong request (consistent with the three requirements): specific inventory of what was wrong ("2,300 doorway pages under /city/"), dated remediation evidence (removal lists, link-cleanup spreadsheets, disavow if used → playbook line 180's two-part disavow rule), and the systemic fix (who/what changed so it won't recur). Contrition without documentation fails; documentation without root-cause fails slower.
- 🟢 Timelines: "Most reconsideration reviews can take several days or weeks, although in some cases, such as link-related reconsideration requests, it may take longer."
- ⚠️ Do not re-request without new work — serial identical requests read as noncompliance. (Cf. the Safe Browsing repeat-offender lockout for the security track → Ch 17 §17.5; manual actions have no documented lockout, but reviewer patience is finite.)
- ⚠️ Buying a domain? Check for inherited manual actions and leftover sitewide URL removals before paying (→ Ch 16 §16.6) — actions travel with the site, not the owner.
18.5 Algorithmic spam enforcement — the other half¶
- 🟢 Spam updates (announced on the Search Status Dashboard → playbook Phase 7) apply the same policies algorithmically; SpamBrain is the named system — "SpamBrain is our AI-based spam-prevention system" (spam-updates page, fetched 2026-08-04). Link-spam neutralization, in the live page's words: "when our systems remove the effects spammy links may have, any ranking benefit the links may have previously generated for your site is lost… Any potential ranking benefits generated by those links cannot be regained" (playbook line 180's basis; the Dec 2022 post's "neutralize the impact of unnatural links" phrasing is 🟡 — post body unfetchable). No notice, no review path: recovery = remove the violation and wait for the systems to reprocess — typically the next spam update for classifier-type hits.
- ⚠️ Diagnosis fork that decides everything (→ Ch 21): drop + Manual Actions entry → fix + reconsider. Drop coinciding with a dashboard-listed spam update + no entry → algorithmic; fix + wait; no one to write to. Drop with neither → it's probably not spam at all; work the quality/technical tracks.
- 🟢 Adjacent quality reclassifications the playbook already carries (lines 243–249): fake bylines, fake freshness, unqualified-YMYL are E-E-A-T harms with no manual action — never promise a "penalty removal" for them; there is no penalty to remove.
Symptoms & diagnosis¶
| Symptom | Likely cause | Where |
|---|---|---|
| Sudden sitewide collapse + GSC notification | Manual action — read the report's scope and pattern | §18.2 |
| Collapse on a dashboard spam-update date, no GSC entry | Algorithmic spam hit — fix + wait, no reconsideration exists | §18.5 |
| Rich results gone, rankings intact | Structured-data action — scoped to display only | §18.2, Ch 12 |
| Coupon/review/"partner" section tanked | Site reputation abuse (action or algorithmic) — sectional containment | §18.3 |
| "Penalty" claimed but Manual Actions report empty | Not a penalty — quality or technical track | §18.5, Ch 21 |
| Reconsideration rejected | Partial cleanup, missing documentation, or unfixed root cause | §18.4 |
| Acquired domain won't rank | Inherited action / sitewide removal / expired-domain-abuse profile | §18.4, Ch 16 |
| Affiliate pages drifting down across updates | Thin-affiliation exposure — add value or qualify links | §18.1 |
Sources¶
Fetched live 2026-08-04 by the drafter: - Manual Actions report (full action list + definition + reconsideration process) - Spam policies (2026-05-15; 16 named policies; site-reputation-abuse definition + examples + consequence) - Spam updates list (SpamBrain definition + link-spam-update effects wording; fetched 2026-08-04) V2 resolutions (2026-08-04): Nov 2024 site-reputation-abuse announcement body still unfetchable (SPA shell) — key clarification wording carried 🟡 per SER writeup; Dec 2022 "neutralize" phrasing demoted to 🟡, replaced in-text with the live spam-updates page wording (🟢).