$ cat Historical Timeline
A sourced chronology of search-policy and ranking events, with historical interpretations separated from confirmed announcements.
1,079 words~5 min read4 tablesREFERENCE
2011 2012 2019 2024 2026
Panda ──▶ Penguin ──▶ BERT ──▶ Core+Spam ──▶ AI-abuse policy
●─────────●───────────●──────────●───────────────●
SOURCED HISTORICAL TIMELINE
A selective chronology of documented announcements, enforcement accounts and research. Each row links to its source and states what the event can establish. Publication, effective and observed dates are distinguished where they differ. This is not a causal history of every ranking change or a measurement of when a technique became popular.
Foundations and documented cases (2005–2011)
| Date / type | Event | Meaning and limit | Source |
|---|---|---|---|
| 2005-01-18 | nofollow announced | Google described a link attribute intended to reduce the ranking incentive for comment spam. This historical treatment was revised in 2019; it does not describe current link handling by itself. | Google: Preventing comment spam |
| 2006-02-07 | BMW reinclusion reported | Matt Cutts reported BMW’s reinclusion after doorway-page removal. This is a specific enforcement account, not a standard recovery timetable. See the case record for scope. | Matt Cutts: Recent reinclusions |
| 2011-02-24 | Quality-ranking update announced | Google announced a change intended to reduce low-quality results, initially in the US. The source describes the rollout and intention; it does not prove that every content farm collapsed. | Google: Finding more high-quality sites in search |
| 2011-04-25 | Overstock reports penalty lifted | A company release filed with the SEC states that a Google penalty imposed in late February had been lifted. Company reporting is primary for its statement, not an independent audit of the full business impact. | Overstock: SEC-filed release |
| 2010-09 | SpamIt closes; GlavMed investigation | A criminal investigation into the operator and the closure of the SpamIt affiliate program coincided with a documented drop in global spam. Attribution is to named operators; this is investigation and reporting, not a completed conviction record. | Krebs on Security |
| 2011 | Spam value chain mapped | The Click Trajectories study followed the pharma spam chain end to end and identified the acquiring-bank layer as its demonetization chokepoint. A peer-reviewed study of the measured era, not a claim about current operators. | Levchenko et al. (IEEE S&P) |
Link-spam system changes (2012–2016)
| Date / type | Event | Meaning and limit | Source |
|---|---|---|---|
| 2012-04-24 | Webspam update announced (Penguin) | Google announced an algorithm change addressing webspam including keyword stuffing and link schemes. The announcement does not date the birth of private blog networks or establish an individual site’s cause of loss. | Google: Another step to reward high-quality sites |
| 2016-09-23 | Penguin becomes part of core | Google described real-time refresh after recrawling/reindexing and more granular spam devaluation. This does not mean all link problems resolve immediately or exclude manual actions. | Google: Penguin is now part of our core algorithm |
Policy and result-feature changes (2019–2023)
| Date / type | Event | Meaning and limit | Source |
|---|---|---|---|
| 2019-09-10 | Link attributes become hints | Google introduced sponsored and ugc attributes and changed how it describes nofollow. The announcement says the attributes are hints for ranking, with nofollow becoming a crawling/indexing hint from March 1, 2020. | Google: Evolving nofollow |
| 2022-08-18 | Helpful content update announced | Google announced the helpful content initiative. This is an announcement date, not a claim that rollout completed that day. Its separate-system terminology is historical; see the March 2024 integration. | Google: August 2022 helpful content update |
| 2023-09-13 | HowTo rich results end on desktop | The September 14 update to Google’s August announcement records the final desktop removal on September 13. HowTo rich results no longer appeared in Google Search; the Schema.org vocabulary itself was not abolished. | Google: Changes to HowTo and FAQ rich results |
Current-policy context and AI research (2024–2026)
| Date / type | Event | Meaning and limit | Source |
|---|---|---|---|
| 2024-03-05 | Core update and three named abuse policies | Google announced scaled content abuse, expired domain abuse and site reputation abuse policies. The same announcement explains that helpful-content assessment is integrated into core systems. The initial site reputation policy had a later enforcement date. | Google: March 2024 core update and new spam policies |
| 2024-05-29 | Google confirms authenticity of leaked documents | The Verge reported Google’s confirmation concerning the leaked material. Authentic documentation does not establish that every field is active, weighted as inferred or usable to change ranking. This row is attributed reporting, not direct publication by Google. | The Verge: Google confirms leaked Search documents |
| 2024-10-21 | US consumer-review rule takes effect | The FTC’s rule addresses specified fake-review and testimonial practices. It does not make every incentive or every search-policy violation a criminal offense. Refer to the regulator’s definitions and exceptions. | FTC: Consumer Reviews and Testimonials Rule Q&A |
| 2024-11-19 | Site reputation abuse language updated | Google’s historical clarification states that first-party oversight does not by itself excuse exploitation of host ranking signals. Read the 2026 revision before using this wording as current guidance. | Google: Updating our site reputation abuse policy |
| 2025-04-06 | UK fake-review provisions take effect | UK government guidance records commencement of the new consumer protections. The applicable fake-review duties and enforcement route require the statute and CMA guidance; a blanket criminality claim is inappropriate. | UK government: Fake reviews and hidden fees banned |
| 2026-03 / 2026-04 | Studies of AI search and indirect injection | Separate preprints examine black-hat manipulation of LLM-enhanced search and indirect prompt injection. Publication months identify the studies, not the date either technique first existed. The AI/GEO and prompt-injection chapters discuss methods and limits separately. | Chen et al.: LLM-enhanced search study |
| 2026-05-07 | FAQ rich results removed | Google’s May 8 documentation log says FAQ rich results stop appearing from May 7. Its June 15 entry removes the documentation. The previous government/health limitation is historical, not the current endpoint. | Google: Search documentation updates |
| 2026-08-28 / 2026-08-30 | Site reputation policy revision and EEA treatment | Google announced revised treatment, including consideration of host involvement and integration, with an August 30 effective date for the described EEA change. Consult the current policy and region-specific wording; no one factor guarantees eligibility. | Google: Update to the Site Reputation Policy |
| 2025 (Jan-Apr, reported) | SEO-poisoning malware campaign | Vendor reporting described an SEO-poisoning campaign impersonating IT and AI tools that reportedly reached 8,500+ SMB users, showing the pharma-hack mechanics repurposed for malware delivery. Scope figures are as reported. | The Hacker News (reporting Kaspersky) |
The second 2026 study is Khodayari et al., Indirect Prompt Injection in the Wild. Current platform language is maintained in Google’s spam policies; UK review-law detail is in the CMA guidance.
WHAT A TIMELINE DOES NOT PROVE
A loss near an update is temporal association. Establishing causation requires evidence about the affected pages, concurrent deployments, measurement, comparison cases and alternative explanations. Older unsourced claims about Florida, “Medic,” the birth of PBNs and universal tactic decline have been removed from this timeline. Named incidents and unresolved outcomes remain in the case database, where their evidence can be assessed separately.