把 Google Search Central blog 2016-2024 关于 webspam、site reputation、hacked sites、link spam、manual actions、disavow 的 26 篇博文做了一次跨年合并。不是逐篇摘要,而是按"政策语义"重新组合 — 这样你做策略判断时能直接读到完整框架,不用回头翻年表。
Source 文件路径见 frontmatter sources: 字段(全部相对于 ~/kb/)。
TL;DR 三句话
- 2024 是分水岭年:Google 第一次系统化把 spam 政策从"传统 link/cloaking/hacked"扩展到针对 SEO 业内灰产的三大新条款 — expired domain abuse(老域名重用)、scaled content abuse(AI 大规模生成)、site reputation abuse(寄生站点 / Coupons Forbes 那一套)。2024-11 进一步收紧 site reputation 措辞,堵掉"白标合作 / 合资 / partial ownership"的洗白通道。
- SpamBrain (2018) 是底层基础设施,2021 后所有 spam update 都基于它。2021 vs 2020 抓出 6x spam sites;2022 vs 2021 抓出 5x;link spam 2022 抓出 50x;hacked spam 抓出 10x。SpamBrain 不只是"分类器",而是一个被复用的"反滥用 robust platform"(harassment、product reviews、scam 都借它跑)。
- Google 反复强调 disavow / spam report 大多数人不需要用。spam reports 只用来"改进算法",不是触发 manual action 的主要路径(2020-07 政策更新明确写死)。disavow 在这套叙事里同理 — 算法 90% 已经 nullify 了 link spam,大多数站长动手只会自伤。
Part 1 — 2024 三大新政策:Site Reputation Abuse + Scaled Content + Expired Domain
2024 年 3 月 5 日 March 2024 core update 与 spam update 同日发布,Chris Nelson 公布三条新 spam 政策,这是 Google 首次系统性回应 2022-2023 业内"parasite SEO"、"AI 内容洪水"、"老域名重启"三大新型黑帽。
Source: 2024/03/core-update-spam-policies.en.md;2024/11/site-reputation-abuse.en.md
1.1 Site Reputation Abuse(寄生站点 / Parasite SEO)
原始定义(2024-03):
"Site reputation abuse is when third-party pages are published with little or no first-party oversight or involvement, where the purpose is to manipulate Search rankings by taking advantage of the first-party site's ranking signals."
典型表现 — 高权重出版物(Forbes、CNN、新闻门户)把 /coupons/、/casino-reviews/、/best-vpn/ 等子目录卖给第三方,第三方放低质内容借母站权重排名。
关键日期:
- 2024-03-05:政策宣布,生效推迟到 2024-05-05 给站点缓冲期
- 2024-11-19:政策语言收紧(见下)
- 2025-01-21:基于 FAQ 反馈做"editorial changes",政策实质不变
2024-11 收紧的核心动作 — 堵 "first-party involvement" 这个洗白通道:
"no amount of first-party involvement alters the fundamental third-party nature of the content or the unfair, exploitative nature of attempting to take advantage of the host's sites ranking signals."
新版语言简化为:
"Site reputation abuse is the practice of publishing third-party pages on a site in an attempt to abuse search rankings by taking advantage of the host site's ranking signals."
FAQ 里几个判定要点(2024-12 补充 + 2025-01 收录入正式 policy):
| 问 | 答 |
|---|---|
| Freelance content 是不是违规? | 不是 — 除非 ALSO 有"taking advantage of host site's ranking signals"的意图 |
| Affiliate content 是不是? | 不是 — 只要 rel="sponsored" 标了 |
| Coupon 区是不是? | 看 "active or close involvement" — 出版商自己从商户 source vs 接 white-label 服务,完全两回事 |
| noindex 之后 manual action 自动撤销吗? | 不会,还得回复 manual action 并 reconsideration |
| 把内容挪到子域 / 子目录? | 不行,会被视为 "circumvention",可能引发"broader actions against a site" |
| 挪到全新无信誉的域名? | OK,前提是新域没建立信誉、且不违反其他 spam policies |
| 老→新 301 重定向 ? | 不要做,会把 site reputation abuse 问题搬过去 |
| 老→新 链接? | 用 rel="nofollow" |
算法层面配合(2024-11 提到):
"we also have systems and methods designed to understand if a section of a site is independent or starkly different from the main content of the site. By treating these areas as if they are standalone sites, it better ensures a level playing field..."
— 这意味着即使没人工 action,寄生子目录也会被自动剥离 site-wide signal,造成流量下滑,但 Google 明确说"this doesn't mean these sub-sections have been demoted or are in violation of our spam policies. It means we're measuring them independently"。
1.2 Scaled Content Abuse(AI 大规模生成)
定义:
"Scaled content abuse is when many pages are generated for the primary purpose of manipulating Search rankings and not helping users... no matter how it's created."
关键政策意图 — 把旧的 "auto-generated content" 政策升级到"不区分人 / 自动 / 混合":
"Our long-standing spam policy has been that use of automation, including generative AI, is spam if the primary purpose is manipulating ranking in Search results. The updated policy is in the same spirit of our previous policy and based on the same principle. It's been expanded to account for more sophisticated scaled content creation methods where it isn't always clear whether low quality content was created purely through automation."
— 翻译人话:你雇 100 个外包写手批量出 SEO 软文,跟你用 GPT 批量出软文,视同 spam。Google 不再纠缠"AI 还是人写"。
和 2023-02 AI 内容指南一致(见 2023 webspam report 引用):
"if AI is used for the primary purpose of manipulating search rankings, that's a violation of our long-standing policy against spammy automatically-generated content."
1.3 Expired Domain Abuse(老域名重用)
定义:
"Expired domain abuse is where an expired domain name is purchased and repurposed primarily to manipulate Search rankings by hosting content that provides little to no value to users."
经典案例(原文给的):
"someone might purchase a domain previously used by a medical site and repurpose that to host low quality casino-related content, hoping to be successful in Search based on the domain's reputation from a previous ownership."
关键边界:
"It's fine to use an old domain name for a new, original site that's designed to serve people first."
— 拿老域名做正常生意没事,但专门买老域名+灌垃圾内容就是 spam。Google 明确这"不是无心之失",是产业链行为("isn't something people accidentally do")。
Part 2 — Link Spam:从 Penguin 到 2022 link spam update 的演化
2.1 时间线
| 时间 | 事件 | 关键点 |
|---|---|---|
| 2016-09 | Penguin 4.0 实时化 | 不再是"周期性运行的算法",而是 core ranking 的一部分 |
| 2017 年报 | 双倍力度移除 unnatural links | 同比 link spam 下降 ~50% |
| 2018 年报 | 持续 nullify bad links | 强调"创造好内容比研究 link 神话更重要" |
| 2019-09 | rel="nofollow" 演化 | 引入 rel="sponsored" 和 rel="ugc";三者变成 hints 而非 directives |
| 2020 年报 | 90%+ link spam 算法已捕获 | paid links、link exchange 已经"made less effective" |
| 2021-07-26 | Link Spam Update | 主目标 affiliate links 未标 rel="sponsored" + sponsored/guest posts;两周 rollout,跨多语言 |
| 2022-12-14 | December 2022 Link Spam Update | 第一次 SpamBrain 接管 link spam;两周 rollout,全语言;同时抓 buying sites + outgoing pass-through sites |
| 2022 年报 | 同比 2021 抓出 50x link spam sites | 验证 SpamBrain 接管效果 |
2.2 Google 反复强调的 link 立场(逐年原话)
2018 年报 — 反"建链神话"宣言:
"We continued to remind website owners that if you simply stay away from building links mainly as an attempt to rank better and focus on creating great content, you should not have to worry about any of the myths or realities."
2019 年报 — nofollow 演进:
"Originally introduced as a means to help fight comment spam and annotate sponsored links, the 'nofollow' attribute has come a long way. But we're not stopping there."
2021-07 link spam update 的执行细则:
- 所有 affiliate 必须
rel="sponsored",手动或动态生成皆然 - 接受 sponsored / guest posts 都得标
- 不标 → 既会触发 manual action,也会触发算法 action
- "Both manual and algorithmic actions may affect how we see a site in Search"
2022-12 SpamBrain link update 的精确措辞:
"Besides using it to detect spam directly, it [SpamBrain] can now detect both sites buying links, and sites used for the purpose of passing outgoing links."
— 注意"used for the purpose of passing outgoing links"这一句:即使你自己没接钱,只要你被 SpamBrain 判定结构上服务于 outgoing link,也会被打。
2.3 link tagging 完整速查表
| 情形 | 应该用什么 |
|---|---|
| 接钱的链接、affiliate | rel="sponsored" |
| 用户生成内容(评论 / 论坛) | rel="ugc" |
| 不想 endorse 但又不广告 | rel="nofollow" |
| 多重场景 | 可以 rel="ugc sponsored" 等组合 |
| 默认正常引用 | 不用任何 rel |
三者从 2019-09 起 Google 都当 hints 处理,不是 directives — 也就是说 Google 仍可能选择性 count 它们做 ranking。
Part 3 — Hacked Sites / Malware:nohacked 系列 + 恢复流程
3.1 整个 nohacked 知识包(2017 系列 4 篇 + 2018 日文版)
| 子主题 | 来源 |
|---|---|
| 侦测篇(How do I know if my site is hacked?) | 2017/12/nohacked-30-how-do-i-know-if-my-site-is |
| 预防篇(Tips on prevention) | 2017/12/nohacked-30-tips-on-prevention |
| 修复篇(Fixing common hack cases) | 2017/12/nohacked-30-fixing-common-hack-cases |
| 资源篇(Useful resources,只有非英文版) | 2017/12/nohacked-30-more-resources |
| 日文专文 hack-resistant | 2018/03/strong-protection-against-hacking |
3.2 三类典型 hack 案件(2017 年报+nohacked 系列里反复提)
| 名字 | 模式 | 修复指南链接(在 web/fundamentals/security/hacked) |
|---|---|---|
| Gibberish Hack | 自动生成大量乱码 keyword 页面 → 用户点击被重定向到色情 / 垃圾站 | fixing_the_gibberish_hack |
| Japanese Keywords Hack | 在随机目录里生成日文页面,挂假货 affiliate;有时把 hacker 账号加成 Search Console site owner | fixing_the_japanese_keyword_hack |
| Cloaked Keywords Hack | 用 cloaking 让页面在抓取时显示 spam、用户访问时是 404 或正常页 | fixing_the_cloaked_keywords_hack |
3.3 侦测六大动作(2017-12 how-do-i-know)
- 看 Google Search 警告类型:
This site may be hacked(挂广告)/This site may harm your computer(挂 malware)/ 红屏(malware/unwanted software/deceptive site / phishing) - 分清 Malvertising vs Hack:Malvertising 是页面上加载了坏广告,不是站本身被改
- 查 open redirects(OWASP unvalidated redirects 那一套)
- mobile incognito 自查 — 有些 bad ad networks 只对移动端触发
- Search Console 验证 + 看消息(关键!)
- 不确定就问专家或 Webmaster Help Forums
2017 年报数据:hacked sites 同比 +32%(2016 vs 2015 是 +180%);61% 被黑站长根本没收到 Google 通知 — 因为他们没注册 Search Console(2017-03 nohacked year in review)
3.4 预防三件事(2017-12 prevention)
- 理解最常见的入侵路径(top_ways_websites_get_hacked_by_spammers)
- 慎用免费版的付费插件 / 主题 — "premium plugins available without payment" 99% 是后门
- WordPress / Joomla 防 botnet — 即使没明显改动,你的站可能已经在 botnet 里跑 spam / DDoS / clickbot
3.5 恢复流程
- 侦测 → 找到入侵点
- 清理(按 hack 类型走对应 fixing 指南)
- filing for reconsideration request(2017 年报数据:84% 申请 reconsideration 的最终成功清理)
- Google 算法重新评估,警告撤销
3.6 SpamBrain 时代的 hacked spam 数据
| 年报 | 数据 |
|---|---|
| 2018 年报(2017 数据) | "detect and remove from search results more than 80% of these sites" |
| 2019 年报(2018 数据) | hacked spam 影响"持续下降",更快侦测 |
| 2020 年报(2019 数据) | hacked spam 比往年更稳定;Cherry Prommawin + Duy Nguyen 联署 |
| 2021 年报(2020 数据) | 侦测能力 +50%,"removed most of the hacked spam" |
| 2022 年报(2021 数据) | hacked spam 减少 70%(SpamBrain 主要功劳)+ gibberish spam on hosting platforms 减少 75% |
| 2023 年报(2022 数据) | hacked site detection 10x 提升 vs 2021 |
Part 4 — Manual Actions + Reconsideration 流程
4.1 manual action 的 Google 内部叙事
历年 webspam 年报反复声明同一个结构:
算法系统是主力 + manual action 是补充 + Search Console 通知是唯一官方沟通渠道 + reconsideration 是出狱通道
具体数据演化(从年报抽出来的 raw numbers):
| 年 | manual action 数量 | 数据点 |
|---|---|---|
| 2015 年报 | 4.3M 消息 | 4.3M messages total 都包含 manual action |
| 2016 年报 | 9M+ 消息 / 10,000+ structured data 站被 manual action | "took manual action on more than 10,000 sites that did not meet the quality guidelines for inclusion in search features powered by structured data" |
| 2017 年报 | 6M+ 消息 related to manual actions(总 45M 消息) | |
| 2018 年报 | 4M+ manual actions(总 186M 消息,96M 成功投递) | |
| 2019 年报 | 4.3M manual actions(总 90M+ 消息) | |
| 2020 年报 | 2.9M manual actions(总 140M+ 消息) | 关键拐点:manual action 持续下降,因为"many spam issues have already been taken care of by our algorithms" |
— 这条曲线讲清楚一件事:Google 长期想退出 manual action 业务,把 enforcement 完全交给算法 / SpamBrain。
4.2 reconsideration 成功率参考
- 2015 年报:reconsideration 申请数 +33%
- 2017 nohacked year in review:84% 的 reconsideration 申请最终清理成功(hacked sites 维度)
4.3 三种 manual action 高发场景(从 Webmaster Guidelines 摘)
- link schemes(2018 年报 / 2021 link tagging / 2022-12 link spam update)
- structured data 不合规(2016 年报 10,000+ 站被打)
- third-party spam(本站被 spammer 滥用 forum / hosting / file uploader / internal search)(2021-05 prevent portions of site from spam — 这篇专门讲这个)
4.4 第三方滥用本站怎么自救(2021-05 prevent portions)
完整 checklist:
- block automated account creation — reCAPTCHA / Securimage / Jcaptcha + 邮箱验证 + 黑名单可疑域名
- 打开 moderation features — 新用户需审核、禁匿名 post、links 需 reputation 门槛
- monitor spam content:
- 注册 Search Console + 看 Security Issues + Manual Actions report
- 用
site:your-domain-name viagra这类 query 检查 - Google Alerts 跑
site:your-domain-name spammy-keywords - 看 server log 的 traffic spike
- 识别终止 spam accounts — 短时间大量注册、同 IP 段、奇怪 user agent、商业用户名
- noindex 或 disallow 不可信内容 —
<meta name="googlebot" content="noindex">或 robots.txtDisallow: /guestbook/ - UGC 链接打
rel="ugc"或rel="nofollow" - 集中 UGC 到一个子目录 —
example.com/user-generated-content-dir-name/而不是example.com/best-online-pharma-buy-red-viagra-online - 保持 CMS 更新 — Akismet / 安全插件
关键引用:"Low-quality content on some parts of a website can impact the whole site's ranking." — 子目录 spam 会拖整站,跟 site reputation abuse 是同一硬币的两面。
Part 5 — Disavow Tool 的政策意图(为什么 Google 反复说"大多数人不需要")
26 篇里没有专门一篇讲 disavow,但多处反复传达同一个口径:
5.1 spam reports 的政策定位 — 2020-07 唯一一篇专题
Source: 2020/07/how-spam-reports-are-used-at-google.en.md (Gary Illyes 署名)
核心一句(2020-07 更新到 quality guidelines):
"we use spam reports only to improve our spam detection algorithms"
Gary 的辩护逻辑:
"Most of the time, it's much more impactful for us to fix an underlying issue with our automated detection systems than it is to take manual action on a single URL or site." "In theory, if our automated systems were perfect, we would catch all spam and not need reporting systems at all."
— 这条政策可以直接迁移到 disavow 的语义:Google 的态度是"我们已经自动 nullify,你别瞎操心"。
5.2 disavow 在 2020 年报的位置 — 工具说明而非主战场
Source: 2021/05/working-with-you-for-safe-and-spam-free-search-2020.en.md
提到 Disavow links tool 的语境是 Search Console 工具列表里的一项,跟 "new version of the Crawl stats report"、"SafeSearch filtering report" 并列出现 — 不再被定位为 SEO 主战略工具。
5.3 反复重申的"link 三神话破除"
跨多年(2018/2019/2021/2022 年报 + 2021-07 link spam update)Google 用几乎相同的话术劝退 disavow:
| 神话 | Google 反驳原话 |
|---|---|
| "差评 link 一定要 disavow" | 2018 年报:"algorithms... made a number of bad linking practices less effective for manipulating ranking" |
| "我得手动管理外链画像" | 2019 年报:"if you simply stay away from building links mainly as an attempt to rank better and focus on creating great content, you should not have to worry" |
| "Negative SEO 攻击需要 disavow 防御" | 2020 年报:"more than 90% of link spam was caught by our systems" |
| "2022 link spam update 后我得自查 disavow" | 2022-12 update:"As we have always emphasized, links obtained primarily for artificial manipulation of Search rankings are link spam. Our algorithms and manual actions aim to nullify these unnatural links at scale" |
— 政策意图很清晰:Google 不希望站长把精力花在 disavow 上,因为(a)算法已经 nullify 了大多数 case;(b)站长自己乱 disavow 反而可能伤到好链接;(c)disavow 工具的存在主要是给那些"自己有意建过 spam links 现在要洗白"的站长用的。
Part 6 — Deceptive Behaviors:back button hijacking / deceptive downloads
6.1 Safe Browsing 的覆盖面演化
| 年 | Safe Browsing 新增覆盖 | 来源 |
|---|---|---|
| 2007 | webmaster tools 集成 | 历史回顾 |
| 2015 | Deceptive Sites + Unwanted Software | 2016-09 引用 |
| 2016-02 | deceptive download buttons | 2017 年报引用 |
| 2016-11 | repeatedly dangerous sites | 2017 年报引用 |
| 2017-03 | macOS 防护扩展 + Chrome predictive phishing protection | 2018 年报引用 |
| 2017-12 | mobile unwanted software | 2018 年报引用 |
| 2018-01 | deceptive Chrome extension installation | 2018 年报引用 |
| 2020-2021 | 反 customer support scam / tech support scam(扩展到 hundreds of millions of searches/yr) | 2021 年报 |
6.2 Search Console Security Issues report 六大分类(2016-09)
Source: 2016/09/more-safe-browsing-help-for-webmasters.en.md
更新后的 Security Issues report 给出 6 类 specific explanations:
- malware
- deceptive pages(social engineering)
- harmful downloads
- uncommon downloads 5.(+ 推断有 phishing / unwanted software 等子项)
每类都配:tailored recommendations + sample URLs + remediation actions。
6.3 Safe Browsing Site Status Tool(2017-03)
Source: 2017/03/updates-to-google-safe-browsings-site.en.md
新版 UI 提供 simpler clearer results,目标用户:"people who are visiting the tool from a Safe Browsing warning they've received, or doing casual research on Google's malware and phishing detection"。技术细节(AS / ASN)挪到 malware dashboard。
6.4 移动端 sneaky redirects + back-button hijacking 模式
Source: 2017 年报
2016 年观察到的趋势:
"we saw a rise in spam that redirects users, without the webmaster's knowledge, to other sites or pages, inserted into webmaster pages using widgets or via ad units from various advertising networks."
— 这是 back-button hijacking / sneaky mobile redirect 在年报里出现的标志性表述。Google 2017 年起持续打击。
6.5 Scams & Fraud(2020 年报 + 2021 年报)
customer support scam / tech support scam 是 2020 年报新增主题:
"many scammers pretend to be offering customer support phone numbers to popular services and products, only to trick users who call in into paying them via bank transfers or gift cards."
数据:
- 2018+ 开始系统化保护,每年 hundreds of millions of searches 被保护
- 2021 年:再扩 40% 的 scam coverage
- 2022 年:扩到所有语言,clicks on scam sites 同比 -50%
6.6 Online Harassment(2021 年报新增)
Source: 2022/04/webspam-report-2021.en.md
SpamBrain 扩展到 harassment 场景,for name queries, "reduce the prominence of sites with exploitative removal practices"(指那种 mugshot 站、收钱删黑历史那种)。
Part 7 — 历史教训:从老 link scheme 到现行框架
7.1 webspam 年报"老语言→新语言"演化
| 时段 | 主流 spam 类型(年报第一段提的) | 主要工具 / 系统 |
|---|---|---|
| 2015 年报 | hacked sites(+180%)+ thin/low-quality content | 算法 + manual + 4.3M messages |
| 2016 年报 | hacked +32% + 移动 sneaky redirects + CMS UGC abuse | Penguin 4.0 实时化 + 9M messages |
| 2017 年报 | hacked spam(detect & remove 80%+)+ CMS UGC + link scheme | "unnatural links 同比 -50%" + 45M messages |
| 2018 年报 | hacked + user-generated spam(-80%)+ link spam | 改 rel=nofollow → sponsored/ugc;186M messages |
| 2019 年报 | scraped/auto-gen content(-60% impact)+ scam/fraud | 90M messages,manual action 4.3M |
| 2020 年报 | SpamBrain 公开亮相 + AI-based spam fighting | 99% spam-free + auto-gen sites -80% + 140M messages |
| 2021 年报 | SpamBrain(+6x sites)+ link spam update + harassment 扩展 | hacked -70%,gibberish on hosting -75% |
| 2022 年报 | SpamBrain link spam +50x,hacked +10x;AI 内容指南发布 | scam clicks -50% |
| 2023(10 月 spam update) | cloaking + hacked + auto-gen + scraped 在 TR/VI/ID/HI/ZH 等语言 | spam update |
| 2024-03 | 3 大新政策落地 + March 2024 core update 同步 | 政策驱动而非系统驱动 |
| 2024-11 | site reputation abuse 政策语言收紧 | 堵 first-party involvement 洗白通道 |
7.2 关键的"教训"金句(可直接做策略 PPT 引用)
"内容质量是 anti-spam 的根本"(2019 年报):
"We think that one of the best ways of fighting spam of all types is by encouraging website owners to just create great quality content."
"99% spam-free 这个比率长期稳定"(2018-2022 年报反复):
- 2017 数据:"less than 1 percent of search results users visited are spammy"
- 2019 数据:"well under 1 percent"
- 2020 数据:"more than 99% of visits from our results lead to experiences without spam"
- 2021 数据:"more than 99% of searches spam-free"
- 2022 数据:"more than 99% of visits from Search are spam-free"
"40 billion spammy pages discovered every day"(2020 数据,2021 年报) "25 Billion pages discovered each day are spammy"(2019 数据,2020 年报) "SpamBrain 2018 launched, 200x more spam sites caught than launch baseline"(2022 年报)
7.3 SpamBrain 自身的演化(独立时间线)
| 时间 | 节点 |
|---|---|
| 2018 | SpamBrain 首次部署(后续年报追溯认定) |
| 2020-06 | 2020 年报第一次公开点名 "AI-based spam detection" |
| 2021 数据 | SpamBrain 抓出 6x sites vs 2020;hacked -70%、gibberish -75% |
| 2022-04 | 2021 年报正式命名 "SpamBrain";扩到 harassment + product reviews + scam |
| 2022-12 | SpamBrain 接管 link spam(link spam update by SpamBrain) |
| 2022 数据 | 5x sites vs 2021;link spam 50x;hacked 10x;crawling-time spam 提速 |
| 2023-10 | spam update 主打多语言(TR / VI / ID / HI / ZH)cloaking + hacked + auto-gen + scraped |
| 2024-03 | March 2024 spam update 与三大新政策同发 |
7.4 Search Console / Removals report 周边工具演化
Source: 2020/01/new-removals-report-in-search-console.en.md
Removals report 三大功能(2020-01 上线):
- Temporary removals — 隐藏 URL ~6 个月(buy time to find permanent solution)
- Temporary remove URL:从结果隐藏 + 清 cache
- Clear cache URL:只清 cache + 清 snippet
- Outdated content — 公众通过 Remove Outdated Content tool 提交,反映"页面已经没有这内容了"
- SafeSearch filtering — 用户通过 SafeSearch Suggestion tool 标记的成人内容历史
— 这是 site owner 应对 hacked spam / 第三方 spam 的"临时止血"工具,跟长期 manual action / reconsideration 是配套用法。
附录 A:跨年 webspam 年报数据卡
便于做演讲 / 报告时直接引用。
| 数据点 | 2015 | 2016 | 2017 | 2018 | 2019 | 2020 | 2021 | 2022 |
|---|---|---|---|---|---|---|---|---|
| User spam reports | 400K | 180K | 90K | 180K | 230K | n/a | n/a | n/a |
| 报告 acted-on 率 | 65% | 52% | n/a | 64% | 82% | n/a | n/a | n/a |
| Hacked sites 同比 | +180% | +32% | -80% impact | 持续↓ | 稳定 | -大幅 | -70% | -10x detect |
| Manual actions | 4.3M | 10K(结构化) | 6M+ | 4M+ | 4.3M | 2.9M | n/a | n/a |
| 总 messages 给站长 | 4.3M | 9M+ | 45M | 186M | 90M+ | 140M+ | n/a | n/a |
| Spam pages discovered/day | n/a | n/a | n/a | n/a | 25B | 40B | n/a | n/a |
| Spam-free 率 | n/a | n/a | <1% spam | <1% spam | <1% spam | >99% clean | >99% clean | >99% clean |
附录 B:Spam 政策核心引用页面索引
直接路径(都是 developers.google.com/search/docs/...):
- Spam policies overview:
/search/docs/essentials/spam-policies - Link spam 节:
/search/docs/essentials/spam-policies#link-spam - Site reputation 节:
/search/docs/essentials/spam-policies#site-reputation - Scaled content 节:
/search/docs/essentials/spam-policies#scaled-content - Expired domains 节:
/search/docs/essentials/spam-policies#expired-domains - Policy circumvention 节:
/search/docs/essentials/spam-policies#policy-circumvension - Qualify outbound links:
/search/docs/advanced/guidelines/qualify-outbound-links - Report spam:
/search/docs/advanced/guidelines/report-spam - Manual actions report doc:Search Console help / answer 9044175
- Reconsideration request:Search Console help / answer 35843
- Helpful Content FAQ:
/search/help/helpful-content-faq
附录 C:本 wiki 的边界
本 wiki 不覆盖:
- 算法系统机制(SpamBrain 内部结构 / Penguin 历史细节) → 见 Google 排名系统 + Core/Spam Updates + 业务信息建立 或 Google 公开论文
- Core updates 演化(独立于 spam update) → 见 GSC Blog: Core Updates 与算法迭代演化(2010-2026)
- E-E-A-T / Helpful Content 政策视角 → 见 GSC Blog: 质量政策与 E-E-A-T 演化(Webmaster Guidelines → Search Essentials)
- Search Console 工具使用细节(Disavow / URL Removal 操作) → 见 Search Console + Analytics + Trends 三件套用法 (流量诊断 / 关键词调研 / Looker Studio)
- Office hours 中的零散 spam Q&A → 见 Google SEO Office Hours 精华(2022.11–2024.08)+ 3 大 FAQ 主题汇总
两条法语 / 日语 only 的 stub 篇(本 wiki 跳过实质内容但保留在 sources 索引):
- 2016/10/equipe-safe-browsing-nouvelle-documentation(法语)
- 2018/03/strong-protection-against-hacking(日文)
- 2017/12/nohacked-30-more-resources(无英文,有 pt-BR/fr/ja/ru/zh-CN/th/zh-TW/tr 8 语言)
来源与关联资料
- https://developers.google.com/search/blog/2016/05/how-we-fought-webspam-in-2015?hl=en
- https://developers.google.com/search/blog/2016/09/more-safe-browsing-help-for-webmasters?hl=en
- https://developers.google.com/search/blog/2016/10/equipe-safe-browsing-nouvelle-documentation?hl=en
- https://developers.google.com/search/blog/2017/03/nohacked-year-in-review?hl=en
- https://developers.google.com/search/blog/2017/03/updates-to-google-safe-browsings-site?hl=en
- https://developers.google.com/search/blog/2017/04/how-we-fought-webspam-webspam-report?hl=en
- https://developers.google.com/search/blog/2017/12/nohacked-30-fixing-common-hack-cases?hl=en
- https://developers.google.com/search/blog/2017/12/nohacked-30-how-do-i-know-if-my-site-is?hl=en
- https://developers.google.com/search/blog/2017/12/nohacked-30-more-resources?hl=en
- https://developers.google.com/search/blog/2017/12/nohacked-30-tips-on-prevention?hl=en
- https://developers.google.com/search/blog/2018/03/strong-protection-against-hacking?hl=en
- https://developers.google.com/search/blog/2018/06/how-we-fought-webspam-webspam-report?hl=en
- https://developers.google.com/search/blog/2019/03/this-year-in-search-spam-webspam-report?hl=en
- https://developers.google.com/search/blog/2020/01/new-removals-report-in-search-console?hl=en
- https://developers.google.com/search/blog/2020/06/how-we-fought-search-spam-on-google?hl=en
- https://developers.google.com/search/blog/2020/07/how-spam-reports-are-used-at-google?hl=en
- https://developers.google.com/search/blog/2021/04/how-we-fought-search-spam-2020?hl=en
- https://developers.google.com/search/blog/2021/05/prevent-portions-of-site-from-spam?hl=en
- https://developers.google.com/search/blog/2021/05/working-with-you-for-safe-and-spam-free-search-2020?hl=en
- https://developers.google.com/search/blog/2021/07/link-tagging-and-link-spam-update?hl=en
- https://developers.google.com/search/blog/2022/04/webspam-report-2021?hl=en
- https://developers.google.com/search/blog/2022/12/december-22-link-spam-update?hl=en
- https://developers.google.com/search/blog/2023/04/webspam-report-2022?hl=en
- https://developers.google.com/search/blog/2023/10/october-2023-spam-update?hl=en
- https://developers.google.com/search/blog/2024/03/core-update-spam-policies?hl=en
- https://developers.google.com/search/blog/2024/11/site-reputation-abuse?hl=en
- GSC Blog: Core Updates 与算法迭代演化(2010-2026)
- GSC Blog: 质量政策与 E-E-A-T 演化(Webmaster Guidelines → Search Essentials)
- Google SEO Office Hours 精华(2022.11–2024.08)+ 3 大 FAQ 主题汇总