TRM Grew ChatGPT Referrals 8,337% in 90 Days. I Copied Their 4 LLMO Pillars Onto 3 Indie Sites. Only 1 Moved the Needle.

TRM 在 90 天内将 ChatGPT 引流提升了 8,337%。我将他们的 4 大 LLMO 支柱应用到了 3 个独立网站上，结果只有一个奏效。

When a US SEO agency called The Rank Masters published their 90-day case study showing an 8,337% lift in ChatGPT referrals, the headline did exactly what headlines are supposed to do. I clicked. Then I noticed the baseline was 8 visits and the post-period was 675. So yes, the percentage is technically true. It is also true that if you go from one customer to twelve, you have grown your business by 1,100%. 当一家名为 The Rank Masters 的美国 SEO 机构发布其 90 天案例研究，声称 ChatGPT 引流增长了 8,337% 时，这个标题起到了它应有的作用。我点击了进去。随后我注意到基准数据是 8 次访问，而期后数据是 675 次。所以，从技术上讲，这个百分比确实是真的。同样，如果你从一个客户增加到十二个，你的业务也增长了 1,100%。

What I actually cared about was the rest of the table. Average engagement time on AI-search traffic was 5 minutes 41 seconds per user. Page views per user climbed to 48. Those numbers are not a percentage trick. Those are people who showed up already interested and stayed. That is the part worth copying. 我真正关心的是表格的其他部分。AI 搜索流量的平均参与时间为每位用户 5 分 41 秒。每位用户的页面浏览量攀升至 48 次。这些数字并非百分比游戏，而是代表了那些带着兴趣而来并留下的用户。这才是值得效仿的部分。

TRM described their playbook as four pillars. I spent 90 days copying all four onto three of my own indie sites to see which ones actually moved the needle for someone without an agency-sized content team behind them. Three of the four were noise. The one that worked was the one I almost skipped. TRM 将他们的策略描述为四大支柱。我花了 90 天时间将这四根支柱应用到我自己的三个独立网站上，看看对于一个背后没有机构级内容团队的人来说，哪些真正能起到作用。结果四分之三都是无效的，而唯一奏效的那个，却是我差点跳过的。

The setup

准备工作

The three indie sites: 这三个独立网站分别是：

• kenimoto.dev — my engineering blog, around 50 articles at the start of the test, four-language stack (EN/JA/PT/ES), already had a llms.txt and JSON-LD on most pages.
• kenimoto.dev —— 我的工程博客，测试开始时约有 50 篇文章，采用四语言架构（英/日/葡/西），大多数页面已具备 llms.txt 和 JSON-LD。
• Site B — a 12-page niche tools site I built for a hobby project, almost zero schema, no author bio.
• 站点 B —— 一个我为业余项目构建的 12 页面利基工具网站，几乎没有 Schema 标记，也没有作者简介。
• Site C — a one-page indie SaaS landing page that ranks for a long-tail keyword, no blog, no schema.
• 站点 C —— 一个单页独立 SaaS 落地页，为一个长尾关键词排名，没有博客，也没有 Schema。

I picked sites at three very different stages on purpose. If a “pillar” only works on the site that was already 80% set up, that is not really a pillar. That is a finishing touch. 我特意选择了处于三个不同阶段的网站。如果一个“支柱”只在已经完成 80% 建设的网站上有效，那它就不是真正的支柱，而只是锦上添花。

Baseline period was 30 days before the test. Treatment period was the 90 days that followed. I used GA4 with the chatgpt.com and perplexity.ai referrer regex from llmoframework.com’s pillars guide, plus the four AI crawler user-agent filters in my server logs to confirm cross-channel pickup. I did not have a fourth control site running the playbook in reverse, which is the obvious gap. I am calling it before someone else does. 基准期为测试前的 30 天，处理期为随后的 90 天。我使用 GA4 并结合 llmoframework.com 支柱指南中的 chatgpt.com 和 perplexity.ai 引荐来源正则，加上服务器日志中的四个 AI 爬虫 User-Agent 过滤器，以确认跨渠道的流量获取。我没有设置第四个反向运行该策略的对照网站，这是一个明显的缺口，我先自己指出来，免得别人说。

The four pillars, as TRM described them

TRM 所描述的四大支柱

For anyone who has not read the original case study, here is the short version of what TRM ran: 对于还没读过原始案例研究的人，以下是 TRM 所执行策略的简要版本：

1. Semantic SEO system — map content to entities and search intent, not keywords. Build topical authority through related entity coverage. 语义 SEO 系统 —— 将内容映射到实体和搜索意图，而非关键词。通过相关实体覆盖建立主题权威性。
2. Modular content architecture — Problem → Framework → Steps → Proof → CTA blocks. Each block stands alone so LLMs can quote it. 模块化内容架构 —— 问题 → 框架 → 步骤 → 证明 → 行动号召 (CTA) 区块。每个区块独立存在，以便 LLM 可以引用。
3. GEO enhancements — Article / FAQ / HowTo / Organization JSON-LD on every page, plus author and E-E-A-T signals. GEO（生成式引擎优化）增强 —— 每个页面上的文章/常见问题/操作指南/组织 JSON-LD，以及作者和 E-E-A-T 信号。
4. Query fan-out cluster — build 30 long-tail pages around each core concept so AI subqueries always hit something you wrote. 查询发散集群 —— 围绕每个核心概念构建 30 个长尾页面，确保 AI 的子查询总能命中你写的内容。

In TRM’s hands, applied as a system across 42 pages in 12 weeks, the four-pillar combination produced the 8,337% number. I did not have 12 weeks and I did not have 42 pages of capacity. What I had was three sites and a fixed budget of evenings. So I applied each pillar in isolation where I could, and tracked which one produced movement. 在 TRM 手中，作为一套系统在 12 周内应用于 42 个页面，这四大支柱的组合产生了 8,337% 的增长。我没有 12 周的时间，也没有 42 个页面的容量。我只有三个网站和固定的晚间时间。因此，我尽可能地孤立应用每个支柱，并追踪哪一个产生了效果。

Pillar 1: Semantic SEO. Flat line on all three sites.

支柱 1：语义 SEO。三个网站均无起色。

I spent the first three weeks rebuilding internal links on kenimoto.dev around entity clusters. Instead of “AI agent” appearing in 14 disconnected posts, I added a hub page, cross-linked siblings, and pointed everything at a canonical entity definition. On Site B I rewrote four pages around their parent entity. Site C got one new “what is X” sibling page. 我花了前三周时间围绕实体集群重建了 kenimoto.dev 的内部链接。不再让“AI agent”出现在 14 篇互不关联的文章中，我增加了一个中心页面，将相关页面交叉链接，并将所有内容指向一个规范的实体定义。在站点 B 上，我围绕其父实体重写了四个页面。站点 C 增加了一个新的“什么是 X”相关页面。

After 90 days, ChatGPT referrals to kenimoto.dev went from 18/month to 23/month. Site B moved from 0 to 2. Site C did not move. That is not zero, but it is also not a pillar. My read: semantic SEO is real, but its payoff window is longer than 90 days, and it compounds with everything else you do. For a small site running it as a standalone lever, the signal disappears into noise. 90 天后，kenimoto.dev 的 ChatGPT 引流从每月 18 次增加到 23 次。站点 B 从 0 增加到 2。站点 C 没有变化。这虽然不是零，但也称不上支柱。我的解读是：语义 SEO 是有效的，但其回报周期超过 90 天，且它会与你所做的其他一切产生复合效应。对于一个将其作为单一杠杆的小型网站来说，信号会淹没在噪音中。

Pillar 2: Modular content. Best for the writer, worst for the test.

支柱 2：模块化内容。对作者最友好，对测试最无效。

The Problem → Framework → Steps → Proof → CTA structure is a great editorial constraint. I rewrote eight existing kenimoto.dev posts to fit it. The articles read better. They are easier to skim. I am personally happier with them. The traffic data did not notice. “问题 → 框架 → 步骤 → 证明 → CTA”结构是一种极好的编辑约束。我重写了 kenimoto.dev 现有的八篇文章以符合该结构。文章读起来更顺畅，更易于浏览。我个人对它们更满意，但流量数据并没有察觉到变化。

ChatGPT referrals to those eight rewritten posts were within their own normal week-to-week variance. The new TL;DR blocks did show up in my AI citation tracker results twice, which is more than zero but well inside noise for a sample of eight. 那八篇重写文章的 ChatGPT 引流仍在正常的周度波动范围内。新的“TL;DR”区块确实在我的 AI 引用追踪结果中出现了两次，这虽然不是零，但在八个样本中完全属于噪音范围。

Pillar 3: Author schema and E-E-A-T. The one that worked.

支柱 3：作者 Schema 和 E-E-A-T。唯一奏效的支柱。

This is the part I almost skipped. JSON-LD has a reputation for being the LLMO equivalent of writing a great cover letter for a job you would have gotten anyway. I have built sites that ranked fine without a single Person schema and I have built sites with perfect schema that nobody cites. 这是我差点跳过的部分。JSON-LD 在 LLMO 领域名声不佳，被认为是那种“即使不写也能得到工作，但写了也只是锦上添花”的求职信。我曾构建过没有任何 Person Schema 也能获得良好排名的网站，也曾构建过拥有完美 Schema 却无人引用的网站。

Three things changed: 有三点发生了变化：

1. I added a Person schema to every author byline on kenimoto.dev, with sameAs pointing to GitHub, X, LinkedIn, and a verified Zenn profile. 我在 kenimoto.dev 的每个作者署名处添加了 Person Schema，并将 sameAs 指向 GitHub、X、LinkedIn 和经过验证的 Zenn 个人资料。
2. I wrote a real /about page with Person + ProfilePage schema, including credentials and a list of published books and articles with WorkExample links. 我写了一个真正的 /about 页面，并添加了 Person + ProfilePage Schema，包括资历以及带有 WorkExample 链接的已出版书籍和文章列表。
3. I added author and publisher fields to the existing Article schema on every post. 我在每篇文章现有的 Article Schema 中添加了作者和发布者字段。

Total time: about six hours. No content was added, no posts were rewritten. Result over 90 days on kenimoto.dev: ChatGPT referrals: 18/month → 127/month. Perplexity referrals: 4/month → 41/month. 总耗时：约六小时。没有增加内容，也没有重写文章。kenimoto.dev 在 90 天后的结果：ChatGPT 引流从每月 18 次增加到 127 次；Perplexity 引流从每月 4 次增加到 41 次。