不，SEO 并没有死：真正改变的是什么
The short version: SEO is not dead, it changed jobs. Zero-click search passed two-thirds of queries in early 2026, and AI Overviews cut organic clicks on the searches where they appear by roughly half. But every one of those answers is still pulled from a traditional index. Indexation, authority, relevance, and structure now decide who gets named, not just who ranks.
Did AI search actually kill SEO?
No, and the data says something more specific than "no." What died was the assumption that a top ranking was the whole game. Zero-click searches, the ones that end without a click to any site, reached 68.01 percent of US Google queries in early 2026, up from 60.45 percent in 2024 (SparkToro, to re-verify). Where an AI Overview appears, the effect compounds. Pew Research Center tracked real browsing data from 900 US adults and found people clicked a traditional result in just 8 percent of searches with an AI summary present, against 15 percent without one (Pew Research Center).
Read that as proof the discipline died, and it reads backwards. Spend tells the real story: the global SEO services market is estimated at 83.98 billion US dollars in 2026, still growing toward a projected 148.86 billion by 2031 (Mordor Intelligence, to re-verify). Nobody pours money into a dead channel. What changed is what that money now has to buy.
8%
of searches with an AI summary present still saw a traditional-result click, versus 15% without one
What genuinely changed
Three shifts are real, worth naming plainly instead of waving at "AI search" as one blob.
| What changed | What it means in practice |
|---|---|
| Zero-click growth | More answers get resolved on the results page itself, so fewer visits go to any single site, yours included. |
| AI Overviews compress clicks | Where an Overview appears, the traditional result underneath it earns roughly half the clicks it used to. |
| New answer engines, new rules | ChatGPT leans on Bing, Perplexity favors fresh and community sources, Copilot surfaces its own citations. Each reads pages slightly differently. |
None of these three retire indexation, authority, relevance, or structure. They add a second reader, one that composes an answer and names a source instead of clicking a link.
Why the fundamentals matter more, not less
Here is the detail the myth skips past. Google states plainly that AI Overviews and AI Mode draw from the same index and ranking systems as ordinary Search (Google). There is no separate AI index to game. A page that is not indexed, not crawlable, or thin on relevance was never found by the old system, and it is not found by the new one either. The floor did not move. If anything, it got stricter, since two audiences now judge the same page instead of one.
If ranking does not guarantee a citation, does it still matter?
Yes, and the evidence is specific, not vibes. Ahrefs analyzed a million keywords that triggered AI Overviews, pulled 1.9 million citation links, and calculated a 0.347 positive correlation between ranking in the top 10 and being cited among the top 3 AI Overview links (Ahrefs, to re-verify). That is real and moderate, not a coin flip and not a guarantee: even the number one result lands in those top citations only about half the time. A separate Ahrefs study of 863,000 keywords and 4 million AI Overview citations found 38 percent of cited pages also sat in the traditional top 10 (Ahrefs, to re-verify). Ranking well no longer guarantees a citation. Not being indexed at all still comes close to guaranteeing you get skipped.
What the myth gets backwards
The myth treats "the job looks different" as proof "the job is gone." It is the opposite. Every mechanism that used to earn a ranking, being indexed, relevant, structured, and backed by real authority, is the same mechanism that earns a citation today. What is new is not a replacement discipline. It is one requirement added on top: write the page so a machine can lift the answer cleanly, not only so a person can scan it. Skip the fundamentals and no clever formatting saves the page. Nail the fundamentals and skip the formatting, and a competitor who did both still outruns you.
The test that settles the argument
Two questions decide it, on any page.
- Can a search engine actually reach this page and understand it?
- Can a machine quote this page's answer without guessing at the rest?
Answer yes to both, and SEO is doing exactly what it always did, just for a second reader now too. Answer no to either one, and the problem is not that SEO died. The fundamentals were never finished.
常见问题
没有。根据谷歌自己的文档说明，AI 概览和 AI 模式运行在与普通谷歌搜索相同的索引和排名系统之上。一个无法被收录或排名的页面，在这两套系统中本来就不会被看见。AI 概览改变的是答案的呈现方式，而不是让页面可被发现的根基。
因为越来越多的问题，直接在结果页上就通过精选摘要、信息面板，如今还有 AI 摘要得到了解答。零点击率在 2026 年初达到美国谷歌搜索总量的 68.01%，高于 2024 年的 60.45%，这意味着更少的查询会最终落到某个具体网站的访问上。
重要，但已经不能保证被引用。Ahrefs 发现，进入前 10 名的排名与被 AI 概览前 3 条链接引用之间，存在中等程度的正相关（相关系数 0.347），同时 38% 的 AI 概览引用，来自同样进入前 10 名的页面。排名能提高被引用的概率，仅此而已。
不会。原创内容、真实数据和清晰的答案，正是答案引擎寻找并引用的对象，因为一手信息源对它们而言是最安全的引用来源。真正会消亡的，是那些只是复述共识的内容，而这类内容在 AI 搜索出现之前就已经很弱了。
SEO 赢得的是排名位置，AEO（答案引擎优化）赢得的，是在生成式答案中被引用的机会。两者共享同一个基础，也就是可被收录、相关且具有权威性，但 AEO 在此之上增加了一层格式要求：写出独立成立的答案，以及有明确出处的事实，让机器无需猜测就能直接采用。
市场给出的答案是肯定的。全球 SEO 服务市场规模，在 2026 年估计接近 840 亿美元，预计还将持续增长到 2031 年。企业是在扩大自然流量投入的覆盖范围，而不是放弃它。
因为这些系统抓取的范围比前 10 名要广得多，包括排名在 100 名开外的页面，以及像 YouTube 这类拥有自身权威信号的信息源。排名只是众多输入因素之一，而不是唯一的门槛。
不能。结构化数据能帮助机器正确理解一个页面，但它无法凭空制造出原本不存在的权威或相关性。它是在转译优质内容，而不是替代优质内容。
两者都有。在出现 AI 概览的地方，某个页面获得的点击会下降，但真正到达的流量往往质量更高，因为访客是循着机器给出的答案而来，而不是随手扫过一排蓝色链接。这个渠道并没有消失，只是它的经济效益发生了转移。
不要再把“获得排名”当作终点。保留每一项曾经赢得排名的基本功，再加上一个新习惯：把答案写得让机器无需猜测就能直接引用。这样的组合依然能被找到，现在还能被引用。