让您的企业参与AI对话（AEO策略指南）

欢迎来到《早期采用者》，这是一档帮助你将AI转化为竞争优势的播客。我是杰西，白天是营销主管，白天、黑夜以及几乎所有清醒时刻都在进行AI实验。

Welcome to Early Adopter, the podcast where we help make AI your unfair advantage. I'm Jess, marketing exec by day, AI experimenter by day, night, and, basically every other waking moment in between at this point.

我是凯尔，你们的技术导师。过去十年我一直在构建解决实际问题的数据系统、产品和AI工具。每周我们会解析那些真正对你业务重要的AI概念、工具或趋势。

And I'm Kyle, your resident tech guru. I've spent the last decade building data systems, products, and AI tools that solve real world problems. Each week, we'll break down AI concepts, tools, or trends that actually matter for your business.

无论你是小企业主、创业者，还是单纯对科技好奇，让我们开始探索吧。就把我们当作带你穿越AI这个既奇妙又时而令人不知所措世界的向导。

So whether you're a small business owner, an entrepreneur, or you're just tech curious, let's dive in. Just think of us as your guides through the wild, wonderful, and sometimes completely overwhelming world of AI.

这里是《早期采用者》。

This is Early Adopter.

欢迎回到《早期采用者》，这档帮助你将AI转化为竞争优势的播客。我是杰西。

Welcome back to Early Adopter, the podcast where we help make AI your unfair advantage. I'm Jess.

我是凯尔。那么杰西，这周过得怎么样？

And I'm Kyle. So, Jess, how's your week been?

有点平静，甚至无聊，没什么特别的事。今天早上我去里士满散了步——如果有人看过《足球教练》的话，里士满是伦敦西南一个真实社区，非常美丽如画又可爱。我和朋友沿着河边散步。

It it's been a little quiet and a a little boring. Not not a whole lot to report. I went for a lovely walk in Richmond this morning. So if anyone, ever watches Ted Lasso, Richmond is a real neighborhood in Southwest London, and it is absolutely beautiful, very picturesque and cutesy. So I went and met a friend, and we walked along the river.

这可能是我这周做过的最有趣的事了。

And that was that was probably the most interesting thing I've done this week.

听起来挺不错的。

I mean, that sounds lovely.

确实。

It was.

我完全迷上了《足球教练》。刚才那段对话让我有点小感动。也许我该找个时间重温一下。哦，你应该看看的。

I'm I'm absolutely obsessed with Ted Lasso. So, yeah, that just got a little ping in the feels. Maybe I should do a little rewatch soon. Oh, you should.

嗯，新一季就要开播了。这部剧真的太治愈了。你知道吗？其实我花了很长时间才入坑。我试了三四次才真正看进去，不知道具体为什么，起初我觉得它有点过于甜腻了。

Well, the new season's coming out. The most wholesome show. Do you know? It actually took me quite a long time to get into that show. I tried it about three or four times before I really got into it, And I don't know what it was about it that just found a little I think I found it a little too sickly sweet.

就像我们这边常说的，有点矫揉造作。

A little too as we say over here, it's a little twee.

我觉得这就是你们英国人特有的气质。你们暂时还接受不了那种纯粹的治愈感，需要花点时间慢慢融化心里的坚冰。

The I think that's that Britishness that you've occupied. You just can't take the wholesome just yet. It takes a little while to, like, you know, break the break the ice off of the heart.

我是说，确实如此。基本上吧。最终我还是入坑了，而且非常喜欢。不过确实，我试了大概三四次才真正看进去

I mean, event yeah. Pretty much. Eventually, I got into it, I really loved it. But it yeah. It took me I I had to try about three or four times to watch it before I actually got into

这部剧。关键是你坚持下来了。我确实坚持了。因为这是有史以来最棒的剧集之一

it. So important thing is you stuck with it. I did. I did. Because one of the best shows of all time.

嗯，因为大家都说他们有多喜欢这部剧。而且我也超爱布雷特·戈德斯坦，他太搞笑了。简直绝了，我知道

Well, because I just everyone kept saying how much they loved it. And also I love Brett Goldstein. He's hilarious. Spectacular. I know.

随着剧情发展，罗伊这个角色的戏份会越来越多。别以为

He gets, he gets a lot more Roy as the show goes on. Don't think

他没戏份。他绝对有。而且我朋友的狗也叫罗伊，不过不是因为这个角色，是致敬足球运动员罗伊·基恩，

he does that. He He definitely does. Also my friend's dog is called Roy. Not after that, after the footballer Roy Keane, who

我觉得他应该就是

is the, I think, was the inspiration for that

这个角色的灵感来源。没错，就是这样。不过这只狗是你见过最小最毛茸茸的吉娃娃博美混种，所以这么个小毛球叫罗伊特别搞笑。话说你这周过得怎么样？

was character. Exactly So, yeah, exactly. But also this dog is like the tiniest, fluffiest Chihuahua Pomeranian thing you've ever seen. So it's hilarious that, this little fluffy dog is, is called Roy. Anyway, how was your week?

是啊，我这周过得非常棒。实际上我见到了我们早期几期节目中的肖恩，我们在现实生活中见面了。哇。

Yeah. My week was great. I actually got to see Sean from some of our earlier episodes. We, we met IRL in real life. Wow.

我们去参加了一个Databricks会议，了解数据领域的最新动态，虽然可能大多数人不会想花时间在这种地方，但作为一个数据极客，

We went to a Databricks conference to see the latest on data, which is, you know, probably not where most people wanna spend their time, but, you know, as a data geek myself,

那是极客的聚会。

it was Nerds.

极客们。有意思。我们参加了一些专题讨论，还参与了一些研讨会。我觉得企业数据社区仍然非常专注于数据专业人士，而不是那些创业者或运营者，我们该如何让数据更普及、更易用呢？

Nerds. Interesting. We we checked out some panels. We were in some workshops. I would say the enterprise data community is still very focused on data people and not necessarily like, you know That's does this work for, you know, the entrepreneur operator, you know, how do we democratize and make this easier?

内容非常非常偏重数据。所以。

It was very, very data heavy. So.

是啊，我去过那种会议，对这些东西还算比较了解。虽然不会深入研究，但能理解。有一次我参加了一个讨论会，坐了大概五分钟就决定离开了。不过在我们开始之前，凯尔，我前几天意识到一件事。

Yeah. When I've been to those sorts of conferences, I have a relatively good understanding of these things. Like I don't get in the weeds with it, but I understand it. And I rang in one panel and I sat there for about five minutes and I was like, nope. Well, before we get started, Kyle, I realized something the other day.

这是我们第二十期节目了。哇哦。庆祝我们的二十期快乐。都快到合法饮酒年龄了。

This is our twentieth episode. Woo. Happy happy twentieth to us. Almost legal to drink.

我是说，20岁喝酒还不合法。如果按年份算的话。

I I mean, 20 old is not legal to drink. Guess if we're taking it in years.

有道理。不过在法国肯定合法。就像狗年算法一样。

Fair enough. Sure it is in France. It's like dog years.

这就对了。嗯，真令人兴奋。我们都

There we go. Well, that's exciting. We're all

长大了。确实令人兴奋。我们都快，快到了。我们的前额叶还没完全发育好呢。

grown That is exciting. We're all almost, almost there. Our frontal lobes haven't fully developed yet.

我自己也还在等那一天。

I'm still waiting on that myself.

没错。总之，我特意提这个是因为我们第十期时确实没提到。所以想确保第二十期时强调下。恭喜我们坚持下来了。现在进入本周主题，我们讨论这个已经有一阵子了，终于要正式探讨它了。

Yeah, exactly. Anyway, I just wanted to flag that because we didn't actually mention that when we passed our tenth episode. So I wanted to make sure that we flagged it for our twentieth. So congratulations us, we stuck with it. So moving on to our topic for this week, we have been talking about this for a while and it's finally time for us to tackle it.

这个概念叫答案引擎优化（AEO）。也叫AIO、GSO、GEO、对话式搜索、情境搜索。名称很多。本播客统一称AEO。这将是个重要话题，最近几周讨论热度明显上升。

And this is the concept of Answer Engine Optimization, AEO. It's also known as AIO, GSO, GEO, conversational search, contextual search. It's got a lot of different names. For the sake of this podcast, we're calling it AEO. It's gonna be quite an important topic and there's been a lot more kind of buzz around it in the last couple of weeks.

这就是我们想讨论这件事的原因。

So that's why we wanted to talk about it.

我记得你第一次跟我提到AEO是在...大概是九个月前，也可能是六个月前？那时候我们好像还叫它GEO。AEO这个叫法大概是一个月前才有的吧，我也不太确定。

I think when you first mentioned AEO to me back in, I don't know, nine months ago, six months ago well, think we were calling it GEO then. AEO, I think was like maybe a month or so ago. I'm not sure.

变化太快了。

It changes so quickly.

变化确实很大，所以我之前还以为这只是又一个营销噱头。但经过深入研究，特别是这周钻研技术层面后，发现这确实从根本上改变了搜索的运作方式。

It changes a lot, which is why I kinda thought it was just another marketing buzz work. Right? But the more we've looked into it, especially this week, dove into the technical side, this is really fundamentally changing how search works.

没错。而且很多人现在还没意识到这点。HubSpot最近的报告显示，现在60%的搜索行为最终都没有点击结果。

Yeah. Absolutely. And, you know, a lot of people still haven't quite caught on to this yet. There are some reports that are showing that 60% of searches now end without a click. That's from a recent report that HubSpot did.

要知道HubSpot是大型CRM和销售平台，这正是他们的专业领域。这本质上意味着AI引擎正在成为新的谷歌。这也是为什么谷歌现在有点慌。但说实话，很多人还没听说过这个概念。

And if you know HubSpot is a big CRM platform, sales platform, and this is exactly in their wheelhouse. Essentially it means that AI engines are becoming the new Google. This is one of the reasons why Google is kind of in panic mode at the minute. But honestly, a lot of people still haven't heard of this concept.

是的，问题之一就是我们还没形成统一的术语体系。不过说真的，如果谷歌代表搜索，那么ChatGPT就是AEO（答案引擎）之类的。现在大家不都开玩笑说'我去问问Chat'嘛。

Yeah. I think that that is one of the problems is we haven't come up with a proper nomenclature Yeah. Absolutely. But, yeah, if search you know, Google is search, then, ChatGPT is AEO or answer engine or whatever have you. And let's be honest, like, most we all joke now about, oh, let me go ask chat.

但实际上，人们就在那里

But actually, that's where people are

这正是我们要探讨的。我认为我们会多谈谈消费者行为的变化，因为这确实是其中非常重要的一部分。所以这实际上会是一个两部分的系列。本周，我们将大致概述它是什么以及为什么重要。然后下周，我们将深入探讨实用技巧和要点，因为这又是一个大话题，内容很多。

going That's to find exactly it. And I think we'll talk a little bit more about the consumer behavior changes because that is a really, really big part of this. So this is actually going to be a two part series. This week, we're going to kind of give a little bit of an overview about what it is and why it matters. And then next week, we're going to deep dive into the practical tips and takeaways, because again, it's a big topic and there's a lot to cover.

所以我们想确保我们对每个部分都给予适当的关注。

So we wanna make sure that we're giving everything the the proper attention.

是的。因为，如果你认为传统SEO对企业来说是改变游戏规则的，我的意思是，我们已经在其中积累了二十、二十五年的经验教训，而企业仍在努力理解。现在，我们有了AEO，这是SEO的进化步骤。我们需要多久才能掌握它呢？

Yeah. Because well, because if you think traditional SEO was a game changer for businesses, I mean, we've been on that now for twenty, twenty five years of lessons and learning and, you know, businesses are still trying to figure out. So, now we have AEO. It's an evolutionary step on SEO. How long is it gonna take us to figure that out?

所以，我们确实需要成为早期行动的一部分，或者说是早期采用者。

So, we definitely need to be part of kind of that early movement or early adopters, if you will.

看你做了什么。

See what you did there.

我知道。我尽力了。而且，你看，如果你能早点弄明白，这绝对是一个巨大的优势。

I know. I try. And, you know, look, it's it can be an absolutely massive advantage if you figure this out early.

特别是因为很多人还在努力理解这个概念。你刚才关于SEO的讨论实际上是个很好的过渡，因为我认为我们应该先简单回顾一下SEO，毕竟AEO（答案引擎优化）是建立在SEO基础之上的。提醒一下，传统SEO（搜索引擎优化）的核心是让网站在搜索结果中更显眼。过去二十年来，这主要围绕关键词研究展开——你根据这些关键词创建内容。

Especially because a lot of people are still kind of wrapping their heads around this. So that was actually a nice segue with your little talk about SEO, because I think let's do a little bit of an SEO refresher because AEO builds on those foundations. So as a reminder, traditional SEO, which is search engine optimization, is all about making your website more visible in search engine results. So for the past twenty years, it's been about keyword research. You create content around those keywords.

你从其他网站建立反向链接并进行技术优化。所有这些都有助于谷歌理解和排名你的网页。目标相当明确：尽可能在相关搜索词中获得更高排名，从而让人们点击进入你的网站。这相当直接。

You build backlinks from other sites and technical optimizations. And all of that helps Google understand and rank your pages. So the goal is pretty straightforward. It's to rank as high as possible for relevant search terms so that people click through to your website. It's pretty straightforward.

但答案引擎优化（AEO）的核心本质截然不同，尽管它建立在SEO基础之上。你不是优化内容以出现在搜索结果列表中，而是优化内容以出现在AI生成的回答里。想想很多人见过却有点讨厌的谷歌AI概览功能——前几天还有人问我怎么关闭它。

But answer engine optimization, AEO, is kind of at its very core. It's quite different, although it builds on those SEO foundations. So instead of optimizing to show up in a list of search results, you're optimizing your content to show up in AI generated responses. So, you know, think of the AI overview that a lot of people have seen and actually kind of hate in Google. I had someone ask me how to switch it off the other day.

或是ChatGPT、Perplexity这类平台。你现在不再追求传统搜索结果的第一名，而是努力成为AI引擎在回答问题时引用的来源。这是完全不同的策略。

Or ChatGPT or Perplexity, like all of those sorts of things. So instead of trying to rank number one in traditional search results, you're now trying to become the source that AI engines quote when someone asks a question. So it is quite a different approach.

正如你提到的传统SEO，关键词、内链建设这些手段能让你看到排名上升。所有努力的目标就是随着排名提升获得更多点击量，因为人们不会滚动太远——他们通常只会点击前两三个结果。

So where traditional SEO was, as you mentioned, you know, keywords, backlinking, you know, between your different pages and you could kind of see yourself moving up in those rankings. All of the goal of that was as you get higher in the rankings, you get more clicks because people didn't scroll down super far. Right? They Yeah. They would click the top two or three.

没错。AEO彻底颠覆了这个逻辑——重点在于被引用而非排名。SEO聚焦关键词，AEO则专注于直接回答问题。

Yep. AEO now flips that. It's about getting cited, not about ranking. So SEO focused on keywords. AEO focuses on answering questions directly.

SEO要求你赢得排名游戏，而AEO是要成为ChatGPT、Claude甚至谷歌AI概览等引擎在回答问题时选择引用的来源。所以，与其指望在谷歌关键词搜索中获得顶部那个蓝色第一链接，你更希望自己的内容被AI调用的众多来源积极提及。

And where SEO required you to win that ranking game, AEO is about becoming the source that these AI engines like ChatGPT or Claude or even Google's AI overview choose to cite when people ask questions. So, again, instead of hoping to win that single number one blue link at the top of the Google keyword, you want your content mentioned positively across many, many sources that the AI pulls from.

确实如此。几周前我们和Eric聊过我最喜欢的谷歌披萨胶水故事。这正是因为那个人问的问题——如何让奶酪粘在披萨上？关于这方面的信息非常少。所以谷歌只能从唯一能找到的来源Reddit获取答案。

Exactly. And I I talked a couple weeks ago when we spoke to Eric about my favorite Google pizza glue story. And this is exactly it because the question that the person asked, which was how do I get cheese to stick to pizza? There is very little information about that. So Google pulled from the only source that it could find, which was Reddit.

这就是为什么最终会出现那个荒谬的回复。本质上这就是核心理念——你正试图成为那个单方面的信息来源。这意味着人们直接在AI概览中获取答案。我们正在见证人们搜索方式、沟通方式和信息获取方式发生的巨大变革。

And that's why you ended up with that ridiculous response. So that is essentially the concept. You're just trying to become that one-sided source. And it means that people are getting their answers right in the AI overview. So we're really really looking at a massive, massive shift in how people are searching and how people are communicating and finding out information.

是啊。我觉得，如果你还在试图为网站引流，这种情况可能有点可怕。因为我们讨论过60%的搜索不会导致用户点击蓝色链接访问你的网站。如果得不到流量，你该怎么办？

Yeah. I think that, I mean, that could be potentially kind of terrifying if you're if you're still trying to drive website traffic because we we talk about this 60% of searches don't result in a click to your website from that blue link. So if you're not getting traffic, what do you do?

完全同意。有一期播客我强烈推荐大家收听，如果想了解更多相关信息。我们肯定会把链接放在节目备注里。这期播客叫Lenny's Pod，主持人采访了一位叫Ethan Smith的嘉宾，他是网站构建平台Graphite的CEO，堪称这个领域的权威。他们与Webflow合作发现，来自AI引擎的流量转化率是普通谷歌流量的六倍。

Yeah, absolutely. And there is a podcast episode that I highly, highly recommend anyone listen to if they want a little bit more information about this. We'll definitely put this in the show notes, but, there is a podcast called Lenny's Pod, and he does an incredibly useful interview with a guy called Ethan Smith, who is the CEO of a company called Graphite. And he is really like the authority in this space, but they work with Webflow, which is a website builder. And they've been finding that the traffic from AI engines was converting at six times the rate of normal Google traffic.

所以效果要好得多得多。稍后我们会讨论为什么能有这样的效果。

So it's much, much, much more effective. And we'll talk a little bit about why that's effective later.

是啊，六倍这个数字太惊人了。这个统计数字让我震惊。即使是2倍或3倍都值得关注，但6倍显然意味着根本性的转变。

Yeah. That six times is incredible. I mean, I was kind of blown away by that stat. Even 2X or 3X would be something to take notice of. But 6X is obviously like a fundamental shift.

没错。但仔细想想用户旅程就很合理——如果有人问ChatGPT'最适合营销机构、能与HubSpot集成且具备高级报表功能的CRM是什么'，而AI推荐了你的产品，说明这个人已经处于转化漏斗更下游阶段。他们寻找的是非常非常具体的解决方案，所以如果你提供这些服务，他们的购买转化率自然会高得多。这不同于只是谷歌搜索'最佳CRM平台'，他们已经有了明确需求，而你的产品正好满足所有条件。本质上他们已经与理解其具体需求的AI进行了咨询，答案就像被端上银盘一样呈现给他们。

Yeah. But it actually really makes sense if you think about someone's user journey, because if someone's asking ChatGPT what's the best CRM for a marketing agency that integrates with HubSpot and has advanced reporting and it recommends your specific product, that person is much further down the funnel. So they're much more likely to convert because they're looking for something that is really, really, really specific. So if you are providing those services, then it makes sense that someone's going to be much more likely to convert to purchase because they're already, rather than just Googling, you know, best CRM platforms, they're already looking for something specific and you tick all of those boxes. So they've essentially had a consultation with an AI that understands their really specific needs and kind of serves it up on a platter.

是的，无论好坏。对吧？我们几乎无条件信任这些AI工具。

Yeah. For for better or worse. Right? We trust these AI tools almost implicitly.

我们

We

确实如此。我们确实如此。要知道，一旦你接触到号称最好的CRM系统，你就具备了资格。你就像进入了漏斗流程，准备就绪。我们经常讨论信任问题，稍后会深入探讨。

do. We do. Know, you're as soon as you get, hey, this is the best CRM, you're qualified. You're like in the funnel. You're ready to go, which, you know, we talk about trust issues all the time and we'll get more to that.

但是

But

好的。那么从实际技术角度来看，这些AI引擎是如何决定引用什么内容的？技术层面具体是如何运作的？

Yeah. So when it comes to the actual technical perspective, how are these AI engines actually deciding what to quote? Like, how how does the technical side work?

当你向ChatGPT或其他大语言模型提问时，它通常会通过检索层获取相关信息。比如你看到它显示'正在搜索网络'——这就是它们添加的检索层功能。这是过去几年的重大改进，使其能引入相关信息，不再受限于我们之前讨论的截止日期限制。检索可能通过谷歌搜索、自有爬虫、必应搜索等多种方式实现。

So when you ask ChatGPT a question or any other LLM, it often uses a retrieval layer to pull relevant information. You know, when you see it on the little thing, say, you know, searching the web, this is a retrieval layer that they've added. It's a big improvement over the last couple years so that it starts to bring in relevant information. It doesn't have that cutoff date like we've talked about. So this could be a web search via Google or its own crawlers or Bing search, a lot of different ways to it.

信息可能来自维基百科等知识库或索引——这是个非常权威的相关资料来源。也可能通过自有浏览器模式（如ChatGPT的浏览模式）连接获取。检索到相关信息后，它会综合这些来源生成答案，但不会单一采用最高排名结果，而是融合多个可信来源的信息进行交叉引用。

It could come from knowledge sources or indexes like Wikipedia, which is a really strong, you know, relevant authoritative source. Or it could connect, you know, to a browser like its own ChatGPT browsing mode. So once it's then retrieved the relevant information, it synthesizes an answer from those retrieved sources. But it's not just the top one, but a blend of the information retrieved. So it'll often quote across several trusted sources.

所以这基本上意味着ChatGPT会自行进行谷歌搜索，对吧？

So it essentially means that ChatGPT is going away and doing its own Google search. Right?

是的，完全正确。

Yeah. Exactly.

但如果它本质上是在自行处理，那么排名还重要吗？

But so does ranking still matter then if it's essentially going away and doing its own?

是的，我认为这正是最有趣的地方。或许会让那些花了二十年学习SEO的人松一口气——排名仍然重要，但只在某种程度上。AEO将建立在SEO的基础上，但它会添加一些独特的层次，或者说是一个非常重要的新步骤。我的意思是，如果Bing或谷歌的排名被LLM用来获取信息来源，排名更高的内容更有可能进入候选信息池。

Yeah. I think this is where it gets really interesting. And perhaps to the relief of everyone who spent, you know, the last twenty years learning SEO, yes, ranking still matters, but only to a certain extent. So AEO is gonna stand on the shoulders of SEO, but it really starts to add some unique layers to it or kind of like its own really important step. And so what I mean by that is, you know, if Bing or Google rankings are used by the LLM to fetch sources, higher ranked content is more likely to be in that candidate pool of information to draw from.

但许多检索系统也严重依赖维基百科和主要出版商作为排名信号。这些来源提供了权威性、时效性和相关性，对LLM来说都非常重要，因为这就是大量上下文信息的来源。

But, you know, many retrieval systems rely heavily on Wikipedia and major publishers for ranking signals as well. And, you know, those give it like authority, recency, and relevance, all very important as we talk about for an LLM because, you know, again, that's a lot of the context.

我认为这也是许多出版商合作关系开始发挥作用的地方。完全同意。如果他们与《纽约时报》等有合作协议，就会从权威来源获取信息，这是好事，因为他们也会大量引用像Reddit这样的平台，而我们知道那里的信息有时可能不太可靠。

I think this is also where a lot of these publisher partnerships start to come into play as well. Totally. Because if they've got a partnership deal with the New York Times or something, then they're going be pulling from an authority, which is a good thing because they also will tend to pull significantly from something like Reddit, where as we know, the information can sometimes be a little iffy, iffy sometimes. Yeah.

这些内容未经出版或引用，而那些主流出版商的内容则是经过审核的。

It's not published, it's not cited, whereas these, you know, major publishers are.

没错，正是如此。披萨胶。我就说这么多，披萨胶。

Yep. Exactly. Pizza glue. That's all I'm gonna say, pizza glue.

确实。所以

Exactly. So

简单总结一下，排名在AEO中仍然重要，但方式略有不同。你可以把它想象成一张入场券。如果你的页面根本排不上名次，它可能永远不会进入Chatchippity或Claude调用的检索池。但一旦进入池中，AI并不在乎你是第一名还是第十名，它只关心哪些内容清晰、真实、易于引用。所以SEO是让你进入房间的门票，而AEO则是让AI真正点你名的关键。

to sum that up quickly, ranking does still matter in AEO, but in a slightly different way. So you can kind of think of it as, as like an entry ticket. So if your page doesn't rank at all, it might never get into the retrieval pool that Chatchippity or that Claude pulls from. But once you're in the pool, the AI doesn't care if you're at number one or number 10, it just cares about which content is clear, factual, easy to quote. So the SEO kind of gets you into the room, but AEO is what makes the AI actually call your name.

所以SEO绝对仍然重要。你必须在SEO上下功夫，但正如你所说，AEO在某种程度上又增加了一层。

So SEO absolutely does still matter. You have to put the effort in with your SEO, but AEO, as you said, kind of adds another layer to it.

是的。我喜欢那个房间的比喻，因为它太贴切了，对吧？它能让你进入房间，但你知道，你仍然需要保持相关性才能被点名。换种方式对比，SEO通过帮助搜索引擎抓取和索引你的网站来工作，然后通过关键词、网站结构等我们讨论过的内容来传递相关性和权威性信号。搜索引擎的排名算法权衡这些信号来决定你的页面在一到十的搜索结果中的位置。而相比之下，要在AI答案中被引用，你需要清晰、基于事实的内容，并以非常易懂的问答形式结构化，这样数据的含义就非常明确。

Yeah. I love that room analogy because it's perfect, right? It gets you in the room, but, you know, you still need to be relevant to get called out. So to kind of contrast this in another way, SEO works by helping the search engines crawl and index your site, and then signaling that relevance and authority through things like the keywords and the site structure and all the things we talked about. The search engine's ranking algorithm weighs those signals to decide where your page shows up in those one to 10 results versus if we compare that to AEO, you know, to be cited in AI Answers, you need clear, fact based content structured in a really understandable question and answer form, so it's like really clear what the data is.

明白吗？然后它还需要得到可信来源的一致提及支持，不仅限于你的网站，还包括许多其他地方，以及那些让你的答案真正易于解析的信号。

You know? And then it needs to be supported by consistent mentions across trusted sources, not just your website, but lots of other places as well, and signals like, you know, things that make your answers really easy to parse.

你是指从数据角度解析，对吧？

You mean parse in terms of a data perspective. Right?

抱歉。对。让我这么说吧。基本上，它让答案变得非常易于消化，或者说，你知道，便于拆分理解。

Sorry. Yeah. Let me say that. Basically, it makes it makes the answers really easy to digest or to like, you know, break apart.

好的。但这就是为什么SEO机构如此受欢迎，因为它们通过整体优化网站来提升搜索结果。它们创建那些与关键词对齐的内容，并建立反向链接来提高你在搜索引擎中的相关性和信任度。如果你曾与任何内容机构合作过，比如我几年前合作过一家做大量内容的机构，那全是关于关键词的，有时甚至损害了实际内容质量。你知道，这真是个平衡的艺术。

Okay. But this is why SEO agencies have been so popular because they improve search results by optimizing your site in general. And they create those keyword aligned content and they build those backlinks to boost your relevance and trust with search engines. If you've ever worked with any sort of content agency or anything like that, like I've worked with an agency that did a lot of content a couple of years ago, and it's all about the keywords, sometimes to the detriment of the actual content. You know, it's a real balancing act.

没错。因为说到底，预期可能是人们并不真正阅读内容。你只是想要网站流量。

Yeah. Yeah. Because at the end of the day, the expectation probably was people aren't really reading the content. You just want the website traffic.

是啊。对吧？完全正确。

Yeah. Right? Exactly.

但这其实是个很好的观点，因为相比之下，AEO机构或任何从事AEO工作的人，为了提升效果，必须塑造非常清晰、可引用的答案。内容确实需要非常优质。然后你需要将这些内容分发到博客、论坛、评测网站等可信来源，任何人们可能寻找相关信息的地方，并以AI引擎能轻松检索和引用的方式构建内容。所以突然间内容变得极其重要。

But, but that's actually a great point because in contrast, AEO agencies or, you know, anyone working on AEO, you need to in order to improve results, you have to shape really clear, quotable answers. Like, the content does need to be really good. And then you need to spread that content across trusted sources like blogs, forums, review sites, anywhere that people might look for relevant information, and then structure that content in a way that AI engines can easily retrieve and cite it. So now all of a sudden the content is super important.

是的。所以与其专注于一篇非常强大的内容或死磕网站，制作那些关键词密集的内容之类，你反而需要无处不在。是这样吗？

Yeah. So instead of focusing on one really strong piece of content or really dialing in on the website and making all of those keyword heavy pieces of content or whatever, you kind of need to be everywhere. Is that right?

完全正确。而且你需要真实地无处不在，如果这说得通的话。所以，我认为对我们观众来说很棒的是，这重新创造了公平竞争的机会。比如在传统SEO中，假设你是一家初创公司对抗那些花大钱、拥有巨大域名权威的老牌玩家，他们已经深耕多年，你基本上很难翻身。你永远无法在那些关键词排名中获得第一第二的位置。

Totally. And you need to be authentically everywhere, if that makes sense. So, you know, what I think is great for our audience as we think about this, it's an opportunity to level the playing field again. So, with traditional SEO, let's say you're a startup going against big established players spending tons of money on this, you know, with massive domain authority, they've been in it for a long time, I mean, you're kinda screwed to go up that. You're never gonna get, you know, number one, number two in those keyword rankings.

是的。但通过AEO，你可能会被提及，比如在相关的维基百科文章、博客，甚至Reddit帖子里，如果你有意向的话。但这可能就是你能立即出现在AI答案中的关键区别。

Yeah. But with AEO, you can get mentioned in, you know, a relevant Wikipedia article, a blog, maybe even a Reddit thread if if you're so inclined. But that that could be the difference between you starting to appear in AI Answers immediately.

没错。这看起来可能很简单，但再次强调，关键词是‘真实’，因为像Reddit这样的平台能非常快速轻易地识别你是否在为了引人注目而发垃圾信息。所以‘真实’这个词相当重要。AI引擎在检测垃圾信息方面越来越强。Reddit有非常严格的社区准则和审核机制，他们不会容忍那些明显在推销的人。

Yeah. So this, it might seem really easy, but again, that keyword is authentically because something like a platform like Reddit will know very easily and very quickly if you are spamming just to get noticed. So the word authentically is quite important. The AI engines are getting really good at detecting spam. Reddit has very strong community guidelines and moderation, they just don't take kindly to people who are very transparently selling.

他们会很快封禁你。但如果你是真心提供帮助，确实在帮人们解决问题，而恰巧提到了你的播客或其他你想推广的内容——我是说，如果作为早期采用者，我们在Reddit上和小企业讨论AI，这样我们就能通过真实的对话和内容帮助他人，解决问题并真正提供价值，从而开始融入Reddit。

They will shut you down really quickly. But if you're genuinely helpful, if you're actually helping people solve problems and it just so happens that you talk about your podcast or, you know, whatever it is that you're trying to, I'm just saying, you know, it's something that, but you know, if we, early adopter, we're on Reddit talking to people about AI for small businesses, that's how we would start getting involved in, in Reddit because we're having those conversations and helping people out through authentic content and, and we're solving problems and actually providing value.

插一句题外话，这点可能值得稍后深入探讨。根据我的观察，我看到很多AI生成的内容被发布，尤其是涉及公司时。人们似乎开始意识到这点，并尝试大量发布类似‘嘿，我在解决这个问题，这是我的公司，感兴趣的话...’的内容，以便出现在AI结果中。但正如你所说，很多人会指出‘这是垃圾生成内容，得把它清除掉’。

One quick little aside, and this might be something to, to drill into later. Just as an anecdotal observation, I'm seeing an I'm seeing a lot of AI generated content being put out there, especially when referencing companies. So, I think people are starting to cotton onto this and are starting to try to, like, put a lot of, like, oh, hey, I'm solving this thing and here's my company, if you're interested in, so that they show up in these AI results. But to your point, you see a lot of people calling them out, like, oh, this is this is shit generated content. Like, we gotta get this out of here.

但你知道，这确实存在一种权衡：既想排名靠前，又要保持真实以免被人指责损害品牌形象。同时，某些子版块的版主会如何踢出这些垃圾信息发布者？我见过很多帖子因为太像广告而被迅速删除。Reddit就像个有趣的微观世界，展现了这一现象的演变。

But, you know, so there there really is that kind of trade off between, you know, trying to rank there, being authentic so that people aren't calling you out and kind of damaging your brand. And then also, like, what is the what are the moderators doing for some of these subreddits to, like, kick those people out? Because I've seen a lot of, like, someone get posted and then get removed if they're getting too spammy. So Reddit's a really interesting kind of little microcosm to see how this is playing out.

完全同意。有趣的是，我刚刚在我们即将制作的节目数据库中记下，应该做一期关于‘AI垃圾内容’的节目，因为这正逐渐成为大问题。正是由于人们开始利用AEO，情况会越来越严重。关于最有效的内容类型——下周讨论实用技巧和策略时我们会深入探讨——最适合AEO的内容其实很简单直接：清晰、简洁、能直接回答问题的陈述。

Totally. And you know what's funny is I literally just put in our upcoming episodes database that we have that we should do an episode around AI slop because that is starting to become increasingly a problem. And for exactly this sort of reason, as people start to cotton on to AEO, it's going to become a bigger and bigger issue. So when it comes to the content that works best, and again, we'll really get into this next week when we talk about practical tips and strategies, but the content that works best for AEO is pretty straightforward and factual. So just really clear, simple statements that directly answer questions.

你只需要非常明确的表述，比如‘Slack能集成2000多个应用’或‘Zoom最多支持1000名参与者’。这些清晰可引用的事实，AI能轻松提取并用于回答。

You just want very clear, like, you know, Slack integrates with over 2,000 apps or, you know, Zoom supports up to a thousand participants. Just very clear, quotable facts that AI can easily pull and use when it's responding.

我认为这提出了一个非常好的问题。你知道，究竟是什么让内容更有可能被选中？

I think that brings up a really good question. You know, what actually makes that content more likely to get picked up?

基础仍然重要。这就是为什么SEO基础对AEO仍然重要。比如清晰的标题、自然语言。自然语言很关键，它本质上就是我们现在的对话方式，以及人们如何自然地提问，而不是

The basics still matter. This is why the basics of SEO still matter for AEO. So it's things like clear headings, natural language. Natural language is an important one. Natural language essentially means just the way that we're talking right now and how people would ask a natural question as opposed to,

那种MBA术语式的表达。

you know, that MBA buzzword speak.

对，完全正确。内容应该听起来像人们实际交谈的方式，而不是充满关键词的机器人式表达。在我看来，SEO的问题就在于它变得...谢谢

Yeah, exactly. Exactly. So it's content that sounds like how people actually talk rather than just very keyword heavy robot speak, which in my opinion is a problem with SEO is that it gets Thanks

天啊，对吧？

god. Right?

是的，变得过于依赖关键词了。

Yeah. It gets too keyword heavy.

有多少次你访问一个网站，读他们的介绍，然后完全不明白他们是做什么的，就因为内容全是关键词堆砌。没错。

How many times do you show up on a website and you read what they do and you're like, I have no idea what you do because it's so keyword heavy. Yep.

现在这确实是个问题。

Now That's it's a real problem.

是啊。现在我们基本上可以说，嘿，这就是我们实际做的，用五岁小孩都能懂的方式解释。这对我来说太棒了。

Yeah. Now we get to basically say, hey, this is what we actually do and make it explain it like I'm five. That's gonna be great for me.

我喜欢，我超爱解释。

Love, I love explaining

相信我，我这儿就是个普通人。我们需要简单的答案。直接告诉我细节就行。

it Trust like I'm simple folk over here. We need the simple answers. Just give me the details.

还有个东西叫结构化数据标记，你可以回听我们那期关于数据的节目，我们讨论过结构化数据。但说实话最重要的是，像人类一样为人类写作。这就是像AI替换这类问题开始出现的地方。

So there's also something called structured data markup, and you can go back and listen to our episode about data, where we talk about structured data. But honestly the biggest thing is, is just writing like a human for humans. So this is where something like AI Swap starts to become an issue.

对。这和对话式搜索也有关系，对吧？

Yeah. And this kind of ties into conversational search, right?

没错，正是如此。想想搜索行为的变化，五年前人们可能在谷歌输入'男士跑鞋'这样的关键词，就三四个词。但现在他们直接和AI进行完整对话，可能会说'我需要一双150美元以下适合马拉松训练的男士防水跑鞋'。

Yeah, that's right. That's right. So if you think about how search behavior has changed, someone, you know, five years ago, someone may have typed something like running shoes for men into Google, right? Just three or four words, but now they're having full conversations with AI. So they might say something like, you know, I need waterproof running shoes for men that are under $150 for marathon training.

最佳选择是什么？几个月前我在找家具时，其实经常用这个功能，因为我当时在网上到处搜寻非常具体的东西，会和聊天界面反复对话，告诉它'这就是我要找的'，'它需要长这样'，'我需要实现这个、这个、这个功能'。最后它给出了一些还算可以的选项。

What are the best options? When I was searching for furniture a couple of months ago, I actually used this a lot because I was, I was scouring the internet trying to find something really specific and I would get into a back and forth conversation with chat where I would tell it, this is exactly what I'm looking for. It needs to look like this. I need to do, you know, this, this, this. And it came back with some okay options.

我最终一个都没选，但至少它帮我稍微拓宽了选择范围。这种来回对话更像是十五到二十个字的完整查询，而不是三两个字的关键词搜索。就像电报（那种驿马快信式的电报）和我们现在的真实对话之间的区别。

I didn't end up going with any of them, but at least it kind of widened my net a little bit. So it's just much more of that back and forth and it's a much longer, like, you know, that's a fifteen, twenty word query instead of a three word keyword search. So it's it's like the difference between a telegram, you know, like the Pony Express telegram, or an actual conversation that we would have now.

没错。用户体验路径完全不同。如果你在第一次十五到二十字的查询后没找到想要的，你会继续提出澄清问题，提供更多背景和细节。突然间，整个上下文窗口可能就变成你用上千字描述需求——这和三个关键词搜索截然不同。这让我想起早年学用谷歌时，还得用引号来做精确关键词匹配之类的技巧。

Yeah. And the user journey is very different. So, you know, know, you're all are always be iterating concept of if you didn't find what you liked after that first fifteen, twenty word query, you're asking it, clarifying questions and giving it more context and more detail of what you're looking for. And all of a sudden, actually then, if you take that entire context window, now it's like a thousand words of you explaining what you're looking for, very different than that three keyword search. So, you know, I remember learning how to Google back in the day, you know, using quotes for direct keyword match and stuff like that.

在现在这种持续迭代的思维模式下，那些技巧基本都过时了。

All of that's kind of irrelevant now in this always be iterating mindset.

完全同意。还要考虑语音搜索的普及让这个趋势更明显。比如开车时问手机天气，不会说'天气纽约'，而是会说'纽约这周末天气怎么样？我在考虑搞个烧烤派对之类的'。

Absolutely. And then you also have to factor in voice search, which amplifies things even more. So, like, if I'm driving and I asked my phone about the weather or something, don't say weather New York. I would say, you know, what's the weather going to be like in New York this weekend? You know, I'm thinking about having a barbecue or something like that.

这才是人类真实的说话方式，没人会只说'天气纽约'。

That's how people talk. They don't just say weather New York.

首先，这个计划太棒了。祝你周末真能办成烧烤派对。不过我觉得...

First off. Love that. Hope you do have a barbecue this weekend. I don't think

我会的，真遗憾。伦敦的天气可不怎么样，确切知道明天会下雨，所以是的。嘿，

I will, sadly. Don't think the weather in London is, mean, know for a fact it's going to rain tomorrow, so yeah. Hey,

你知道吗？其实确实如此。我们过去常在雨中烧烤。

You know? Actually, that's true. Used to barbecue in the rain.

确实是这样。

That is very true.

借口。要像个冠军一样行动。

Excuses. Play like a champion.

凯尔，我觉得你不会想让我来烧烤的。我会把任何食材都烤成炭。我的厨艺可不怎么样。

I don't think you want me to barbecue, Kyle. I will charcoal any item that I put on that barbecue. I am not known for my cooking skills.

有道理。有道理。好吧。但让我们直击核心问题——如果AI直接给人答案，企业要怎么获取流量呢？

Fair enough. Fair enough. Alright. But let's let's let's get if we get to the heart of this. So if AI is giving people answers directly, you know, how are businesses supposed to get traffic?

要知道，过去25年我们一直在追逐流量。流量。流量。更重要的是，如果没有流量，我该怎么赚钱？

You know, we've spent the last twenty five years traffic. Traffic. Traffic. And then more importantly, like, how do I make money if I'm not getting any traffic?

嗯，这可是个价值十亿美元的问题。这也正是为什么人们需要了解AEO并对其保持警觉。想象一下，当你浏览网站时，那更像是个耗时费力的个人流程。你很可能会因为厌倦点击各种链接而半途放弃。

Well, that is the billion dollar question. And that's exactly why people need to know about AEO and be be aware of it. If you think of it like when you would go to websites, right. That's a much more like individual time consuming process. And you're probably going to give up halfway because you're tired of clicking on various links.

现在你可以在聊天中获取所有信息。这将变成一个更自然的过程，但我们在实践中开始看到成效。再次强调，这一切都非常非常新，甚至就在过去几个月里才真正爆发。要知道，情况可能会变化，但这是目前我们观察到的现象——品牌权威正在发挥作用。

Now you're getting all the information within the chat. And so it's going to become like a much more natural process, but what we're starting to see work in practice. And again, this is all incredibly, incredibly new, even within the last couple of months, it's, it's really kicked off. You know, this might change, but this is what we're seeing right now. That's working brand authority.

因此即使没人点击链接，只要被引用为某个领域的专家，就能建立巨大的品牌资产。人们会开始搜索你，或者因为多次看到你作为专家解答而信任你的权威性。这是非常重要的。其次，虽然AI搜索正变得零点击——我是指所有交互都在聊天框内完成，就像我刚说的——但复杂购买决策仍会驱动网站访问量。

So even if someone doesn't click through, if you're being cited as an expert on the topic that builds massive brand equity. So people start to look you up or people start to trust you as an authority because they've seen you as the expert answer multiple times. So that's a really big thing. And then second, while AI searches are becoming zero click and by zero click, I mean, everything's happening within that chat. What I was just talking about, Complex purchasing decisions are still driving website visits.

对于B2B买家来说，你会想要核实信息。在做决定前会想深入了解。所以虽然可能用AI搜索指明方向，或让你注意到一些之前不知道的工具，但最终你还是会访问网站获取更多信息，查看案例研究等资料。

So for a B2B buyer, you're going to want to verify information. You're going to want to dig deeper before you make decisions. So while you might use the AI as the AI search as point you in the right direction or kind of make you aware of some tools that maybe you weren't aware of before, you're still going to go to that website just to get more information, get case studies, you know, whatever the case may be.

这非常有趣，因为实际上通过AI推荐点击过来的人资质要高得多。就像销售漏斗概念，你基本上已经经过AI预筛选，已经了解公司情况，而且它直接回答了你问题。所以我们应该会看到来自AI引擎的流量转化率非常高。

Which is super interesting because, you know, effectively, the people who do click through from an AI recommendation are far more qualified. You talk about like the funnel. You know, you're basically been pre vetted by the AI. You already know about the company now and it answered your direct question. So, we should see really high conversion rates from traffic coming from AI engines.

没错，这非常关键。AEO正在成为一个渠道，而非取代所有渠道。所以你仍然需要传统SEO、付费广告、自然社交推广等所有方式。

Yeah, absolutely. And this is really key. AEO is becoming a channel. It's not replacing all channels. So, this is why you do still need traditional SEO, paid ads, organic social, all of that.

如果忽视这点，就意味着你会错过那些已经在AI引擎里做出购买决策的客户，因为这种情况正变得越来越普遍。明白吗？

If you ignore it, it means that you're missing customers who are already making purchase decisions inside the AI engines, because that is becoming increasingly common. Okay.

让我们看一个具体实例。塔吉特百货刚刚宣布彻底重构他们的搜索策略，称之为生成式商务。哦，这个时髦词真棒。没错，正是如此。

Let's take a look at a concrete example of how this is playing out. So, Target just announced they're completely rethinking their search strategy around what they call generative commerce. Oh, love that buzzword. Yeah. Exactly.

这不仅是SEO优化，而是彻底改变了他们看待产品发现的方式。

This isn't just tweaking their SEO. It's a total shift in how they think about product discovery.

这个用例非常有趣。我要先说明一点，我尝试寻找过小型企业的用例——毕竟塔吉特资金雄厚，有专业团队处理这类项目，但小企业目前还不具备这种条件。我确实很难找到初创或小型企业的案例，因为这项技术还处于早期阶段。以塔吉特为例，他们不再像过去那样针对'女士冬季外套'这类关键词优化单品页，而是创建贯穿整个客户旅程的综合性内容体验。

This is a really interesting use case. And I'm going to caveat this with, I did try and find a use case for a smaller business because Target obviously has a lot of money behind it. They have people and agencies that they can throw at this sort of thing, but they just don't exist yet. I really struggled to find a use case that was a startup or a smaller business because it's just still such early days with a lot of this, but in the target instance, so instead of optimizing individual product pages for keywords like winter coats, women, which is what they would previously have done. They're creating comprehensive content experiences that kind of walks through the entire customer journey.

比如当用户询问AI：'芝加哥通勤穿什么冬季外套既能应对客户会议又不超过200美元'——塔吉特希望成为这个问题的答案。这要求变得极其具体，因此他们正在构建所谓的场景化产品故事。这类内容不仅罗列产品特性，还会阐述产品如何融入真实生活场景，非常有意思。

So if someone asks an AI, what's the best winter coat for commuting to work in Chicago, that's styling us enough for a client meeting and under $200. Target wants to be that answer. So again, it's getting much, much more specific. So they're building what they're calling contextual product stories. So this is content that doesn't just list the features, but talks about how the product fits into real world scenarios, which is really interesting.

现在他们的冬季外套内容会包含：适用零下15华氏度、可机洗、能搭配商务休闲装、尺码说明等。相比过去那种关键词堆砌的产品描述，现在提供了更完整的信息图景。

So their winter coat content will now include things like works in temperatures down to 15 degrees Fahrenheit or machine washable, or, you know, you can wear over business casual clothing or, you know, what the sizes are. So it's giving a much fuller picture rather than just again, that really keyword heavy product description that it would have had previously.

作为产品人，这让我特别兴奋。我们之前提过，比如你的Shopify店铺，产品描述需要更详细。这正是AI获取信息并理解上下文的地方——它不会只看产品名称。

Yeah. This makes the product person to me super excited. You know, I think we've, we've kind of mentioned this in the past, like in your Shopify store, if you have one of those, your product descriptions need to be a lot more detailed and descriptive. This is where, you know, AI is going to find that information and use the context. It's not just gonna look at the name of your product.

AI会通过描述判断相关性。最让我兴奋的是，所有关于客户旅程、痛点解决方案的工作——特别是在B2B领域——现在都能被AI充分利用。这真的很令人期待。

It's gonna look at that description to figure out if it's relevant. But I think what makes me excited is all of that work, you know, and especially in like a a b to b relevance, all that work about customer journeys and customer pain points and how you solve a thing is now going to get used by the AI. So it's kind of exciting.

确实如此。他们正在精心构建这些信息，以便AI引擎能够轻松解析——又提到这个词了——并引用。根据零售行业专业杂志《Retail Dive》的报道，早期数据显示，通过AI推荐发现塔吉特商品的顾客，其购买意向比传统搜索结果高出47%。但我觉得塔吉特策略更高明之处在于，他们不仅优化了初始发现环节，而是着眼于整个对话过程。

Totally. And they're really structuring that information so the AI engines can easily parse, again, there's that word, and quote it. So according to Retail Dive, which is an industry trade magazine for the retail industry, early results are showing that customers who discover Target product through AI recommendations have a 47% higher purchase intent compared to traditional results. But I think one thing that's even really smart about Target's approach is that they're not just optimizing for the initial discovery. They're thinking about that entire conversation.

再次强调，这个完整的客户旅程可能涉及与AI讨论冬季准备、职场着装规划或礼物挑选。他们希望参与整个对话，而非仅针对单一产品查询。这种与传统SEO产品发现截然不同的方式非常有趣。

Again, that entire customer journey that someone might have with an AI about getting ready for winter or planning a work wardrobe or finding gifts. They really want to be part of that whole dialogue, not just for one product query. So it's a very interesting, very different approach to the traditional SEO product discovery.

是啊，听起来他们几乎把这些AI引擎当作私人购物助手来对待。

Yeah. I mean, it sounds like they're treating these an AI engine almost like a personal shopping assistant.

完全正确。我们之后肯定会讨论沃尔玛及其超级智能代理，这对AEO非常重要，但今天先不展开。请放心我们已为此准备了完整案例研究。但你看，结果不言自明——转化率提升明显，因为当用户访问网站时，他们已经处于更深层的决策阶段了。

Absolutely. And we will definitely at some point talk about Walmart and their super agents because this is very relevant for AEO, but we're not going to get into that today. Just know that we've got a whole case study just about that. But, you know, the results are speaking for themselves because they're seeing these higher conversion rates because by the time someone gets to the site, they're much further along in that decision making process already.

对。我认为很多企业没意识到的是，这不仅仅是关于为网站带来更精准流量的问题。

Yeah. I think what a lot of businesses are missing about where this is headed is, we're not just talking about better qualified traffic to your website.

没错。因为终极目标根本不是把用户引向网站。平台正在努力构建的是完整的端到端购买体验。举个例子，只要看看合作动向就明白了：Perplexity已与PayPal达成平台内购合作，这就是未来趋势。

Right. Because the end game is not about sending people to your site at all. The ultimate goal, and this is what the platforms are working towards, is building a complete end to end purchasing experience. So, as an example, you can see where this is going purely if you look at the partnerships. Perplexity already has a PayPal partnership for in platform buying, and that is where this is going.

确实。这挺疯狂的。想象一下，我认为这就是我们正在迈向的场景。对吧？没错。

Yeah. So, I mean, that's kinda wild. I mean, imagine I think this is the scenario we're headed towards. Right? Yep.

比如说，你进入JatGPT，提出需求：'我需要一个15人营销团队使用的项目管理工具，要集成Slack，具备X、Y、Z功能，但月费不能超过50美元/用户'。Chatty Boutique会提供三个选项，你选中一个并在ChatGPT里完成支付——砰！交易完成。我甚至从没访问过那个官网。

Someone or, you know, you go into JatGPT, you say, I need a project management tool for a 15 person marketing team with Slack integration and x y z features, but I need it to be under $50 per user per month or whatever your criteria may be. Chatty Boutique, again, gives you three options. You pick one, you enter your payment details inside of ChatGPT, and now boom, you're a customer. I I never even visited that website.

没错。如果你的产品没被提及，没出现在那三个推荐选项里，你不仅损失了一个官网访问量，不仅是损失了漏斗顶端的流量——而是直接丢掉了整笔交易。连用付费广告重定向的机会都没有。

Yep. So if your product wasn't mentioned in that conversation, if your product wasn't those three options that were provided, you didn't just lose a website visit. So you didn't even just lose that top of funnel traffic. You lost the entire sale. So you lost that conversion and there is no opportunity for your paid ads to retarget them.

销售团队没有跟进机会。完全没有二次机会，因为整个流程——搜索、发现、购买——全在AI搜索引擎里完成，彻底颠覆了我们所知的传统营销模式。

There's no chance for your sales team to follow-up. There is no second chance at all because all of it, the search, the discovery, and the purchase all happened within the AI search engine completely upending everything that we know about traditional marketing.

是啊。传统销售漏斗完全消失了，而你甚至毫不知情。根本不知道发生了什么，对吧？

Yeah. That the traditional funnel has completely disappeared, and you don't even know about it. You don't even know what happened. Right? Yep.

所以正如你所说，AI对话成了完整的买家旅程——从发现到购买全程都在这里完成。而你甚至不知道自己已经出局了。

So the AI conversation becomes the entire buyer journey from, like you said, discovery all the way to purchase. And you had no idea that you lost out.

正是如此。所以说AEO（天啊这个词真拗口）...所以说AEO优化至关重要。这不只是品牌曝光的问题，而是要在消费者做购买决策的关键时刻出现在他们面前。

Exactly. So that is why AEO God, it is such a, it's such a hard one to say. That's why AEO optimization is so critical. It's not just about brand awareness. It's about being present at the exact moment that someone makes that purchasing decision.

而这个关键时刻，正越来越多地发生在这些AI平台内部。

And that moment is increasingly happening within these AI platforms.

是的。当我们谈论企业从AI流量中获得六倍转化率时，这很可能将是最后一代能看到这个指标的群体了。要知道，未来的客户在终极场景下，可能根本不会有可追踪的点击或流量。

Yeah. So when we talk about businesses getting six times higher conversion rates from AI traffic, that's actually gonna be probably the last generation to see that metric. You know, future customers, when we get to this end game, know there might never be a a trackable click or traffic at all.

没错。购买行为完全发生在对话中。如果你没有参与那场对话，对那部分客户群体来说你基本上就是隐形的，这绝对不是你想要的。所以这就是我们要讨论这个问题的原因——提前思考这个问题至关重要。

Yeah. The purchase happens entirely in the conversation. And if you're not part of that conversation, you are invisible essentially to that customer segment entirely, which is absolutely not what you want. So that's why we wanted to talk about this. It's so important to be thinking about this.

所以过去专注数据测量的那个我，内心可能正在经历某种消亡。但关键问题是：工具在哪里？测量维度是什么？我们到底该如何追踪这些数据？

And so the data person inside of me focusing on measurement for the last however many years is, you know, maybe dying a little bit inside. But, you know, what are the tools? What does the measurement side look like? You know, how do we even track this stuff?

这正是当前最大的挑战之一。当你的内容被AI回答引用时，传统指标如自然点击量就变得无关紧要。你需要开始追踪诸如AI回答中的声量份额——被提及次数、品牌提及频率，Graphite创始人Ethan Smith在Lenny播客中称之为'引用优化'。

So that is one of the biggest challenges right now. Traditional metrics like organic clicks become less relevant when your content's being quoted in AI responses. So you're going to need to start keeping track of things like share of voice in AI answers. So how many times are you mentioned, branded mentioned frequency what Ethan Smith, who's the founder of Graphite and did the Lenny's podcast. It's what he's calling citation optimization.

简单说就是你的被引用频次。现在确实涌现了一些新工具，比如Graphite公司的AEO追踪器，这也是为什么这位创始人能成为该领域权威。我强烈推荐听那期播客，内容非常精彩。

So like just how often you're cited. There are definitely some new tools that are emerging. So Graphite, this company has an AEO tracker. That's one of the reasons why this guy is the authority on it. And I do highly recommend listening to that podcast, because it is really interesting.

还有几家公司正在开发AI可见度指标，但目前仍处于早期阶段，实际可用的追踪测量方案还非常有限。

And there are a couple of other companies that are building AI visibility metrics, but it's still such early days that there actually isn't a ton of tracking and measurement options quite yet.

是的。这个话题我们下周战术研讨会上会深入探讨。虽然AI机器人的流量有时会被隐藏或标记为未知（如果它们清除了请求头等信息），但显然还是有方法识别这类流量。关键是一台AI机器人的访问可能意味着你在AI回答中被提及了数千次——看似单一流量背后可能隐藏着指数级的影响力。

Yeah. I think that's something we'll cover more next week during the more tactical session. You know, there's obviously ways to see traffic coming from AI bots, although sometimes that is hidden or identified as unknown, if they scrub headers and stuff. But, you know, still traffic from an AI bot could mean thousands of mentions in AI answers. So you might see one AI bot come to you for answers, but that could be, you know, there could be an exponential amount of, like, let's call it traffic that you're getting from that.

那么对于感到又要掌握新事物而压力山大的听众们来说，我得说这确实有点像面对一个黑箱。对吧？你有什么建议？今天该从何入手？

So for our listeners who are feeling overwhelmed by yet another thing they need to master, and also, I gotta say, you know, something that feels a little bit like a black box. Right? What is what's your advice? What do you start doing today?

我们下周肯定会深入探讨这个问题。这也是我们把这两期节目分开的原因。因为今天这期主要是介绍背景知识，说明其重要性。下周我们才会真正深入讲解。但我觉得关键是要从你的专长领域开始。

I mean, we'll definitely get more into this next week. Like, that's why we wanted to separate out these two episodes. Cause I think today is all about, or this episode is all about just giving some background on why it's important. Then next week we'll really get into it. But I think a lot of it is just starting with your expertise.

选择一个你真正比他人更了解的领域，围绕该主题的常见问题创作真正有用、清晰的内容。就像我刚才说的，如果我们在Reddit的小企业论坛上讨论如何使用AI，就能逐步积累引用和专业知识，开始被更多人引用。但核心在于真实性，在于用创造价值的方式分享专长，不要贪多求全。毕竟很多人甚至还没听说过这个概念。

So pick one area where you genuinely know more about the topic than other people and create really helpful, clear content around the questions that people ask about that topic. So again, it's what I was just talking about with, you know, if we were to go onto Reddit and start talking to people about, you know, in the small business forum about how to use AI, would be us building up our citations, building up our expertise and starting to be able to get quoted a little bit more. But again, it's all about the authenticity. It's all about you sharing your expertise in a way that brings value, but don't try and boil the ocean. A lot of people haven't even heard of this yet.

光是现在收听这期节目，你就已经领先于许多尚未理解这个概念的人了。所以早期采用者确实能获得优势。

Even by listening to this episode right now, you are going to be ahead of a lot of people who haven't even wrapped their head around this concept yet. So early adopters, do really have an advantage with this one.

下周我们将讨论具体策略。包括实用战术、工具、实施框架等所有能帮你获得早期优势的内容。

And next week we'll get tactical. Specific strategies, tools, implementation frameworks, everything you can use to kind of get that early, early adopter advantage.

没错。看，我们正在示范——关键词重复。因为理解AEO的重要性只是第一步。

Yeah. How many times can we say? See, we're just doing it right now. We're a keyword, keyword, keyword. Because again, understanding why the AEO matters is just the first step.

真正的价值在于掌握如何针对答案引擎进行优化，以及当前最有效的方法。毕竟三个月后我们可能就得做更新节目了，这个领域变化太快。

The real value actually comes from knowing how to optimize for the answer engines and what's working at this moment in time. Cause again, we're probably gonna have to do an update episode on this in three months because it's gonna change.

完全正确。今天的关键要点是，如果你要从这一切中学到一件事，那就是搜索正在从‘找到最佳页面’转变为‘获取最佳答案’。如果你想让你的业务成为这些答案的一部分，你需要开始像ChatGPT这样的答案引擎一样思考，而不仅仅是像过去多年那样优化搜索排名。

Totally. So the key takeaway for today, if you're gonna learn one thing from all of this, search is shifting from find the best page to get the best answer. And if you want your business to be part of those answers, you need to start thinking like an answer engine, like ChatGPT, not just optimizing for search rankings like you have been for the past however many years.

没错。所以这不是要取代你正在做的事情，而是要让你的方法适应一个AI日益成为你的内容和客户之间中介的世界。所以不要完全放弃SEO。现在有很多关于SEO已死的说法。

Exactly. So it's not about replacing what you're doing. It's about evolving your approach for a world where AI is increasingly the middleman between your content and your customers. So don't give up on SEO entirely. There's lots of SEO is dead stories going around right now.

SEO并没有死。这只是SEO的进化。好吧，

It's not. This is just an evolution of SEO. Alright,

杰斯。现在让我们深入探讨对我们听众真正重要的AI新闻。本周有一个故事对我们的观众来说非常重要，即使你现在还没有意识到。

Jess. Let's now dive into AI news that actually matters for our listeners. And this week, there's one story that is pretty important for our audience, you know, even if you don't realize it yet.

所以，你是在说OpenAI和NVIDIA的合作关系。对吧？正是如此。

So, you are talking about the OpenAI NVIDIA partnership. Right? Exactly.

事情是这样的。OpenAI和NVIDIA宣布了一项战略合作，OpenAI将部署价值10吉瓦的NVIDIA AI系统。为了让大家有个概念，这相当于足够运行一个小型城市的计算能力。

So, here's what happened. OpenAI and NVIDIA announced a strategic partnership where OpenAI is going to deploy 10 gigawatts worth of NVIDIA's AI systems. And to put that in perspective, that's enough computing power to run a small city.

首先，我们是否应该提醒大家NVIDIA是什么？因为在这个领域它是个大名字，但普通人可能除了新闻之外没听说过它。

Well, first of all, do we want to just remind people what NVIDIA is? Because that is a big name in this space, but your average person might not have heard of them beyond a news story.

好吧，我要提醒大家，英伟达CEO黄仁勋说过，他正在将AI推向他预想的方向。没错。我们在

Well, I'll remind everyone with Jensen Huang, the CEO saying that he is moving, AI exactly where he intends it to go. So Exactly. We saw him at

我们在伦敦科技周上见到他，这是他的原话。

we saw him at London Tech Week, and that is a direct quote.

这让你稍微了解到，虽然不想自夸，但在某种程度上，他确实是推动AI革命的核心人物之一。

So that gives you a little instance that he, you know, not to toot his horn, but I guess in some ways, he really is one of the driving forces behind the AI revolution.

他确实是。没错。

He definitely is. Yeah.

这是因为英伟达制造的GPU和芯片为大量AI处理能力提供动力。他们确实处于技术前沿。他们最初为游戏开发的GPU技术，后来应用于比特币挖矿等领域，真正推动了AI革命的发生。现在他们通过持续研发更强大的处理单元，推动AI实现更多推理和思考能力。他们就像是支撑AI发展的基础构件。

And that is because NVIDIA builds the GPUs or chips that power a lot of the AI processing power. They're really on the cutting edge. They kind of their advancements with GPUs, which was originally for like gaming and then eventually Bitcoin mining and all that stuff, it really enabled the AI revolution to happen. And now they are pushing it forward with continually, building and creating stronger and stronger processing units so that AI can do a lot of the reasoning and thinking. So they are like the fundamental building block that is enabling AI to happen.

他们现在也是全球最具价值的公司之一，对吧？是的。这就是为什么黄仁勋能说出那样的话。

They're also one of the most valuable companies in the world now, aren't they? Yeah. This is why, this is why Jensen can say that sort of thing.

不过说真的，如果我说自己没持有一点他们的股票，那是在撒谎

But I mean, lying if I said I didn't own a little bit of stock to I get in

我是说，这很公平。但我们的观众为什么要真正关心这个？因为如果你是一家只有20名员工的初创公司，他们为什么要关心这种基础设施的故事？

mean, fair. But why should our audience actually care about this? Because if you're a a startup with 20 employees, like, why should they care about this infrastructure story?

是的。我是说，这些芯片可不便宜。你知道，最便宜的型号可能也要5美元左右。而且主要的是，我去过一些超微计算机处理NVIDIA的活动，那里...

Yeah. I mean, these chips aren't cheap. You know, I think some of the cheapest models are $5 or something like that. And primarily, you know, I've been to some of these, you know, micro, you know, super micro computer processing NVIDIA events, and it's

听起来，哇，这很令人兴奋。是的。

like, Wow. That you sounds exciting. Yeah.

但是，你知道，这就像是大型公司部署配备这些GPU的巨大数据中心。过去这些资源相当难以触及。这次与OpenAI的合作公告，我认为开始让更多人能够接触到这些强大的AI组件。现在如果你想为企业运行高级AI模型，比如用于客服或内容生成，你会受限于计算能力和成本。你必须深入研究谷歌或云环境才能运行很多这类应用。

And but, you know, it's it's like massive companies deploying huge data centers with these GPUs. It's been pretty inaccessible. This announcement in this partnership with OpenAI, I think starts to democratize a lot of that access to, you know, these powerful AI components. Right now, if you wanna run advanced AI models for your business, say for customer service or content generation, you're limited by computing power and cost. You're having to delve into the, you know, Google or, like, the cloud environments to be able to run a lot of that.

这次合作意味着OpenAI能够以显著更低的成本提供更复杂的AI能力。我们谈论的是能够处理所有这些复杂推理的AI，这些以前可能只有那些打算自建AI数据中心的大型科技公司才能拥有。

This partnership means OpenAI can offer much more sophisticated AI capabilities at significantly lower cost. So we're talking AI that can handle all of this complex reasoning that was probably only available to really large tech companies who are gonna deploy their own AI, you know, data centers.

那么你的意思是，这最终会渗透到我们日常使用的工具中？

So this then so basically, what you're saying is that this will trickle down to the tools that we use on a daily basis?

完全正确。更好的ChatGPT模型，为小企业提供更强大的API访问，能同时处理视频、音频和文本的AI工具。我们常说AI让中小企业、初创公司能使用与企业客户相同类型的工具，而这次合作将进一步推动这一点。不过有一点需要明确，这是个注意事项。

Totally. Better chat GPT models, more powerful API access for smaller businesses, you know, AI tools that can handle video, audio, text simultaneously. You know, we're always talking about how AI allows, you know, small businesses, startups to have access to the same types of tools that enterprise customers have. And I think this partnership will only further allow that. But, you know, one thing to be really clear about, and this is a kind of a caveat.

这些声明，尤其是这一个，有点像障眼法。说是最高可达10吉瓦的规模，对吧？但这并不意味着OpenAI明天就会实际部署。它更像是一个展示雄心或我们发展方向的标题。

These announcements, and especially this one, is a bit of a shell game. So the announcement is up to 10 gigawatts. Right? It doesn't mean OpenAI is actually deploying that tomorrow. It's more of a headline to show ambition or where we're trying to go.

所以这与其说是一份有保障的合同，不如说更像是表明OpenAI和英伟达正在达成某种共识的信号，可以这么说吧？

So it's less of a guaranteed contract and kind of more of a signal that OpenAI and NVIDIA are aligning, shall we say?

没错，完全正确。它指明了发展方向——大规模、更低成本的计算能力，以及我们可以通过ChatGPT订阅服务接入的更庞大、更先进的模型。但具体时间表和细节都还很模糊。

Exactly. Exactly. It tells us the direction of travel, where we're going. Massive scale, cheaper compute, you know, bigger, more advanced models that we can tap into with our chat GPT use, you know, subscription. But the timeline and the specifics of all that are still very fuzzy.

正因如此，这或许更像是提振股价的手段，而非实质性的收入或利润增长点。

Hence why, you know, this is maybe more of, like, drive the stock price up than it is actually, like, here's revenue on the or bottom line, you know, enhancement.

说到工具普及化，本周其他新闻里我们确实有很多产品发布。之所以要特别强调这些，是因为随着假日购物季临近，这些工具将逐渐面世——事实上我们已经看到这个趋势了，它们将特别助力零售和电商领域。比如亚马逊刚推出的'卖家助手'，本质上是代理型AI（我这么说是比较宽松的定义，毕竟大家对'代理型'都有不同理解）。

And speaking of tools getting more accessible, some of our other news stories this week, we actually have a lot of product announcements. And the reason why we wanted to flag these is as we start to come into the holiday shopping season, you're going to start to see some of these tools become more available, which I think is exactly what we are seeing, because these are all going to help particularly in the retail and commerce space. So Amazon has just launched something called Seller Assistant. So it's essentially agentic AI. And I say that loosely, because we all know what our thoughts are on agentic, but it's yeah, exactly.

带引号的'代理型AI'，专为他们平台卖家服务的。

Finger quotes. It's AgenTic AI for their marketplace sellers.

但这又是个绝佳案例，说明企业级AI正在向中小企业开放。该工具能自动管理库存、优化定价（这点非常重要）、响应客户咨询，甚至处理部分物流工作。

But this is another, you know, perfect example of enterprise level AI becoming available to smaller businesses. The tool can automatically manage your inventory, optimize your pricing. That's very important. Is very important. Respond to customer queries and even handle some logistics.

非常实用。

Very useful.

特别有趣的是，这展示了平台如何开始将AI直接构建到生态系统中。再次强调，你不需要成为这方面的专家。亚马逊基本上是在说，这里有一个AI代理将为你运营业务。这无疑是我们开始看到的发展方向。

What is particularly interesting about this is that it's showing how the platforms are starting to build AI directly into the ecosystems. So again, you don't need to be an expert with a lot of this. Amazon is basically saying, here's an AI agent that's gonna run your business for you. So it's definitely, again, the direction that we're starting to see things go into.

说得好，这个趋势我们随处可见。谷歌刚推出了AI生成的产品摘要功能。比如有人搜索'最佳无线耳机'时，不再只是获得链接列表，而是会看到一个AI整理的摘要，包含优缺点、价格、推荐，甚至可能有个性化记忆功能，让你看到'哦，这个特别适合我'。

Great, great point. That that trend we're seeing everywhere. So Google just rolled out AI generated product summaries in their AI mode. So if someone searches for best wireless headphones, instead of getting a list of links, like we talked about, they get an AI curated summary with pros, cons, pricing, recommendations, maybe even a little memory mode coming into place where you can see like, oh, this is actually really good for me specifically.

这一切都把我们带回到今天的主题。这就是为什么AEO如此重要。因为如果谷歌的AI在总结产品和做推荐，你必须确保你的产品被纳入讨论。实际上，既然谷歌推出了这个功能，就意味着无论你愿不愿意，都不得不跟进。

Which all of this brings us right back into our main topic for today. This is exactly why AEO matters. Because if Google's AI is summarizing these products and making these recommendations, you need to make sure that your products are in that conversation. And actually if Google is introducing this, it means you're going to have to you're going to be forced into doing this, whether you like it or not.

完全同意。尽管AI让很多工具变得平民化，但大公司正在布局，如果我们不抢占先机，他们最终会主导市场。还有个值得关注的产品更新：Perplexity刚推出了一款AI邮件管家，年费200美元。它能阅读、优先处理邮件，起草回复甚至安排日程。

Absolutely. As much as AI democratizes a lot of these tools, the larger players are starting to do this and they will eventually own the market if we don't get ahead of So definitely something to keep track of. One more that caught my attention, another product update. Perplexity just launched an AI agent that manages your email, for $200 a year. So it reads, prioritizes, drafts, responses, and even handles scheduling.

200美元年费乍看挺贵，但比很多企业花在产品管理工具上的费用低。这个定价很有意思，值得研究。

So $200 a year seems pretty expensive off the, like, right off the bat, but that's less than a lot of businesses spend on product management tools. So it's interesting. It's interesting. I think it's worth checking out.

说到这个，让我想起Superhuman这家公司。几年前他们号称推出了'史上最佳邮件工具'，我记得价格大概是每月40美元左右...

Well, and I think an interesting comparison for this is the company Superhuman, which a few years came out and was like, this is the best email tool forever. And I remember, I forget how much it cost. It was like $40 a month or

是啊，价格相当昂贵。我用过之后感觉挺失望的。

Yeah. It was quite expensive. I've used it and I was underwhelmed.

考虑到Gmail是免费的，这价格确实偏高。要是五年前，这种现成的AI助手开发成本会高得离谱，对吧？

It seemed expensive given Gmail's free. And five years ago, this kind of AI assistant to build off the shelf would have cost Oh. A ridiculous amount of money. Right?

完全同意。

Absolutely.

当然比雇兼职助理便宜多了。这个定价很有意思——每月20美元左右，我们反复看到这个价位。

And it's certainly cheaper than hiring a part time assistant. Yep. You know, it's an interesting price point. It's not that much. Again, it's basically like $20 a month again, that we're seeing over and over again.

说实话，自己动手做这个也不难。所以我不确定这是否最划算的投资，但可能他们是在布局生态系统，想整合各种...

I mean, honestly, you could build this yourself pretty easily. So Yeah. I'm not sure if it's the best use of funds, but maybe it's starting an ecosystem play for them about, you know, building all different I

我觉得正是这样。不过本周我们有不少产品发布，这些公告的共同点是什么？

think that's probably exactly what it is. Yeah. But what we've got quite a few product announcements this week. What's the common thread within all of these?

我们经常讨论这个，我再次强调：高级或企业级AI的门槛正在快速消失。无论是英伟达的基础设施、亚马逊谷歌的平台集成，还是像Perplexity邮件助手这样的直接应用，中小企业甚至个人创业者，现在都能获得过去只有巨头才有的能力。好了杰西，新闻到此为止，该进入观众最爱的环节——AI翻车现场。

I think we talk about it quite often, and I'll just reiterate here, which is the barriers to advanced or enterprise grade AI is disappearing very quickly. Whether it's infrastructure from NVIDIA, platform integration from Amazon and Google, or direct like Perplexity's email agent, small, medium sized businesses, you know, startups even, you know, one person solopreneurs, you're getting access to capabilities that used to be exclusive to the only the largest of companies. Alright, Jess. Enough news. Time for everyone's favorite segment, AI Gone Wrong.

我要说，本周的故事对那些在业务集成中使用ChatGPT的人来说感同身受。

And I will say this week's story hits close to home for anyone using ChatGPT with business integrations.

是的。这更像是一个警示故事，而非我们通常喜欢关注的那种搞笑的AI出错案例。牛津大学的研究员宫村英士最近发现了一些ChatJPT的安全隐患。他仅凭一个电子邮件地址就让ChatJPT泄露了私人邮件数据。

Yeah. This is kind of a cautionary tale more than a more than kind of the funny AI gone wrong that we usually like to focus on. So Eito Miyamura, who is a researcher at Oxford, just discovered some, not great security risks with ChatJPT. He basically got ChatJPT to leak private email data just using someone's email address.

另外，你那个名字的发音真是绝了。我要是...

Also, you crushed that, pronunciation of that name. I would

谢谢夸奖。我确实练习过那个发音，因为我想...嗯...把它念对。

just thank say you. I did, I did practice that one because I wanted to Yeah. Get it

太棒了。但他怎么仅凭一个电子邮件地址就做到了？这有点令人惊讶。

Love it. But how did he do that with just an email address? That is kind of surprising.

它利用了ChatGPT的新模型上下文协议（MCP），这个我们之前讨论过。

So it exploited ChatGPT's new model context protocol, which is something that we've talked about. MCP's.

MCP。

MCP.

是的，MCP功能。没错，就是让ChatGPT能连接Gmail、日历和其他商业工具的特性。记得我们把它比作USB接口。

Yep. MCP's. Yep. It's the feature that lets ChatGPT connect to Gmail, calendars, other business tools. Remember we, we described it as like a USB port.

本质上确实如此。但攻击者会利用这个邮箱功能发送看似正常的日历邀请，实际上用极小字体或白色文字隐藏了恶意指令，让人难以察觉。

That's essentially what it is. But the way that it used the email address is an attacker sends an innocent looking calendar invite, but hidden in tiny fonts or white text so you can't see it are malicious instructions.

这真有意思。顺便说个趣事，我在LinkedIn上看到有人分享，他在个人资料里用白色字体藏了份金枪鱼食谱之类的内容。现在收到的工作邀约里要是有金枪鱼食谱，就知道是机器人发的了。

That's really interesting. I I as a quick little aside, I saw something on LinkedIn where a guy was talking about how on his LinkedIn, he adds in, like, in, you know, white letters, a a tuna recipe or something like that. And so now he's getting all these job reach outs with, like, tuna recipes in so he knows if it's a bot.

对，这招很聪明。早有人用这方法处理简历——在空白处藏白色文字能命中大量关键词，帮助绕过ATS追踪系统。

Yeah. Which is really smart. I've seen that for some people were doing that for a while with with resumes or with CVs. They were doing that. If you put white text in, you can hit a bunch of keywords and it and that will help bypass the ATS tracking systems.

求职者追踪系统。没错。

Applicant tracking systems. Yeah.

就是这个，我正想不起这个词呢。所以这不算新点子，显然能用于娱乐或实用场景。但这次...这次可不是什么好事。

That's it. That's the that's the word I was looking for. So it's it's not a new idea and it can obviously be used for entertaining or practical use cases. This was not. This was not not good.

是啊，这次更恶意些。你想，用户很无辜地说'小Chat帮我查日程'，它既会读取正当请求，也会执行隐藏的恶意指令。没有防护机制的话，它根本分不清区别。

Yeah. This one is a little more malicious, but, you know, so, I mean, you say very innocently, hey, chatty pity, check my schedule. It's gonna read both the legitimate request, and then it's gonna read the hidden malicious comment. And, you know, without guardrails, it's not gonna know the difference.

没错。它可能会给你提供日历摘要，但同时也会暗中搜索你的邮件，并将敏感数据发送到攻击者的服务器。由于这一切发生在OpenAI的云端，你的网络不会留下任何痕迹。所以你根本无从察觉。

Exactly. So it might give you your calendar summary, but it will also secretly search your emails and send sensitive data to the attacker's server. And because this happens in OpenAI's cloud, there's no trace on your network. So you would have absolutely no idea.

这正是可怕之处。完全正确。可追溯性是个问题。我的意思是，我猜你能访问邮件里的所有内容，

That's the scary part. Exactly. Traceability is a problem. And I mean, I assume you have access to everything from your email,

对吧？所有内容。包括财务报告、密码、私人对话等等。演示中显示ChatGPT被诱骗搜索特定的敏感信息并发送出去。所以这个演示显然没有出错。

right? Everything. So financial reports, you know, passwords, personal conversations. So the the demo showed ChatGPT being tricked into searching for specific sensitive information and then sending it out. So this is a demo that didn't go wrong, clearly.

是的。它表现得太‘好’了，但我猜他们已经发现了这个问题。希望他们正在修复。

Yeah. It went too right, but I assume they've seen this. Yes. Hoping they're putting in a fix.

对。问题被披露后他们打了补丁。显然这是在研究环境中进行的，是由牛津大学或与牛津有关的研究人员完成的。所以他们已经修复了。

Yep. So they patched it after it was disclosed. Obviously, this was done in a research environment. So this was done by, you know, at the University of Oxford or by someone who works with Oxford. So they have patched it.

但令人担忧的是，这种攻击恰恰针对那些让AI对企业特别有用的集成功能。比如邮件连接器、日历访问，这些正是企业真正需要的功能。

But what is concerning is that this targets exactly the integrations that make AI particularly useful for businesses. So your email connector, your calendar access, those are features that companies actually want.

是的。我认为这也印证了我们很早就强调的观点：假设你输入AI系统的任何内容都可能公开。就像我们常说的‘纽约时报测试’。没错。

Yeah. I think it also reinforces what we've been saying very early on. Assume anything you put into AI systems could potentially become public. It's the New York Times test, as we like to call it. Yep.

要知道，无论是漏洞、数据泄露还是法律要求，这些公开的AI和大语言模型总会存在风险。

You know, whether through bugs, breaches, or legal requests, there's always going to be a risk with some of these public AI and public LLMs.

完全同意。随着AI智能体变得更强大，攻击者肯定会开发出越来越有创意的利用手段。就像垃圾邮件发送者或诈骗电话那样，他们的手法正变得极其老练。所以我们必须保持警惕，显然需要在安全方面做到滴水不漏。

Totally. And as the AI agents get more powerful, I think you can expect attackers to find increasingly creative exploits. It's just like spammers, you know, or scam calls who are now just getting really sophisticated. So it's just something to be aware of. And obviously you need to be really tight on your security.

没错。让我们总结一下关于答案引擎优化的几个关键要点。

Right. So let's wrap up with some key takeaways from our deep dive into answer engine optimization.

首先，搜索正在发生根本性变革，这是最重要的。我们正从'找到最佳网页'转向'获得最佳答案'。如今60%的搜索以零点击结束，这个数字只会继续攀升。现在还只是初期阶段，就已经达到60%了。

First, search is fundamentally changing, and that's the big one. You know, we're moving from find the best page to get the best answer. 60% of searches now end without a click, and that number is only going to climb. We're still early days, and it's only it's already at 60

确实处于早期。其次，AEO虽不同于SEO，但建立在SEO基础上。不再是追求关键词排名第一，而是要获得AI引擎的引用。所以优化目标从点击量变成了在多个可信来源中被引用。

Very early days. Second, AEO is different from SEO, but it builds on top of SEO. Instead of ranking number one for keywords, you're trying to get quoted by AI engine. So instead of optimizing for clicks, you're optimizing for citations across multiple trusted sources.

对。第三点，适用于AEO的内容必须直白且基于事实。要制作可被引用的陈述，而不是花哨的——恕我直言——充满MBA营销套话的文案。AI引擎需要清晰可验证的信息才能放心引用。

Yep. And third, the content that works for AEO needs to be straightforward and factual. Think quotable statements rather than flowery, I hate to say it, MBA marketing buzzwordy copy. AI engines need really clear, verifiable information that they can confidently cite.

值得记住的是，这实际上创造了公平竞争环境。传统SEO时代初创企业往往难以抗衡老牌网站的权威性，那些历史悠久或资金雄厚的玩家总能占优。而通过AEO，你可能明天在相关Reddit讨论中被提及，就能立即出现在AI生成的答案里。

It's definitely worth remembering that this actually levels the playing field because with traditional SEO startups often couldn't compete against established domain authority. So anyone who's been around for a really, really long time or can throw a lot of money at it. With AEO, you can get mentioned in a relevant, you know, Reddit thread tomorrow, and you can start appearing in AI answers immediately.

要知道，我们齐聚于此的原因在于，那些先行者将会看到极其惊人的成果。记住，像Webflow这样的公司实现了比传统谷歌流量高出六倍的转化率。当AI为用户提供了实质性的咨询服务后，这些用户的匹配度会大幅提升。他们了解你的品牌，清楚你的业务。

And, you know, the reason we're all here, the early movers on this are gonna see really, really incredible results. Remember, that six time higher conversion rate from companies like Webflow, you know, compared to traditional Google traffic. Users are so much more qualified when an AI has given them kind of essentially a consultation. They know your brand. They know what you do.

他们带着具体痛点前来咨询，现在正主动访问你的网站。这简直就是营销合格线索中的黄金资源。

They had a specific pain point and they asked about it, and now they're coming to your website. Like, that is, you know, marketing qualified lead gold.

关键在于，AEO并不会取代传统营销渠道，但忽视它就意味着你将错过一个日益重要的接触点。任何现在就开始适应的人，都将获得显著的先发优势。

So the bottom line is that AEO is not going to replace traditional marketing channels, but ignoring it means that you're going to miss an increasingly important touch point. And anyone that starts adapting to it now is going to have a really significant head start.

没错。下周我们将深入探讨具体策略、工具和实施框架——我们最爱的就是好框架，

Yep. And next week we're going to get tactical with specific strategies, tools, implementation frameworks. We love a good framework,

我们才不喜欢什么框架。

We don't love a framework.

但现阶段，只需明白变革正在发生，而且速度非常快。

But for now, just understand that the shift is happening, and it's happening very quickly.

以上就是本周的全部内容。如果觉得本期节目有帮助，请务必在您收听播客的平台订阅。记得我们在社交媒体上也有账号，您可以在LinkedIn、Instagram和TikTok上关注@earlyadopter（注意adopter里没有字母e）。

So that is everything for this week. If you found this episode helpful, please make sure to subscribe wherever you get your podcasts. And remember that we are on social media, you can follow us at earlyadopter on LinkedIn, Instagram, TikTok. Remember, there's no e in adopter.

如果你开始注意到AI引擎回答问题，如果你有关于AI流量的有趣故事，或者你自己有个小目标案例，我们非常希望能看到一些围绕这方面的小型企业案例，因为确实如此。

And if you start noticing AI engines answering questions, if you have an interesting story about AI traffic, if you've got your own little target example, we'd love to see some small business examples around that because Absolutely.

这很难

It's hard

目前还不太清楚。你知道，人们在这方面实际投入了多少。如果你有好的例子，请发邮件至hello@earlyadopter.ai（注意没有字母e的adopter）。我们很乐意听取听众的反馈。

to find right now. It's not clear It is. You know, how much people are really doing around this. So if you have some great examples, email us at hello@earlyadopter.ai,adopterwithoutthee. We love hearing from listeners.

同样值得关注的是你自己的行为。你是如何寻找信息的？你还在用传统的谷歌搜索，还是已经开始使用ChatGPT了？我父亲告诉我他用ChatGPT买了新笔记本，这帮助他确定了想要哪款。

It's also worth noticing your own behavior. How are you finding things? Are you still searching things with traditional Google or have you started using ChatGPT? My dad told me he bought a new laptop using ChatGPT. It helped him figure out which laptop he wanted to use.

哇。

It's Wow.

那可是笔大开销。相当可观的一笔消费。

That's a big purchase. That's a sizable purchase.

确实是大额消费。我想因为他有非常具体的要求，聊天机器人确实帮他指明了方向。所以，观察你自己的行为吧，因为每个人的消费方式都在因此发生巨大变化。总之，我们期待下周继续讨论AEO。感谢收听《Early Adopter》，在这里我们助力让AI成为你的不公平优势。

It's a sizable purchase. And I think because he had very specific requirements for it, you know, the chat really helped him, you know, guide him in the right direction. So yeah, a look at your own behavior because everyone's consumer behavior is changing quite drastically around this. Anyway, we look forward to talking more about AEO next week. Thank you for listening to Early Adopter, where we help make AI your unfair advantage.

下次见之前，请保持创新精神与好奇心。

And until next time, keep innovating and and stay curious.

再见，杰斯。

Bye, Jess.

再见，凯尔。

Bye, Kyle.