SEO in the Age of AI (AEO—AI Experience Optimization)

229
00:57:33.300 –> 00:57:43.879
Julia Nimchinski: Next up SEO. In the age of AI and Aeo AI experience optimization. Welcome to the show, Jonathan, Neil, Mary and Ellie.

230
00:57:44.100 –> 00:57:45.329
Julia Nimchinski: How are you doing

231
00:57:45.980 –> 00:57:46.330
Jonathan Kvarfordt: Age.

232
00:57:46.330 –> 00:57:46.950
Marie Haynes: Hello!

233
00:57:47.570 –> 00:57:49.189
Eli Schwartz (ProductLedseo.com): To be here. Thank you for having me

234
00:57:50.350 –> 00:57:52.369
Julia Nimchinski: Let’s do a quick round of introductions

235
00:57:53.660 –> 00:57:55.139
Jonathan Kvarfordt: Oh, ladies first, st please

236
00:57:55.810 –> 00:58:25.559
Marie Haynes: Sure I’m Marie Haynes. I used to be a veterinarian kind of a weird thing for a vet to be talking to you about. AI. But I absolutely love learning. What Google is doing have been full time in SEO since 2012, and when Chatgpt came out in 2022, it was all at the same time that I was trying to learn more about Google’s machine learning systems. And so today, I find myself studying everything I can about AI, and how we can help businesses with it.

237
00:58:27.690 –> 00:58:51.539
Eli Schwartz (ProductLedseo.com): Eli Schwartz. I am the author of product led SEO. I. I work with companies to build SEO strategies that are focused on users and revenue rather than keyword, centric and product led SEO is experiencing a ton of interest. Now that AI is taking away many of the tactics that used to work just a couple of years ago. So excited to dive into this conversation

238
00:58:52.960 –> 00:59:06.020
Neil’s iPad: My name is Neil Patel. I’m the co-founder of a global ad agency called Np, digital we heard we help companies of all sizes grow through just getting more traffic on all the major platforms and drive more revenue

239
00:59:07.650 –> 00:59:33.299
Jonathan Kvarfordt: Love that my name is Jonathan Carford. Some people call me Coach K. And I’m super excited to be here with my bald brother Neil and Eli and Marie so good to see you all, and they dive in. I’m currently the head of growth at momentum, and I’ve also been in the Gtmai world for a long time, and this is a topic I’m super passionate about, and want to learn from you all. So I’m excited. So should we dig in, Julia? Should we keep going? It’s all right. Let’s do it all right. So I say some questions before. But I’m really curious on.

240
00:59:33.440 –> 00:59:41.429
Jonathan Kvarfordt: I’m asking a question that I did not prep before, which is, what are you most excited about in the next 6 months? 9 months this year with AI. Let’s go with Marie Eli, and then Neil

241
00:59:41.810 –> 01:00:06.020
Marie Haynes: Oh, gosh! The most excited is probably agents, and there’s so much we don’t know about agents, but I think it’s gonna radically change how the web works and much of what we talk about even today will be outdated by the end of the year, because, things are gonna change. So much. So I’m excited about that. I’m also a little bit nervous about that. But yeah, there’s so much opportunity that’s on the horizon

242
01:00:07.050 –> 01:00:09.370
Jonathan Kvarfordt: Love, that when they give

243
01:00:09.370 –> 01:00:32.609
Eli Schwartz (ProductLedseo.com): I’m really excited by AI in general, and people just getting smarter because they get more information. Many people are not good at the skill of searching for things online, and AI allows you to be bad at it. So now, yes, there’s a lot of hallucination and misinformation. But you’re also not typing in like, I love watching my kids and parents and friends who aren’t good at it, typing things in and seeing the information they get.

244
01:00:32.740 –> 01:00:47.809
Eli Schwartz (ProductLedseo.com): And AI allows them to type the things that they just want to do, and it it learns from them what their intent is, and from on a business side. I think it’s gonna blow up all old SEO. And and like, we’re just gonna have to be more user focused. So those 2 tie together

245
01:00:48.520 –> 01:00:48.840
Neil’s iPad: Yeah.

246
01:00:49.160 –> 01:00:49.639
Jonathan Kvarfordt: Everything, now

247
01:00:49.640 –> 01:00:56.149
Neil’s iPad: The thing thing I’m looking forward to is data and analytics. Ai just makes it much easier if you look at

248
01:00:56.150 –> 01:01:20.119
Neil’s iPad: ad spend, there’s a ton of ad spend being spent globally, but depending on the source. 60 to 80 ish 1 billion dollars is wasted just on digital ad, spend, right? So imagine how much is wasted on all forms of marketing, whether it’s SEO digital ad spend right people targeting the wrong keywords on SEO, showcasing the wrong information or social media campaigns, just losing money.

249
01:01:20.120 –> 01:01:25.640
Neil’s iPad: Imagine in real time getting better and better help of

250
01:01:25.720 –> 01:01:38.989
Neil’s iPad: AI. Just analyzing your campaigns and telling you what to spend more money on what? Not to, because everyone claims they look at their analytics. But if you actually look at how many corporations make adjustments based on the data. It’s very slim.

251
01:01:39.770 –> 01:01:51.720
Jonathan Kvarfordt: Yeah, one thing I love about yourself, anil is just all the data and analytics you guys share on on your zip and socials. It’s super insightful. And going to your point, Eli, about intent, I want to ask the question I had beforehand. I sent you before around

252
01:01:52.770 –> 01:01:58.219
Jonathan Kvarfordt: the whole intent matching. So I had a friend of mine in a SEO agency works very closely with Google. Talk about how

253
01:01:58.820 –> 01:02:20.780
Jonathan Kvarfordt: one of the things that is shifting is like the intent based search. So AI has to understand the intent of the user and then match that with the intent of a content type, and how that shifts how you think about it from a keyword based structure or SEO keywords. So I’m curious. Let’s start with. I’ll start with Neil, then, Eli, and go to Marie. What’s your take on that? How do you operationalize, you know, Matching intent to what someone may be searching for with AI

254
01:02:21.580 –> 01:02:42.680
Neil’s iPad: Yeah. So it’s actually very similar to traditional SEO, when someone types in a keyword, they’re going to a page and looking for a solution to their problem. What is the solution to their problem? Well, it depends on the intent, right, like, what do they really mean? From that search? And the same thing goes with AI. You got to figure out

255
01:02:42.680 –> 01:03:01.699
Neil’s iPad: how to create information and put it out there. That solves literally all the edge cases to give people what they’re looking for. And a great example of this is, if you ask a person what is the fastest animal? In most cases, they’ll say, hey? It’s a cheetah. Well, is a cheetah the fastest animal.

256
01:03:01.750 –> 01:03:26.799
Neil’s iPad: Are you? Including air animals, sea animals? Right? The cheetah is faster than sea animals, but there’s birds that go up to 200 plus miles per hour, and AI is much better at understanding all these edge cases, because each person’s intent is going to be different, based on their own personal experiences and what they’re looking for. So you have to focus on creating content that really solves

257
01:03:26.800 –> 01:03:33.010
Neil’s iPad: all the edge cases to do better and be there, no matter what their quote. Unquote, original purpose of that search is

258
01:03:33.440 –> 01:03:34.000
Jonathan Kvarfordt: Yeah.

259
01:03:34.890 –> 01:03:35.900
Jonathan Kvarfordt: Anything. Eli.

260
01:03:37.430 –> 01:03:53.710
Eli Schwartz (ProductLedseo.com): I think that the best thing about AI is that it blows up the entire concept of like, what keywords should I be focused on. And what are the keywords that get the most searches? Because keywords were a really bad representation of what people were looking for? We were forced into

261
01:03:53.920 –> 01:04:08.500
Eli Schwartz (ProductLedseo.com): typing the words we wanted when really we wanted concepts and services and products. And AI allows us to have that. So you know, a common question I’ll get is like, how do I rank higher in AI like, what keywords do I use?

262
01:04:08.730 –> 01:04:20.230
Eli Schwartz (ProductLedseo.com): And that’s not even a thing like if you think about how a how Llms even work is. They read content. They understand the content. You don’t actually have to use the word in content

263
01:04:20.370 –> 01:04:44.830
Eli Schwartz (ProductLedseo.com): to rank on it because AI understands it, and that’s something. Google has been getting better at it. But AI just skips ahead, and you no longer need to have it. So if you really want to do well in AI and really understand what users are looking for, you have to be that user. You have to switch to the other side of the table and feel that user’s pain and not like they’re looking for something because they were forced into typing 2 to 3 or 4 keywords into a search browser.

264
01:04:44.830 –> 01:04:54.970
Eli Schwartz (ProductLedseo.com): They’re typing entire concepts. They’re explaining what they’re looking for. So if you are not what they’re looking for, you should not be in that AI response. And when it came to SEO,

265
01:04:54.970 –> 01:05:00.070
Eli Schwartz (ProductLedseo.com): people would rank on things they shouldn’t have ranked on, because that was the way the algorithms were set up

266
01:05:01.540 –> 01:05:03.620
Jonathan Kvarfordt: Interesting, Dr. Marie

267
01:05:03.830 –> 01:05:31.020
Marie Haynes: So a few years ago, Frederick Debu from Bing put out. He was interviewed, and he said something really important. He said, seos need to switch from keyword research to intent research. And at the time it just kind of we didn’t pay a whole lot of attention to it. But it’s so much more important in the age of AI. Now, in 2018, when Google released Bert, you probably remember, Bert, I didn’t realize that Bert is a language model.

268
01:05:31.020 –> 01:05:58.049
Marie Haynes: and when Google announced it, they said it was in a blog post where they said, understanding keywords or no understanding searches better than before. And Google’s goal as a search engine has not been to match pages with keywords. It’s been to understand, what is the searcher really looking for? So here’s something that I love to do, whether we’re trying to be seen in language model responses, or whether we’re trying to rank better on search

269
01:05:58.050 –> 01:06:04.890
Marie Haynes: is to use a language model like, I really like to use Gemini, the thinking models of Gemini.

270
01:06:04.890 –> 01:06:29.719
Marie Haynes: And instead of doing keyword research, I’ll have this page. And I’ll say to Gemini. What is the main intent of a searcher who lands on this page? And what are 3 related micro intents for that searcher? And then I want to make sure that my page meets those intents, not only meets them, but is really easy to find the answer to each of those, and that’s far better than just trying to get the keyword on the page certain number

271
01:06:29.720 –> 01:06:51.090
Marie Haynes: of times, because whether you’re a search engine or whether you’re chatgpt or some other AI based tool that’s trying to get people a good response. What the tool is trying to do is understand? Like, what does this user really want and what content actually gives them the answer that they want. So understanding, intent is everything in today’s marketing.

272
01:06:51.770 –> 01:06:56.770
Jonathan Kvarfordt: I’m going to share this a little bit because I go. It’s attached to this intent concept. But I’m curious on

273
01:06:57.010 –> 01:07:21.530
Jonathan Kvarfordt: making content, or I should say, file types that are specific to AI not necessary for humans. So like one of the questions I had, I think we’ll start with you, Dr. Marie is, what’s the best way to structure content. So AI crawlers or AI search engines can parse it or understand it? So is it Api accessible documentation, is it Json Endpoints? Is it machine readable assets like, is there things we need to do differently on the website? So it’s AI friendly versus HTML, coding

274
01:07:21.530 –> 01:07:46.479
Marie Haynes: I think it depends on what type of content you have. So if you have an e-commerce site, then schema is super important, you want both AI and search engines to be able to understand all of the content, the specific things about the dimensions, the things that users expect to find for the vast majority of content. The schema, the markup is not as important as the words on the page.

275
01:07:46.480 –> 01:08:14.280
Marie Haynes: So you need to use language that is easy for machines to parse. And so that means being direct, often concise, in your answer. This is why sometimes we see that AI generated content can rank really well at first, st because it looks good to machines content that’s written by machines, looks good and is easy to understand by machines. And then there’s another factor to that, though, as well with either search or with AI,

276
01:08:14.280 –> 01:08:32.879
Marie Haynes: is that those systems are constantly refining so that they’re not just trying to find content that looks good to machines, but content that people like as well. And so you have to find the balance of something that is written in a way that it’s easy for the machine to understand what our point is.

277
01:08:32.880 –> 01:08:57.180
Marie Haynes: and also something that humans like, because humans humans often can tell when something’s written by AI, and and get turned off by that. And so we have to find that balance. But and then there’s there’s all sorts of controversy over whether schema, you know, does Schema Markup help you? And it turns out that like, if you’re looking at language models, they actually, when they’re training on your content, strip away all of the markup

278
01:08:57.180 –> 01:09:17.060
Marie Haynes: and convert the text into tokens numbered tokens. So I, personally don’t think that schema matters so much unless you have again e-commerce or something very specific. But if you’re looking at just regular content, and you’re trying to meet user intent, then I like to use

279
01:09:17.060 –> 01:09:32.279
Marie Haynes: phrasing that makes it really clear something that you know, in the days of Google using featured snippets, I mean, they still do. But when featured snippets 1st came out we wanted to create content, that Google would be happy to to use as a concise answer in a featured snippet

280
01:09:33.210 –> 01:09:36.180
Jonathan Kvarfordt: Totally okay, Neil. I know you got data on this one. What do you think

281
01:09:36.985 –> 01:09:59.990
Neil’s iPad: She’s spot on. So we don’t necessarily look at it as like the file type. It’s actually more the text on the page. If you look at these models out there, they’re crawling the web and gathering information just like a search engine would right. Some of them have partnerships for data feeds. But it’s more about the text on the page. And what is the quality of the text on the page?

282
01:09:59.990 –> 01:10:06.880
Neil’s iPad: It’s not about an exact keyword. And are you trying to optimize for that? It’s more so. Think about a topic.

283
01:10:06.880 –> 01:10:18.160
Neil’s iPad: Can you cover it very thoroughly? And when I mean thoroughly, I’m not talking about word count. Back in the day SEO used to be about. Here are the 5 keywords I’m going for on this page, and I have a 5,000 word article.

284
01:10:18.160 –> 01:10:38.757
Neil’s iPad: It’s more about. Here’s the topic. I’m gonna cover it very thoroughly. So I can match the intent of the user. But I want to be concise with my answer, and give it to them as quick as possible, but concise enough where they can get it fast, but they still don’t have any more questions around that topic, right? Because if you’re giving

285
01:10:39.220 –> 01:10:50.939
Neil’s iPad: forward response and it doesn’t solve what the user is looking for, it doesn’t matter. That’s too concise. But at the same time people don’t need 2,000 words when they could have got it in 2 sentences

286
01:10:51.440 –> 01:10:53.280
Jonathan Kvarfordt: Yeah, I get it. What do you think, Eli?

287
01:10:53.580 –> 01:11:16.380
Eli Schwartz (ProductLedseo.com): Yeah, I think technical SEO for AI is the same as it’s ever been. One of the biggest challenges that Google solved 25 years ago was crawling the web faster, being able to do it cheaper than every one of their competitors. And that’s why they survived. It wasn’t just the ranking algorithm. It was the fact that they had the most complete index. And that’s not something that any of these new AI products have solved better than Google.

288
01:11:16.380 –> 01:11:30.700
Eli Schwartz (ProductLedseo.com): So if you want to be crawled by any of these crawlers, and they’re all using crawlers. You need to follow those same best practices of having server side rendered content of having the content be cheap for the crawler to render when it. There’s scripts, and there’s complex things on the page.

289
01:11:30.800 –> 01:11:52.360
Eli Schwartz (ProductLedseo.com): It is expensive, and they will deprioritize it. Some people call that crawl budget, but it’s really just the crawlers looking out for their own expenses. So those best practices still exist. Schema helps make it cheaper because they can. Again, if they trust you, they can instantly recognize what it is. Is it e-commerce? Is it a service. Is it a product? So all those best practices remain the same.

290
01:11:52.660 –> 01:12:15.690
Eli Schwartz (ProductLedseo.com): The only thing these these Lms are doing different are rendering responses differently. But the way they’re gobbling up. The Internet is the same. And if they have more resources. And right now there’s a ton of money, they can do that faster and more frequently as they’d run out of money. That’s when they might do things like, I’ll only visit this website once a month or once a year. So the frequency and the freshness of the content matter

291
01:12:15.690 –> 01:12:26.609
Eli Schwartz (ProductLedseo.com): more for the engine than for the website. So you want to be accurate and accurate to the point that when, if the crawler doesn’t come back and this gets complicated, when you’re of a let’s say sas pricing.

292
01:12:26.610 –> 01:12:29.329
Eli Schwartz (ProductLedseo.com): So if you have sas pricing and you update it frequently.

293
01:12:29.330 –> 01:12:41.679
Eli Schwartz (ProductLedseo.com): The Llm. Won’t necessarily have that updated pricing. If it has not come back. Google does a really good job of knowing when to come back. And the Llms have to solve that. If it’s not, Google has to solve that just like Google has solved it

294
01:12:43.030 –> 01:13:06.899
Jonathan Kvarfordt: Yeah, I’m loving this, you guys, I need to have your brains just dump it into my brain. So I can get this and actually execute on it. Neil, I’m gonna ask you, I’m ask you each one question and have you kind of dive into some of your hopefully specialties for you is about predictive search. So you talk a lot about that in social. I’ve seen recently, Neil, a lot of your brands are always about, how do you predict user intent or user, predictively, what they’re searching for.

295
01:13:07.030 –> 01:13:08.870
Jonathan Kvarfordt: So how

296
01:13:09.080 –> 01:13:18.970
Jonathan Kvarfordt: how should we think about seos when it has AI driven predictive search? Or how do we make sure we’re hitting on what AI is proactively trying to find for someone to capture their? Their question

297
01:13:20.110 –> 01:13:42.330
Neil’s iPad: As a business. When it comes to SEO, you just got to be thorough in your marketing, and cover a topic in every single way as well as hopefully have a product or service that solves people’s problem. But when you’re looking at the predictive side, let’s use Google, for example, because that one that’s 1 that everyone can relate to. Most people may have an iphone or android device. If you have an android device.

298
01:13:42.600 –> 01:13:44.070
Neil’s iPad: they have data on you

299
01:13:44.220 –> 01:13:55.079
Neil’s iPad: if you’re using Gmail, even if it’s on an apple phone or chrome, they have data on you sites, running Google analytics. They have more data, right? Your search history. They have more and more data.

300
01:13:55.080 –> 01:14:17.750
Neil’s iPad: So when you think about this, they’re they’re eventually have this big repository of information which they can use to better understand what you, John, or Eli, or Marie, or Neil, or whoever it may be, is specifically looking for, and show them that now, as a website, you may say, Hey, I want to show up for all these people, and I want to capture it.

301
01:14:17.750 –> 01:14:40.020
Neil’s iPad: That is very, very tough, because unless your product or service hits all demographics. You know, people in different regions and all those little different things you’re not going to show up for everyone, and you just have to be okay with that. But the more applicable your product is for different people. Different price points, different regions, etc. You’re much more likely to show up.

302
01:14:40.020 –> 01:14:57.439
Neil’s iPad: But the world of search is going to change, in which, instead of 10 blue links which has already gone away from it’s more so going to be, every person eventually is going to have a unique experience. And they’re going to showcase information for that individual. But what we’re seeing is these platforms trust brands

303
01:14:57.620 –> 01:15:02.450
Neil’s iPad: more than people who don’t have a brand. And when I say people, I’m more see mean companies.

304
01:15:02.700 –> 01:15:28.689
Neil’s iPad: because one thing that’s been consistent is, they believe that brands are less likely to put out false information on the web compared to companies and individuals who don’t have brands. So if you want to do well in this new world. It’s really important to focus on building up your brand individually corporation. Right? It’s gonna you’re what you’ll see is you have a higher likelihood of showing up

305
01:15:30.370 –> 01:15:44.320
Jonathan Kvarfordt: Can you? Can we double click into that for a second? Because conceptually, I understand what you mean. But how does someone actually do that. How do you shift from being just a company to a brand in the AI world besides just pumping out a ton of AI generated content from chat, gpt like, what’s the concept you have? People focus on

306
01:15:44.890 –> 01:16:12.689
Neil’s iPad: So the 1st 2 are good product and good service. If you look at Air Jordans, for example, everyone knows the brand. They’ve seen it. They google it. They generate more sales from people knowing it than someone searching the keyword. You know. Basketball shoes, for example, has one of the biggest brands out there. So the 1st thing is good product or good service, because that gets your brand out there, and it creates word of mouth marketing just like how most people you know on this call right now don’t live in China.

307
01:16:12.690 –> 01:16:22.830
Neil’s iPad: or probably almost everyone doesn’t live in China. And we all know about the wowie threefold phone, right? Cause the brand got out there created a ton of press and buzz.

308
01:16:22.830 –> 01:16:26.119
Neil’s iPad: So good product, good service 2 is being Omni channel.

309
01:16:26.120 –> 01:16:49.740
Neil’s iPad: So, being on social media, the average person uses 6.7 social networks per month, according to sprout social, where they’re actually logging in each month into at least 6.7 social networks. The next thing is is, do general marketing SEO paid ads, etc. So when you start doing all this stuff, your brand’s getting out there. Your corporate brand and Llms. Notice, hey? How many people are

310
01:16:49.740 –> 01:17:02.409
Neil’s iPad: talking about this brand on the web, is it law? Is it? Little is the sentiment positive, because they’re much better actually understanding what people are saying versus what they’re not. These things impact how likely you are to show up in the long run.

311
01:17:02.830 –> 01:17:07.199
Jonathan Kvarfordt: Thank you, Neil. I love that, Dr. Murray. I have a question for you. Are you ready? Batter up?

312
01:17:07.669 –> 01:17:20.260
Jonathan Kvarfordt: So if an AI model misclassifies a site as low quality. How can a company correct its AI reputation, or is there new methods besides just traditional SEO tactics to be able to regain lost reputation

313
01:17:20.920 –> 01:17:23.560
Marie Haynes: I think there’s a couple things in this question, because

314
01:17:24.210 –> 01:17:34.680
Marie Haynes: sometimes an AI model well, we all know that they can get things wrong, that they hallucinate from time to time in terms of your reputation, though I don’t know

315
01:17:35.950 –> 01:17:55.299
Marie Haynes: that always happens. Actually, let me backtrack on this because I reviewed a site recently, and they’re very well known. They have a really good reputation, and when I asked Chatgpt and Gemini tell me what you know about this company. They both brought up this one bad review. They’ve had

316
01:17:55.530 –> 01:18:17.990
Marie Haynes: thousands and thousands of really good reviews, except that one bad review came up from just one website. And so what do we do? I think this is nothing really has changed in terms of what we do for SEO. That’s kind of reputation management. And so if there is negative information on a language model, it usually is because it’s

317
01:18:17.990 –> 01:18:42.939
Marie Haynes: it’s been fed that it’s got it off of the web. I remember when Chatgpt 1st came out there was somebody who was saying to me, Well, it’s awful! It doesn’t know anything. It keeps telling people these horrible lies about me. And when I looked up the horrible lies they were all over the Internet. So it wasn’t the language models fault it was that the reputation across the Internet was not good. And so I think and I’ll what you asked Neil. I love

318
01:18:42.940 –> 01:19:06.170
Marie Haynes: your answer, Neil. I think that Neil’s spot on that brand being known for something is so important. If you look at Google’s quality rater guidelines, which is what they try to model their ranking systems after and what their machine learning systems are trying to replicate. The word reputation is in there many times, and that doesn’t just mean your star rating. It means, what are you known, for.

319
01:19:06.170 –> 01:19:33.290
Marie Haynes: like the people who are on this panel, we all have a reputation for being known for understanding SEO. Otherwise we wouldn’t be on this panel. And so if you have a product, a website, something that you want to get found on search or on language models, the key is to find a way to get known, and that might be Pr, it might be advertising. It might also be just like creating something awesome that nobody’s ever created before

320
01:19:33.290 –> 01:19:48.439
Marie Haynes: and producing original stuff, and then, once it starts to spread, you become known for that. So I think when it comes to things being wrong on language models some of it. We can’t fix some of it. We can fix with the words that we put on our website.

321
01:19:48.440 –> 01:20:12.639
Marie Haynes: you know. If you ask if I ask Chatgpt for information about me, it pulls stuff from my about page, from my homepage, and so I have some control over what language models say about me. And then, if there’s something inaccurate, I think what I would do is ask the language model. Where did you get that information from? Why do you think that? And then you need to trace back? Was it bad training?

322
01:20:12.640 –> 01:20:18.819
Marie Haynes: Do I need to flood the web with accurate information, so that this inaccurate stuff doesn’t show up

323
01:20:20.120 –> 01:20:23.560
Jonathan Kvarfordt: Oh, this is! This is good golden stuff, Kayla, you ready? You’re up next

324
01:20:23.560 –> 01:20:23.960
Eli Schwartz (ProductLedseo.com): Right.

325
01:20:24.000 –> 01:20:36.910
Jonathan Kvarfordt: So AI models are bypassing traditional search ranking factors and pulling direct answers. So how does product led SEO strategy involve in a world where search engines are shifting away from ranking, and towards more AI generated summaries

326
01:20:37.490 –> 01:21:01.510
Eli Schwartz (ProductLedseo.com): So product led SEO. I mean, it was not so. I didn’t predict AI, but it aligns perfectly with the AI future, because product led SEO is really all about building your product for the SEO searcher. So if you’re not doing product led SEO, you’re going to a keyword research tool and figuring out what you think your user would search for without really understanding your user, without deep diving into the persona. And the pain points your user.

327
01:21:01.510 –> 01:21:14.699
Eli Schwartz (ProductLedseo.com): And one of the companies that prompted me to even start writing product. Let SEO was a friend of mine had started an insurance company, and most of the insurance companies have been around since. Long before Google, like their websites, were registered long before Google.

328
01:21:15.090 –> 01:21:35.829
Eli Schwartz (ProductLedseo.com): So this company wanted to rank for insurance, and they looked at all the search volume around the keyword insurance, and that was not something they were going to rank for. But when they did paid marketing, when they did their Facebook ads. They somehow attracted people that had complicated insurance problems like past duis or felonies or complicated things, and that was their niche.

329
01:21:35.830 –> 01:21:56.790
Eli Schwartz (ProductLedseo.com): But when it came to SEO. All they did was pump, search, car insurance content out there car insurance plus zip code car insurance plus city. And it sort of worked, but then actually got they got penalized because they overdid it. But product led SEO. If they would have done that, would have had them building products around and products in that case would have been search pages around what they served.

330
01:21:56.790 –> 01:22:16.640
Eli Schwartz (ProductLedseo.com): What can you expect to pay if you have these bad things in your history. So product. Let SEO aligns with an AI future, because now AI either knows who you are or you tell them, and you say, Hey, I am a, you know, a 25 year old, and I don’t like to pay my bills, and I have really bad credit score. What’s the best car Insurance company for me. Where do I find that car insurance?

331
01:22:17.210 –> 01:22:30.810
Eli Schwartz (ProductLedseo.com): That’s what product led. SEO would be. You have created that content for that particular SEO searcher. Because you understand that, SEO searcher rather than you created a piece of content because it had high monthly search volume. So

332
01:22:30.910 –> 01:23:00.599
Eli Schwartz (ProductLedseo.com): it’s not that you’re doing, SEO. And you’re doing your traditional SEO. You’re understanding your users. Whenever I start a consulting project, it’s really about asking the client who are their users. What did they have on their pitch deck that convinced a company or a Vc. To partner with them and give them a ton of money when they partner with companies. What are they special at? Why do they exist? That should be what is on their SEO not is what has high monthly searches in their, you know, broader category

333
01:23:01.970 –> 01:23:08.890
Jonathan Kvarfordt: I’m going to ask one last question that you all have a 3 have a chance to answer, so I’ll start with Dr. Marie. Then, Neil, and then Eli.

334
01:23:08.900 –> 01:23:33.630
Jonathan Kvarfordt: But question is this a lot of people listening? Are are marketers for go to market teams from the hard skills, hard skills, Exchange group. And so they’re trying to obviously capture more customers. So if you could give someone, whether in a series A To series D or even public facing company of like, what do you do right now with AI to make sure your website is performing correctly and getting people in there that are the right person for the right problem. What do they do right now? What’s the actual step? You’d have them? Do

335
01:23:33.940 –> 01:23:34.740
Jonathan Kvarfordt: we think

336
01:23:35.610 –> 01:24:00.770
Marie Haynes: I think the 1st thing is to just learn all you can about how AI is changing things. I think we haven’t even touched on the way. Search is going to change in the next couple of years. It’s kind of like if we had been around in the time when electricity 1st started lighting up light bulbs. There’s no way we could anticipate what we’re doing today on a computer screen. And so it’s very hard for businesses to prepare.

337
01:24:00.770 –> 01:24:25.690
Marie Haynes: So if I had a business with a bunch of staff, what I would do is have at least one person who’s in charge of staying on top of the AI things that are happening. And then I would have regular meetings to just brainstorm on how our companies can use this information. I’d look into agents. We don’t have time to talk thoroughly about what agents are going to be and do. But it’s going to radically change how the web

338
01:24:25.690 –> 01:24:50.619
Marie Haynes: works and change what websites are necessary, and how websites are used by businesses. And I would look into more in regards to images. We’ve seen Google lens being used more. And project Astra from Google is coming out in glasses that will radically change. Mark Zuckerberg predicted that smartphones will be no more very soon that

339
01:24:50.620 –> 01:25:07.180
Marie Haynes: we’ll all just basically be conversing with the web through smart devices like glasses. It’s hard to grasp, so I could go on and on. But I would say, just keep learning about how AI is changing the world and be ready to experiment and try new things

340
01:25:08.760 –> 01:25:29.960
Neil’s iPad: Quick, because I know we’re running out of time. But 1st off searching the land has a really great piece. I forgot who published it on there. But there’s a direct correlation with ranking on Google and ranking on Bing and getting mentioned by a lot of these AI platforms. So the 1st off is, make sure you follow traditional SEO best practices.

341
01:25:29.960 –> 01:25:38.920
Neil’s iPad: The second thing that I would recommend is big corporations, or even small corporations, stop posting a ton of crap search engine journal

342
01:25:38.920 –> 01:26:07.759
Neil’s iPad: had, and excuse my language. The Search Engine Journal had an article that broke down roughly. It was either 4.6 or 4.8 billion pieces of content are being put out and created each and every single day. That’s a lot with AI. You can expect that number to increase. People don’t want more information. What they want is quality, unique, amazing information that helps them solve their problem. And as a organization, if you can focus on quality like really, really high end quality that people haven’t seen before.

343
01:26:08.030 –> 01:26:21.420
Neil’s iPad: you’re likely to eventually start climbing to the top of any platform because these days searches everywhere, because even Instagram, Facebook, etc, they all have AI models that are powering their search. And their experience

344
01:26:22.440 –> 01:26:24.160
Jonathan Kvarfordt: Love that. Thank you, Neil. What do you think, Eli?

345
01:26:24.160 –> 01:26:45.149
Eli Schwartz (ProductLedseo.com): Mine might be a bit of a hot take, but I would just ignore AI. I think that businesses should focus on being a really good business and build a really good product. And brand because AI engines and word of mouth and every other channel that people find your businesses are just channels. So if you’re not a good product, if you don’t have your customers don’t love you, they’re not going to find you through those channels.

346
01:26:45.150 –> 01:27:03.829
Eli Schwartz (ProductLedseo.com): So if you focus on doing great by your customers creating loyalty, building a really good product, building a really good brand, you likely will show up in AI. And at that point, and we’re talking to new businesses. At that point you can tweak around the edges to figure out how to show up better. But 1st and foremost build a really good product, build a great brand, and do great by your customers

347
01:27:05.320 –> 01:27:17.749
Jonathan Kvarfordt: I can add my one thought to that is that the more and more I use AI because I use it literally all day, every day, the more I’m figuring out that it really is a amplifier of my own original thought, and if I can use it in that way to amplify my own

348
01:27:17.880 –> 01:27:31.370
Jonathan Kvarfordt: view of the world versus it, trying to replace my view of the world the better. And then it hits everything you guys are talking about where it’s about what I what can I bring to the market as an individual, as a brand? That’s unique to me, you know, that would serve the customer.

349
01:27:31.830 –> 01:27:32.969
Jonathan Kvarfordt: There you go, Julia.

350
01:27:33.850 –> 01:27:50.889
Julia Nimchinski: Much for bringing some clarity to the chaos every time I’m opening my Linkedin feed, especially on the SEO topic. It’s really hard to make sense. What’s actually happening? Where should we invest? Where? What’s going to happen? And yeah, how to optimize all of the content and websites and

351
01:27:51.801 –> 01:28:04.320
Julia Nimchinski: our community is already asking, is this recorded? So to continue learning from you, Eli, Neil, Mary Jonathan, where should our folks go? Newsletters? Yeah.

352
01:28:04.920 –> 01:28:06.070
Julia Nimchinski: do a quick one

353
01:28:06.680 –> 01:28:07.020
Jonathan Kvarfordt: Yeah.

354
01:28:07.020 –> 01:28:30.830
Marie Haynes: I’ll start, sure. So I have a newsletter where I publish way too much on what’s happening with AI and search every week. You can get that at Newsletter or mariehaines.com slash newsletter, and also a [email protected] is a bunch of really really amazing people who are just trying to be the leaders as AI changes our world

355
01:28:32.390 –> 01:28:37.099
Eli Schwartz (ProductLedseo.com): And mine is product. Let seo.com. That’s my newsletter, and of course check out my buckets on Amazon

356
01:28:37.870 –> 01:28:42.280
Neil’s iPad: Cool my ad agencies. Np, digital, and my blog is neilpatel.com

357
01:28:43.650 –> 01:28:50.589
Jonathan Kvarfordt: And I work for momentumio, which is agents for Gtm. Teams, and then gtmaiacademy.com nice and easy