YouTube2Text

Curso GEO 2026: Cómo hacer que la IA cite tu web — Transcript

by Pedro SEO · 6,708 words · 954 segments · language en · Watch on YouTube

Full transcript

  1. 0:00Hey there, everyone. I was thinking
  2. 0:01about whether or not to make this video
  3. 0:03, but I did it. So, Geo 2026, I'm going
  4. 0:05to explain my opinion on GEO, why you
  5. 0:08should accept it, and why you should
  6. 0:10take it into account in your marketing
  7. 0:13strategy. So, I know that there are
  8. 0:15many SEOs who, when they see the title
  9. 0:17of the video and so on, are going to
  10. 0:19come here just to criticize. At least
  11. 0:21watch the first part of the video, then
  12. 0:23if you want, you can stay until the end
  13. 0:25, but it all makes a lot of sense. Okay
  14. 0:28, first of all, I'd like to know who
  15. 0:29here supports Geo. You don't need to
  16. 0:33answer because I'm going to answer for
  17. 0:35everyone here, and I understand that
  18. 0:37you're with me if you do SEO, that GEO
  19. 0:39is obviously this here, no. SEO is geo,
  20. 0:43we all know that, including me. Okay,
  21. 0:46so I don't think so, there's any
  22. 0:48discussion here. In fact, John Müller
  23. 0:50already said that if we want to
  24. 0:52position ourselves in the AI mode
  25. 0:54overviews, we have to keep doing SEO as
  26. 0:57usual. So, we're talking about GEO
  27. 0:59still being SEO, it's all the same and
  28. 1:01that's it. So, Pedro, why are you
  29. 1:03creating a Geo course for me? Because,
  30. 1:06much to our regret, Geo exists.
  31. 1:08Generative Geo at JN Optimization
  32. 1:10basically means that when you ask an
  33. 1:12artificial intelligence something, that
  34. 1:15artificial intelligence quotes you,
  35. 1:17right? In other words, when you search
  36. 1:19for something on Google, the answer
  37. 1:20that comes up at the top will appear as
  38. 1:22your result, right? I'm going to
  39. 1:23explain how we've done it with several
  40. 1:25clients and others. I have quite a few
  41. 1:27real cases that I want to share with
  42. 1:29you, techniques and strategies; but
  43. 1:31it's true, I want to be transparent
  44. 1:33with you, that for me, the "o" in the
  45. 1:35word geo is what fails me. Because,
  46. 1:37after all, the user is using a
  47. 1:39generative engine, but the optimization
  48. 1:42part is very sketchy because it's not
  49. 1:44exactly known how it works, right? Then
  50. 1:47we'll break it down a bit and I'll
  51. 1:49explain it to you a bit more, but
  52. 1:50anyway, you should know that from my
  53. 1:52point of view, the O of optimization is
  54. 1:53the part that, well, we take it with a
  55. 1:55grain of salt. But basically, there's
  56. 1:57geo because now when you search on
  57. 1:59Google, you see this result, right? So
  58. 2:02this is Geo, let's call it whatever you
  59. 2:04want, but this is GEO. And our faithful
  60. 2:07friend Sander Pichai, CEO of Google
  61. 2:09since 2015, already said last year in
  62. 2:12May that this result, the AI overview,
  63. 2:14already has 1.5 billion searches per
  64. 2:16month or users who use this function
  65. 2:19per month. So it's a function that's
  66. 2:21going to stay, whether we like it or
  67. 2:22not. It makes perfect sense. After all,
  68. 2:25what Google has always pursued, what
  69. 2:27Google wants, is for you to find what
  70. 2:29you're looking for as quickly as
  71. 2:31possible. So, if you ask something and
  72. 2:33I'm going to give you the answer
  73. 2:35directly, you don't have to go to a web
  74. 2:37page, much better, right? We already
  75. 2:39experienced this with fragmented texts,
  76. 2:41right? The typical frequently asked
  77. 2:43questions that appeared up there in
  78. 2:45Google, which cited your website and so
  79. 2:46on, but now it is directly generated by
  80. 2:48artificial intelligence, so it is
  81. 2:50something that is going to stay, we are
  82. 2:52very clear about it. In fact, I have
  83. 2:54been spreading the message for years
  84. 2:55that this was going to happen. Here,
  85. 2:57for example, I have a quote in
  86. 2:59Entrepreneurs that explained that
  87. 3:00artificial intelligence will make many
  88. 3:02businesses cease to exist on Google. It
  89. 3:04is already happening. People use Chat
  90. 3:06GPT as a search engine instead of
  91. 3:08Google, right? It is already the
  92. 3:10world's fifth most visited site. So,
  93. 3:12here I am, well, I am doing it in
  94. 3:14several presentations, and in fact I
  95. 3:16coined the term SEO on control and SEO
  96. 3:18off control: an SEO that we could
  97. 3:20control and an SEO that we could not
  98. 3:22control. This SEO that we could not
  99. 3:24control is based on all this text
  100. 3:26generated by artificial intelligence
  101. 3:28and so on, which works in a completely
  102. 3:30different way. So getting back to the
  103. 3:33topic, for those of you who are still
  104. 3:36rejecting Geo, give me a moment to
  105. 3:38explain why we should accept it, mainly
  106. 3:41because there is a difference between
  107. 3:44SEO and GE. Basically, the goal of SEO
  108. 3:47is to appear in a search result. So,
  109. 3:50within the SERPs, SEO is links. GEO is
  110. 3:54what we call an AI-generated response,
  111. 3:56it is not a website link. So, that's a
  112. 3:59different story, right? The format is
  113. 4:01completely different. At the end of the
  114. 4:03day, within SEO, what we are doing is
  115. 4:05working on keywords. In GEO, it is more
  116. 4:07based on questions, conversational
  117. 4:09intent and so on, okay? In search
  118. 4:11intent, which is also often talked
  119. 4:13about within SEO. In short, SEO can
  120. 4:15work on its own. Geo depends on SEO. So
  121. 4:19, we can say that geo is essentially a
  122. 4:20part of SEO, if we want to call it that
  123. 4:22way. But obviously between you and me,
  124. 4:25regarding SEO, geo is still the meme
  125. 4:28that we are seeing here, okay? I mean,
  126. 4:31I already know that in order to do geo
  127. 4:33I have to do SEO, it's super expensive,
  128. 4:36I understand, but we SEOs should accept
  129. 4:38geo as a term. Why? Mainly because it
  130. 4:41exists. We're already seeing, right?,
  131. 4:42the results that come up at the top of
  132. 4:43Google. So, this is already the main
  133. 4:45reason as to why we really ought to
  134. 4:46accept it. Secondly, the client is
  135. 4:48going to search for it anyway. So the
  136. 4:51client is looking for GEO. Just this
  137. 4:53morning, it seems unbelievable, but
  138. 4:55this morning a potential On Lead client
  139. 4:56came to me talking about GEO. I can't
  140. 4:58believe it when they told me that,
  141. 5:00right? They told me, "No, what I want
  142. 5:02is to run Google Ads campaigns and also
  143. 5:03begin working on the GEO aspect." I
  144. 5:06said, "Okay, okay, the client is
  145. 5:07already looking for it, that's it, you
  146. 5:09know? If you have an agency that offers
  147. 5:11SEO services, you have to know that
  148. 5:13this is going to happen to you. And if
  149. 5:15you have any doubts, go to Google
  150. 5:16Trends and search for Geo Agency or
  151. 5:18enter it in the whole world. You'll see
  152. 5:20that it's pointing up, okay? So, people
  153. 5:22are already looking for Geo agencies.
  154. 5:24Don't ask me why this is happening, but
  155. 5:26it is happening. And if you go to tools
  156. 5:29like HRF or SEMRASH, you'll also see
  157. 5:31terms like Geo Agency, Geo Agency
  158. 5:33Alicante, Geo Agency, Geo Agency
  159. 5:35Barcelona and so on. They have few
  160. 5:37searches, yes, it's true, but they're
  161. 5:38starting to have searches. And at the
  162. 5:40end of the day, accepting is easier
  163. 5:42than competing. If you accept it, you
  164. 5:43don't have to give explanations. The
  165. 5:45deniers are the ones who have to
  166. 5:47justify that it doesn't exist. And
  167. 5:49ultimately, if you don't accept it,
  168. 5:51someone else is going to do it for you
  169. 5:52and you're going to get ahead of it.
  170. 5:54Also, being SEOs, Oh man, we're going
  171. 5:57to get a lot of businesses interested
  172. 5:59in SEO, since Geo equals SEO, so it's a
  173. 6:01much sexier way to sell it, because hey
  174. 6:03man, do you want Geo? Yeah, yeah, it's
  175. 6:06not like Geo is part of SEO, okay? But
  176. 6:08this way, if we SEOs accept it, we can
  177. 6:10create a better ecosystem within our
  178. 6:13service, within what they want, right?
  179. 6:15I mean, at the end of the day we know
  180. 6:18that it's goye, let's accept it and say
  181. 6:20, yes, if this is part of SEO, so and
  182. 6:22so on, right? Let's say, no, no, that's
  183. 6:23something else because such and such.
  184. 6:25Then the guy on duty is going to come
  185. 6:27along and with some trafficker-style
  186. 6:29term and he's going to take over that
  187. 6:31niche in the sector and take part of
  188. 6:33our work, right? So, hey, let's accept
  189. 6:36it, period and that's it. And the thing
  190. 6:38is, we've seen this before, for example
  191. 6:40, Sin Elis 'growth hacking, who
  192. 6:42invented it in 2010 and many SEOs and
  193. 6:44Marketers hated it because it wasn't
  194. 6:46anything new. Growth marketing of what?
  195. 6:48Or the inbound marketing that HSpot
  196. 6:50coined in 200526. So, it didn't
  197. 6:52describe anything new technically, did
  198. 6:54it? After all, it was content, it was
  199. 6:56SEO, it was email and so on, but with a
  200. 6:58different name, a different category.
  201. 7:00After all, we love coming up with new
  202. 7:02terms, right? So, hey, I don't care if
  203. 7:04you want to call it Geo, AO, SEO cona,
  204. 7:07whatever. I don't care. What we really
  205. 7:10want here when we do SEO is for a
  206. 7:13business to earn more. I say this
  207. 7:15because there are many SEOs, especially
  208. 7:17technical SEOs, who still think about
  209. 7:20HS1, metas, interlinking and other SEO
  210. 7:22strategies, a bit outdated because they
  211. 7:25are still the core, they are still the
  212. 7:27basis; but today SEO is much more than
  213. 7:29all that and, at the end of the day,
  214. 7:32when we do SEO, what we care about is
  215. 7:34the client's pocketbook, or so it
  216. 7:36should be. We shouldn't be fighting for
  217. 7:38a position number one with a keyword.
  218. 7:40No. we should fight to increase the
  219. 7:42organic traffic of a web page. When we
  220. 7:44do SEO, let's not forget, we fight to
  221. 7:47earn more money. It's that simple. So,
  222. 7:48what we want is for a business to earn
  223. 7:50more through a marketing strategy that
  224. 7:52is SEO, organic positioning, GEO, call
  225. 7:54it what you want, okay? But we should
  226. 7:57accept it. After all, it doesn't matter
  227. 7:59if the cat can catch the mouse. It's
  228. 8:01what I was saying a little earlier. The
  229. 8:02client wants sales, so it doesn't
  230. 8:04matter if you do it with Go, with SEO
  231. 8:06and so on. So, at the end of the day,
  232. 8:08the term is the least important thing,
  233. 8:10it's just that now people are searching
  234. 8:11and the result is up there. People
  235. 8:13wonder how to get up there. Well, we
  236. 8:15have the solution. It's done with GEO
  237. 8:17and, obviously, it's part of the
  238. 8:18foundation of SEO. And as I mentioned
  239. 8:20at the beginning, nobody knows the
  240. 8:22exact formula for GEO or SEO. Google
  241. 8:25does give us guidelines and tells us a
  242. 8:27bit about what to do to position within
  243. 8:29Google, but when we're talking about
  244. 8:31Geo, it's a much bigger deal. We're
  245. 8:34talking about artificial intelligence
  246. 8:35in neural networks. SEO ultimately
  247. 8:37works with algorithms, right? With
  248. 8:40Panda, Penguin and so on. Yes, the
  249. 8:42entire Google system is changing now.
  250. 8:44I've been saying for years that Google
  251. 8:46is increasingly implementing artificial
  252. 8:48intelligence in its search engine with
  253. 8:51neural networks like Rankbrain and so
  254. 8:53on. So, what's happening is that it's
  255. 8:55no longer a simple game of chance, it's
  256. 8:58no longer the ABC of SEO that we knew,
  257. 9:00that used to work. We would include a
  258. 9:02good heading structure, good meta tags,
  259. 9:04the keyword by shoehorn, throughout the
  260. 9:06content, a little link building if
  261. 9:08necessary and you would be positioned.
  262. 9:10Now the game has changed and those who
  263. 9:12are getting results are those who are
  264. 9:14understanding the rules of the game.
  265. 9:16And I'm telling you this firsthand
  266. 9:18because I've been struggling with a
  267. 9:19search engine like YouTube to do SEO.
  268. 9:22In fact, here you can see a video that
  269. 9:23I uploaded in 2018, the first video on
  270. 9:25my YouTube channel. If you go to that
  271. 9:26video and go to the learning section,
  272. 9:28I'll show you what I said here. is that
  273. 9:30one of the reasons for starting this
  274. 9:32YouTube channel is to study how to do
  275. 9:34SEO within the platform. And do you
  276. 9:36think that today I have the magic key
  277. 9:37to do SEO within YouTube? No, no. This
  278. 9:40is a jungle, it's wild with, let's say,
  279. 9:42algorithms that you have no idea what's
  280. 9:44going on in there. The same is
  281. 9:45happening with Google. So, if you're a
  282. 9:48Google SEO, you should know that you're
  283. 9:50going to have to get used to factors or
  284. 9:52sectors that you won't have any control
  285. 9:55over how to position yourself up there,
  286. 9:57okay? So, we've been fighting with this
  287. 10:00for years and we've really achieved it.
  288. 10:03Although I don't have the magic key to
  289. 10:04do SEO on YouTube, I've managed to
  290. 10:06position myself with the GPT chat
  291. 10:08course and even with just GPT chat. How
  292. 10:10to make ads on Instagram that have
  293. 10:12700,000 views or how to grow on
  294. 10:14Instagram, for example, which has
  295. 10:16another 600,000 views. Or this SEO
  296. 10:17course, for example, I uploaded it a
  297. 10:19year ago, it has almost 1 million views
  298. 10:21. So, I know Doing SEO within YouTube.
  299. 10:23Well, what do you want me to tell you?
  300. 10:24I wish I had the magic key and all my
  301. 10:26videos had these views. What I do know
  302. 10:29is that there are certain factors that
  303. 10:31help your videos rank on YouTube. And
  304. 10:34with Geo, I'm not going to tell you
  305. 10:35that we've discovered it, but we do
  306. 10:37have several success stories within the
  307. 10:38agency and that's what I want to share
  308. 10:40with you in this training. And I'm
  309. 10:42telling you that I'm not alone in all
  310. 10:43this. All this has been thanks to the
  311. 10:45Palo Seco team who are researching,
  312. 10:47testing, and testing every day, whether
  313. 10:50with our own internal projects or with
  314. 10:52several of our clients on how to
  315. 10:53position ourselves within the LLM and
  316. 10:55so on. And within the agency we've
  317. 10:58actually detected how clients are
  318. 11:00receiving traffic from sources like
  319. 11:02Chat GPT, which is reflected in sales.
  320. 11:05Well, we have several success stories
  321. 11:08that we document on our website and,
  322. 11:10thanks to all this, we're appearing in
  323. 11:12the media and at different events that
  324. 11:14call us to spread the message. So, if
  325. 11:16you're willing to accept Geo, update
  326. 11:19yourself, learn, Let's get started.
  327. 11:21Okay, let's get started. Let's get into
  328. 11:24it. First of all, before I explain how
  329. 11:26Geo works, I want you to understand how
  330. 11:29a modern search engine works, okay?
  331. 11:31We're talking about Geo here, we're
  332. 11:33talking about artificial intelligence,
  333. 11:35so understanding how these search
  334. 11:37engines work in 2026 is essential for
  335. 11:40you to understand the current landscape
  336. 11:42, okay? So I'm going to get a little
  337. 11:44technical, don't worry if you forget
  338. 11:46something, if you miss something, okay?
  339. 11:49You can leave a comment, I'll try to
  340. 11:50solve it. And if you join the community
  341. 11:52, which you can do for free from the
  342. 11:54link in the video description, you can
  343. 11:56ask whatever you want there. Okay,
  344. 11:58let's take it one step at a time. First
  345. 11:59of all, when you use a modern search
  346. 12:01engine, the first thing you do is ask
  347. 12:03that modern search engine, right? You
  348. 12:06write a query. A query would be like
  349. 12:08the prompt you do on Google, right?
  350. 12:11That would be the first thing we do.
  351. 12:13The user writes to the artificial
  352. 12:14intelligence. What happens here is that
  353. 12:17the artificial intelligence detects
  354. 12:19that you have written something and
  355. 12:21says," Hey, this person has an
  356. 12:23intention to do this. "Then it starts
  357. 12:25to break down the query that you have
  358. 12:27made. Once it understands what you want
  359. 12:30to find or what you want to search for,
  360. 12:33what it does is use different internal
  361. 12:35agents, okay? So, as you can see here,
  362. 12:38it has agents A, B, C, D. There are
  363. 12:40many more, obviously, but so that you
  364. 12:42understand, okay? So, what it does here
  365. 12:44is after this question it consults with
  366. 12:46all the agents that it has, okay? These
  367. 12:49agents are going to make different
  368. 12:51questions to the internet or to its
  369. 12:54database to gather information from
  370. 12:56what you have previously asked. It
  371. 12:59breaks all that down into different
  372. 13:00queries. So, what it does is, well, it
  373. 13:02asks one, question two, question three,
  374. 13:04question four and it will group all
  375. 13:06that information that it is going to
  376. 13:08gather, okay? Whether from its database
  377. 13:10or from the internet. And finally, once
  378. 13:12it has all that information, it does
  379. 13:14the query check. This means, hey, does
  380. 13:17all this make sense? Does it work? If
  381. 13:19it does, it moves on to the next one.
  382. 13:21step. If it doesn't work, it goes back
  383. 13:23here, quer modification, and it goes
  384. 13:25back up here again and loops again and
  385. 13:28asks itself again until it has a good
  386. 13:30answer and moves on to the next phase,
  387. 13:32which is the next phase that we all
  388. 13:34like and we all see that it is the
  389. 13:36final result, okay? Here it returns
  390. 13:39that magical result that we all
  391. 13:40hallucinate, but that would be a bit of
  392. 13:42the process of how a modern search
  393. 13:44engine like Chat GPT works. Okay, Pedro
  394. 13:47, why are you telling me all this? Okay
  395. 13:48, basically so that you understand how
  396. 13:51a modern search engine works today.
  397. 13:53Once it returns this magical result,
  398. 13:54the key is not in the magical result,
  399. 13:56which I also want to explain to you a
  400. 13:58bit about how it works, but rather in
  401. 14:00the queries that these agents make here
  402. 14:02, okay? These queries that these agents
  403. 14:05make are cool because what we can do
  404. 14:06with them is detect the questions that
  405. 14:08have been made on the internet. That is
  406. 14:11known as query funout. So, query
  407. 14:12funouts are those specific searches
  408. 14:14that these mini-agents, by calling them
  409. 14:16that somehow, have made within search
  410. 14:17engine. So, if you, for example, search
  411. 14:19for how to plan a trip in spring, what
  412. 14:21the AI is going to do is ask itself,
  413. 14:23well, hey, the weather in Japan in
  414. 14:25April and May, the best places to go to
  415. 14:27Bel Sakura, Japan Prima 2025 festivals,
  416. 14:29tips for Spanish speakers? I don't know
  417. 14:31how many of these questions it's going
  418. 14:32to ask itself and then, based on all of
  419. 14:34this, it returns the result that we all
  420. 14:36love and that we all freak out about.
  421. 14:37Okay? The good thing about querifouts
  422. 14:40is that if you're a fairly technical
  423. 14:42person, you can detect which keywords
  424. 14:45these agents have used on the internet
  425. 14:47to return that result, so you can then
  426. 14:50use these questions to use them on your
  427. 14:52website, for example, right? So, for me
  428. 14:55, this is key and fundamental for doing
  429. 14:58SEO or GEO in 2026, okay? Detecting how
  430. 15:01agents search and using these queries
  431. 15:03within your content plan, in your FAQs,
  432. 15:06and so on. And okay, now finally, I
  433. 15:08also want you to understand how
  434. 15:10artificial intelligence generates that
  435. 15:12content, okay? So, this is what I want
  436. 15:14to do. that you understand it too.
  437. 15:15After all, they work with four phases,
  438. 15:18tokens, chunks, in mendings and
  439. 15:20semantic search. Tokens, tokens
  440. 15:22ultimately represent the semantic unit
  441. 15:24that the model can convert into numbers
  442. 15:27, okay? After all, what artificial
  443. 15:29intelligence does is that if you search
  444. 15:32for digital marketing, boost online
  445. 15:34sales, what it will do is divide this
  446. 15:36into different words and each word will
  447. 15:39have a vector, okay? So, all this works
  448. 15:42in a vector way. Artificial
  449. 15:43intelligence, chat GPT, Gemini, Clot,
  450. 15:45all the others do not understand text,
  451. 15:48they understand vectors. So, everything
  452. 15:50works by vectors within artificial
  453. 15:52intelligence. And what they basically
  454. 15:54do is use chunks. Chunks are fragments
  455. 15:56of text divided into small blocks so
  456. 15:58that AI can process them and compare
  457. 16:00them to understand their meaning
  458. 16:02separately. For example, in this
  459. 16:04section here, which would be a text
  460. 16:05that we would see, what it does at the
  461. 16:07beginning, well, listen, is segment it
  462. 16:09into HTML, but it is not going to to
  463. 16:10take everything specifically. What it
  464. 16:12will take are different parts, because
  465. 16:14each chunk is made up of different
  466. 16:16tokens, so that is the maximum that it
  467. 16:18could collect. Then, with the endings,
  468. 16:20after all, it is still a numerical
  469. 16:22representation of a text and what it
  470. 16:24basically does is analyze the entire
  471. 16:26string of characters in the different
  472. 16:28numbers of vectors. So, as you can see
  473. 16:30here, for example, it has the three
  474. 16:32chunks, it has the entire string of
  475. 16:34vectors and then when it goes to the
  476. 16:35semantic search what it wants to do is
  477. 16:37the cosine similarity, okay? To make
  478. 16:39ourselves understood, if what we have
  479. 16:41here has a cosine similarity of 0.81,
  480. 16:43it means that this result is the most
  481. 16:45appropriate or the most semantically
  482. 16:47similar to the query that it was
  483. 16:49looking for. Then, it is the chosen
  484. 16:51result. After all, artificial
  485. 16:53intelligence understands numbers, it
  486. 16:55understands vectors and what it will do
  487. 16:57is that if it detects that this is the
  488. 16:58best result, it will return the result
  489. 17:00to you. For that same reason, many
  490. 17:02times you will see that it hallucinates
  491. 17:03and that a result totally hurts you.
  492. 17:05invented that makes no sense. Basically
  493. 17:07because he has based it on the
  494. 17:08similarity of the cosine and since he
  495. 17:09has seen that this was greater than the
  496. 17:11other, well that's what he has told you
  497. 17:12, okay? Since the model does not look
  498. 17:14for exact words, but semantic
  499. 17:16similarity of meaning. So, he doesn't
  500. 17:18care what you tell him, because he is
  501. 17:20going to analyze his numbers and give
  502. 17:21you the result that he is going to get.
  503. 17:23And for him, as I mentioned before, the
  504. 17:25greater similarity of the cosine is the
  505. 17:27most relevant result for him,
  506. 17:29regardless of whether it makes sense or
  507. 17:31not. And okay, I'll stop with the
  508. 17:32technical part, because I'm sure many
  509. 17:34people here will say," Hey, okay, yes,
  510. 17:36Pedro, and how do we do geo? "" Well,
  511. 17:38I'll explain it to you below. To
  512. 17:39enhance the geo, we have different ways
  513. 17:41of doing it. First of all, use content
  514. 17:44in question-answer format. LLMs are
  515. 17:47trained with millions of frequently
  516. 17:49asked questions, okay? In other words,
  517. 17:51questions and answers, questions and
  518. 17:52answers, questions and answers. In fact
  519. 17:54, we've been training them for years.
  520. 17:56Every time we have to complete one of
  521. 17:58these puzzles, it's not to check if
  522. 18:00you're a robot, it's simply to train
  523. 18:02their artificial intelligence. I hate
  524. 18:04to say it, but they don't care. What
  525. 18:06they want is to train those patterns to
  526. 18:08identify them and then make them work
  527. 18:10better. So, when we complete these
  528. 18:12puzzles, it's to make their artificial
  529. 18:14intelligences smarter. They are trained
  530. 18:17with millions of frequently asked
  531. 18:18questions, which is X, how does it work
  532. 18:20? And what they want is to understand
  533. 18:23the pattern so that the models
  534. 18:24recognize it and be a reliable source.
  535. 18:26Optimize the direct answer, right?,for
  536. 18:28the click. Okay? So, in the past, what
  537. 18:31SEO did was a lot of content, a lot of
  538. 18:34keywords. Now what Geo wants is for you
  539. 18:36to enter the clear, direct answer, to
  540. 18:38Just stick it in your first response
  541. 18:40and that's it, you know? I mean, you
  542. 18:43put it in the first paragraph, there's
  543. 18:45no need to explain everything else and
  544. 18:47then give the content, right? So,
  545. 18:49that's what used to be done with SEO.
  546. 18:52Unfortunately, we've ruined the
  547. 18:53internet a bit. Let's remember, right?,
  548. 18:55the gigantic texts that we put at the
  549. 18:57end of everything in product categories
  550. 18:59or something like that to do SEO, which
  551. 19:01still works today because Google still
  552. 19:03swallows it. But in the long run, when
  553. 19:06it gets smart and sees that all the
  554. 19:08text we put in there with keywords
  555. 19:10using semantics, okay, very good, and
  556. 19:12generated with it too, very good, but
  557. 19:14we know that over time all that won't
  558. 19:16work as well because what they want is
  559. 19:19data that you can analyze directly and,
  560. 19:21above all, that they can cite in the
  561. 19:23first response, okay? Okay, for example
  562. 19:25, in the case of zona F, a driving
  563. 19:27school that we've been managing for a
  564. 19:29long time, for example, well if You ask
  565. 19:31, "Can I drive motorcycles, trucks or
  566. 19:33vans with a B license?""Yes, with a B
  567. 19:36license you can drive vans and light
  568. 19:37trucks. And that's it. And then I'll
  569. 19:39explain a little more, right? From the
  570. 19:40moment you get your license, that's it.
  571. 19:41We explain this in the first sentence,
  572. 19:43you see? I mean, we don't ramble on or
  573. 19:45add more text here, we explain
  574. 19:46everything right away. Or for example,
  575. 19:48if you search for price, Barcelona car
  576. 19:50license, well the same. Here we have,
  577. 19:52look, these are the positions that we
  578. 19:54like, you know? You search and here you
  579. 19:56get price, getting the license, ta ta
  580. 19:57such. Here you get another player, okay
  581. 19:59? But then we both appear here. We also
  582. 20:01appear organically below, okay? So here
  583. 20:03we are taking up three spaces in the
  584. 20:05SERPs with a simple search and this is
  585. 20:07because it is taking our content, let's
  586. 20:10remember the chunks, the endings and so
  587. 20:12on. So, what it is going to do here is
  588. 20:14take a compilation of all these pages,
  589. 20:16including ours, and that's why we
  590. 20:18appear in the first results, because we
  591. 20:20are the most relevant answer to this
  592. 20:22type of question. That in all this, you
  593. 20:24know more and more that people are no
  594. 20:25longer going to search for the price of
  595. 20:27a car in Barcelona, they are going to
  596. 20:28start searching, hey, I want to get a
  597. 20:29driving license in Barcelona, I'm 18
  598. 20:31years old, how much does it cost? No,
  599. 20:32explain to me how the subject works. No
  600. 20:33, these are going to be much longer
  601. 20:35queries. So, now you know: act when you
  602. 20:37have to create that very clear, very
  603. 20:39direct content, and don't fight for
  604. 20:40articles of 2,000, 5,000 words, because
  605. 20:42really...well, it's not that it's not
  606. 20:44necessary, but it's that, hey, it works
  607. 20:46; but, depending on the case, if they
  608. 20:48are frequently asked questions, the
  609. 20:50answer should be as clear and direct as
  610. 20:52possible. The next thing, your own data
  611. 20:54and original statistics. LLMs, IAS, are
  612. 20:57always going to cite data that no one
  613. 21:00else has. So, we have to document
  614. 21:03everything that we do that makes us
  615. 21:05unique. Some of our own studies, a
  616. 21:07survey of our clients, some internal
  617. 21:10benchmark that we have, is much more
  618. 21:12citable than not compiling what others
  619. 21:14are saying. That's why AI-generated
  620. 21:16content doesn't work very well in
  621. 21:18certain sectors, because if you want to
  622. 21:20compete in certain markets, everything
  623. 21:22has already been invented, everything
  624. 21:24has been said. So, if you create
  625. 21:26content with IAP, it will be very
  626. 21:27difficult for you to differentiate
  627. 21:28yourself and position yourself or for
  628. 21:29you to be the chosen option. That's why
  629. 21:31the key here is always to make those
  630. 21:34lists, those studies and document them
  631. 21:36internally. This will have far more
  632. 21:38value than just doing something
  633. 21:39generated by consensus and that is all.
  634. 21:41After all, artificial intelligence has
  635. 21:43already moved on to the internet, it
  636. 21:45has already crawled all the text
  637. 21:47content that exists. The next step will
  638. 21:50be audiovisual content, so if you
  639. 21:52continue pursuing SEO natively, that is
  640. 21:54, with text, it will be much more
  641. 21:56difficult for you than if you don't
  642. 21:58implement audiovisual content within
  643. 22:01your content strategy. That's why at
  644. 22:03our agency in Palo Seco we are
  645. 22:05increasingly committed to video
  646. 22:07strategies within YouTube, where we can
  647. 22:09also do SEO and when you search within
  648. 22:12Google, the first results are already
  649. 22:14video. So, Second tip: start
  650. 22:16implementing your own studies or
  651. 22:18documentation that doesn't exist on the
  652. 22:21internet, and if you can include video
  653. 22:23content, all the better. The next point
  654. 22:26you need to keep in mind when doing geo
  655. 22:28is to use directories. LM models don't
  656. 22:31really think about which is the best
  657. 22:33fryer. Let's remember, they use tokens,
  658. 22:36chunks, and what they do is look at
  659. 22:38vectors, right? They see those numbers.
  660. 22:40So, when you search for which is the
  661. 22:42best air fryer, what it does is think
  662. 22:45about which is the best fryer, what it
  663. 22:47does is go to different rankings or
  664. 22:49different websites that talk about
  665. 22:52which are the best. So there you go,
  666. 22:54take that list. Okay. Now, before I
  667. 22:56show you an example, I'll explain it
  668. 22:57here, okay? So, if you search," I want
  669. 22:59to buy a mountain bike, I'm a beginner
  670. 23:02and I have a budget of € 500. What do
  671. 23:04you recommend for me to start with? ""
  672. 23:05What the AI does, you see? It gives you
  673. 23:07this result, but if you go to the
  674. 23:09sources, it will take you to some
  675. 23:11sources where it tells you, well: "the
  676. 23:13best mountain bikes", "the best
  677. 23:15mountain bikes for less than € 500",
  678. 23:17"the 10 best mountain bikes" in English
  679. 23:19, you see: *mountain bikes*. I mean, as
  680. 23:21you can see, they are all directories.
  681. 23:23If you go in here, you will see
  682. 23:25different ones, a list with different
  683. 23:27mountain bikes for less than € 500,
  684. 23:29right? Well, these directories are the
  685. 23:31ones that we have to implement in our
  686. 23:33content strategy. We SEOs do the same,
  687. 23:35those of us who have learned about all
  688. 23:36this and the top ones, we do, well, hey
  689. 23:38, a ranking of the seven best SEO
  690. 23:40agencies in Barcelona, okay? And here
  691. 23:42we do a comparative list with all the
  692. 23:43agencies, we include the agencies and
  693. 23:45so on. This article can be improved a
  694. 23:47lot from what I'm seeing. After all,
  695. 23:49it's doing this and obviously here we
  696. 23:52ally ourselves with our competitors.
  697. 23:54Well, hey, Peter Leit gives me a link
  698. 23:56to me and I put a link to Peter Leit
  699. 23:58and obviously we put it as do follow
  700. 24:00because unity is power, okay? So, the
  701. 24:02key here is going to be this: that it
  702. 24:04is a "pain", in quotes, to have to
  703. 24:06mention competitors. But for me this is
  704. 24:08key, because I have a competitor, Peter
  705. 24:10who is a player that I have in
  706. 24:11Barcelona and we compete for the same
  707. 24:13keywords, I have spoken with him many
  708. 24:15times, which is a bit like what we were
  709. 24:17talking about at the beginning about
  710. 24:19accepting geo or not. I mean, at the
  711. 24:21end of the day, Peter talking about SEO
  712. 24:23helps me because it makes the service
  713. 24:25known, it makes SEO known, it gives
  714. 24:27importance to SEO. So that I, Peter,
  715. 24:29also explain about Geo also helps you
  716. 24:31so that people demand the service more
  717. 24:33because I give importance to the sector
  718. 24:35, okay? Now, outside of all the content
  719. 24:37and so on, but simply so that you
  720. 24:39understand the philosophy that
  721. 24:41competition is very healthy and it is
  722. 24:43super good. In fact, I wish we had also
  723. 24:45compared ourselves with Peter and with
  724. 24:46some other competitors. prices, we have
  725. 24:48explained how to do everything, because
  726. 24:50at the end of the day what we do is
  727. 24:51make people need us more. The
  728. 24:53philosophy of our competitors is
  729. 24:55understood, so this is being healthy. I
  730. 24:57was going to say it last, but I have
  731. 24:58one more extra tip at the end. Few SEOs
  732. 25:02do this. We at Palo Seco. We have been
  733. 25:04doing it for a long time and we can
  734. 25:06easily see this in other projects and
  735. 25:07so on, which is something that we see
  736. 25:09that most are not doing. Structured
  737. 25:11data, the Markup schema, in other words
  738. 25:14, Jason LD well implemented, explicitly
  739. 25:16tells any LLM system with RAC, who are
  740. 25:19you? What do you do? Where are you? and
  741. 25:21that you respond. That is super
  742. 25:23important to have and most companies,
  743. 25:26even if they do not use it. What I
  744. 25:28recommend is that you use this tool,
  745. 25:30the validator, enter the URL here and
  746. 25:33you will see if your schema is done
  747. 25:34correctly and so on. If you do not know
  748. 25:37what I am talking about, I hope you
  749. 25:38join the community for free and there
  750. 25:40you will see different resources on
  751. 25:42this. And finally, that This is where
  752. 25:44the real deal comes in and where people
  753. 25:47start to hate me, the 2026 EEAT. This
  754. 25:50is where we really delve into the
  755. 25:52community and where I think most SEOs
  756. 25:55fail and that's why they're not getting
  757. 25:58results. Now, when we do SEO, what we
  758. 26:01want to do, as I said at the beginning,
  759. 26:03is for our client to sell and invoice
  760. 26:05more. So we have a lot of factors to
  761. 26:07work on. After all, EAT is the
  762. 26:09definitive filter that Google uses to
  763. 26:11prioritize human experience and real
  764. 26:13authority in the face of the massive
  765. 26:15rise of AI-generated content. Because
  766. 26:18I'm sure there are a lot of SEOs here
  767. 26:19who create content with artificial
  768. 26:21intelligence, which is supported and is
  769. 26:23great, but you know that if you don't
  770. 26:25have a digital footprint, it's going to
  771. 26:27cost you a lot. So, some strategies
  772. 26:28that we've used, for example, the whole
  773. 26:31issue of brand queries, what Google and
  774. 26:33these LLMs do now is for users to
  775. 26:35search for your product with your brand
  776. 26:37. Those are signals that we send to the
  777. 26:41algorithms that, hey, our product is
  778. 26:42better. What they want is for people to
  779. 26:45search for your product with your brand
  780. 26:46. Search for your business online. We
  781. 26:48are achieving this with Pedro Seo, with
  782. 26:50my personal brand, and with Palo Seco,
  783. 26:52with my agency, and it is increasingly
  784. 26:54getting more searches. One strategy
  785. 26:56that we use to boost brand queries is
  786. 26:58the one we use with this course here.
  787. 27:00We did the complete Chatpt 2025 course,
  788. 27:03which worked very well, but what we did
  789. 27:05to boost this course was to make
  790. 27:07different videos on Instagram where we
  791. 27:10told them the typical thing about
  792. 27:12commenting on X and sending you the
  793. 27:14link to the video. However, when they
  794. 27:16commented on X, we did not send them
  795. 27:18the link to the video because if I had
  796. 27:20sent them the link to this hour-long
  797. 27:22video, what I would have achieved is
  798. 27:24that the user from Instagram would
  799. 27:26click on the video, watch 2 minutes,
  800. 27:28and leave. So you say, ", we are
  801. 27:30sending signals to YouTube that this
  802. 27:32video is of poor quality, there is no
  803. 27:34user retention." So, what we did was
  804. 27:36tell them, "Hey, get ready because you
  805. 27:38are going to start using CHGPT like
  806. 27:39never before.""Blah blah blah. The
  807. 27:41course is on YouTube, you'll find it by
  808. 27:43searching for" chat GPT course. "Pedro,
  809. 27:45obviously the CTR was much lower,
  810. 27:48people barely clicked. Well, they
  811. 27:50barely clicked. It was sent, well, it
  812. 27:52was sent in this case, in this case it
  813. 27:54was sent 1000 and some times, but since
  814. 27:56we repeated the format several times,
  815. 27:57all this was multiplying for us. So,
  816. 27:59what we basically achieved is that you
  817. 28:01went to Fuente Trá, searched for
  818. 28:03YouTube and you saw how" Chat GPT
  819. 28:05course Pedro "was searched for a
  820. 28:07thousand and some times. This was
  821. 28:09giving Google signals that people were
  822. 28:11searching for Pedro's Chat GPT course.
  823. 28:13And who was that Pedro? Well, not us.
  824. 28:16So here, as you can see, for example,
  825. 28:17we were positioned with CHG GPT, which
  826. 28:19was like 16,000 searches, CHGBT course,
  827. 28:21CHGBT course, well, how do you say"
  828. 28:23chapp "? So all this was adding up
  829. 28:24searches for us and this was ultimately
  830. 28:26reflected in brand queries. People were
  831. 28:28searching for my course, We boosted it
  832. 28:30here. And then we also saw this
  833. 28:32reflected in Google. If you searched
  834. 28:34for a free GPT chat course, it said
  835. 28:36there were many, but who actually
  836. 28:37appeared here? Well, we did. Okay? So,
  837. 28:39at the end of the day, what we were
  838. 28:41doing was boosting the entire area of
  839. 28:42brand queries in order to improve our
  840. 28:44site's overall EEAT. The next point is
  841. 28:46brand content, link building, but
  842. 28:48quality. Let's forget about buying
  843. 28:51links with keywords that point to our
  844. 28:53website and so on. It works, it still
  845. 28:56works, but its days are numbered. What
  846. 28:58we have to do are press releases that
  847. 29:01people want to share, that people want
  848. 29:03to read, and so on. That's why we focus
  849. 29:05on appearing in articles in highly
  850. 29:07recognized media outlets, and it's not
  851. 29:09an article that someone has written or
  852. 29:11a text that makes no sense, but rather
  853. 29:13a person comes, interviews you,
  854. 29:15understands everything, and creates an
  855. 29:17article that people then want to read.
  856. 29:19And that's also why we organize
  857. 29:21in-person events to give that brand
  858. 29:23authority. In this case, I got together
  859. 29:25with Peter to do a In-person event in
  860. 29:27Barcelona that was great, it was very
  861. 29:29well received, and ultimately we do
  862. 29:30that to boost the entire EEAT part.
  863. 29:32After all, that's what building brand
  864. 29:34authority is. What I want you to take
  865. 29:36away is that the key is really in your
  866. 29:38digital footprint. The number of times
  867. 29:41people are searching for you, people
  868. 29:43are finding you online. I also really
  869. 29:45like something I read, I don't know who
  870. 29:47said it, but links or link building,
  871. 29:49even if it's a no-follow article, still
  872. 29:51adds up to me. And even literal text,
  873. 29:54you can have them mention your brand,
  874. 29:56but don't include the link. Having them
  875. 29:58mention your brand helps a lot. That's
  876. 29:59especially true on Reddit. I don't know
  877. 30:01where I read it, but on Reddit they
  878. 30:03collect a lot of information, they read
  879. 30:05all the posts, and having your product
  880. 30:07or brand mentioned helps everything
  881. 30:08else add up. Reddit in general, okay?
  882. 30:11Then the whole issue of social networks
  883. 30:13, YouTube and so on, we know that all
  884. 30:15this is, come on, essential. That's why
  885. 30:17we in the community have a module where
  886. 30:19we explain how to grow on social
  887. 30:21networks, because because Even if we do
  888. 30:23SEO, if we do social media it helps our
  889. 30:25SEO indirectly. It’s still part of
  890. 30:28off-page SEO, SEO. If you’re an SEO,
  891. 30:30you’ve probably heard that on-page
  892. 30:31SEO is about link building and little
  893. 30:33else, right? Off-page SEO is ultimately
  894. 30:35everything outside of your page. So
  895. 30:37it’s the whole digital footprint part
  896. 30:38. It’s much more complicated than
  897. 30:40on-page SEO. We can do on-page SEO very
  898. 30:42easily. In fact, now an AI generates a
  899. 30:44web page ready with all the minimum
  900. 30:46viable SEO requirements. So, you have
  901. 30:48to do all the on-page SEO work, which
  902. 30:50is what it costs. We at the agency are
  903. 30:52very clear about it: a company or
  904. 30:53business that comes with well-developed
  905. 30:56social media, a powerful YouTube
  906. 30:57channel, and a certain authority within
  907. 30:59its sector. We can even guarantee
  908. 31:01results. We know that what we’re
  909. 31:03going to do will work much better.
  910. 31:05That’s why, many times when we do all
  911. 31:07the SEO, what we do is enhance the
  912. 31:09entire off-page part, whether it’s
  913. 31:11with social media, email marketing,
  914. 31:13campaigns, etc. And finally, all the
  915. 31:14local SEO part, which is I know that
  916. 31:16more than one person here is already
  917. 31:18working very well on it. Well, this is
  918. 31:21another huge melon that we also talked
  919. 31:23about within the community, and I
  920. 31:24remind you that you can subscribe for
  921. 31:26free using the link in the description.
  922. 31:28There, of course, there are a lot of
  923. 31:30strategies and so on. Well, we're
  924. 31:31almost done, as you can see. Well, I
  925. 31:33hope that if you've made it this far,
  926. 31:36I've convinced you to accept SEO. SEO,
  927. 31:39sorry, you already accept SEO, or even
  928. 31:41geo. And of course, I've already said
  929. 31:43it, I don't care what they call it. I
  930. 31:45mean, I don't really care. What we want
  931. 31:48to do with CO is to improve things for
  932. 31:51our clients, to increase their sales,
  933. 31:53and so on. So, if you're someone who
  934. 31:55wants to learn modern SEO, SEO that
  935. 31:58really works, accompanied by an agency
  936. 32:00that's achieving it with its clients, I
  937. 32:03want you to know that we have a
  938. 32:05community where we explain all our
  939. 32:07processes, we give you unlimited
  940. 32:09support, and you'll have access to our
  941. 32:12roadmap with the step-by-step
  942. 32:14instructions that you should follow to
  943. 32:16work on your SEO and get more traffic.
  944. 32:19and, above all, get more sales. If you
  945. 32:21want more information, I'll leave a
  946. 32:23link in the description of this video
  947. 32:25and tell you that I'd love to see you
  948. 32:27inside, I'd love to meet you, so I hope
  949. 32:29to see you there and anything, any
  950. 32:31doubts, questions, the usual. And now,
  951. 32:33if you liked the video, give it a good
  952. 32:35like, subscribe if you're not
  953. 32:36subscribed. See you on the road to
  954. 32:38continue learning. A big hug.

About this transcript

This page contains the full transcript of Curso GEO 2026: Cómo hacer que la IA cite tu web by Pedro SEO, generated from the public captions YouTube serves with the video. The transcript has 6,708 words across 954 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.