Get the replay by email
About 61 minutes, in 15 short segments. You’ll have 72 hours from the moment you first open it. If you buy the course, this module comes with it, along with all the others, and you keep them for life.
Already signed up? Open the link in your welcome email.
Inside this workshop
- ✓The deeper purpose of SEO. Serve your people, clarify your message, then update what spreads.
- ✓The three doors people use to find you through AI. By name, by kind, and by question.
- ✓Websites, not only content platforms. Why AIs learn far more from websites than from content platforms.
- ✓Checks you can run on your own site. A site search and a Common Crawl lookup, to see what search and AI can read.
What people said after watching
“George has an amazing ability to take complex topics and make them understandable, practical, and manageable. His Will AI Recommend You? webinar gave me a clear understanding of why being discoverable through AI matters and, more importantly, the specific first steps I can take to make it happen.”
“George takes the lid off the mysteries of SEO in the age of AI. His insights are practical and actionable. I've never uncovered any of this before now. It's very powerful.”
“Great webinar, you were relaxed, available, and un-salesy. The content sold itself.”
“I did not feel overwhelmed with the amount of content in this course. I highly recommend his courses.”
Your replay
Your replay is open until … (your time).
Two ways through this workshop
Use either one, or both. Each remembers your place.
Click on any of the options above.
0:00 of 1:00:42
Pick up where you left off: …
About this workshop
This was George's free live workshop Will AI Recommend You? on Tuesday, October 27, 2026, recorded as 15 short videos (about an hour in all).
Try these four things from the workshop:
- Pick three pages to update. Name the three pages on your site most worth updating soon (1.2).
- Ask an AI who you are. In a logged-out, private browser window, ask an AI who you are and what kind of work you do. Note what's missing or wrong, and choose a short title you'll use everywhere (1.3).
- Install two free Google tools. Set up Google Search Console and Google Analytics with your AI's help, today if you can (1.7).
- See which pages Google knows. Type
site:followed by your web address into Google and see which of your pages it knows about (1.8).
Module 1’s Lessons:
- 1.0Welcome2 min
- 1.1The deeper purpose of SEO - a flywheel10 min
- 1.2Update what spreads5 min
- 1.3Three doors, one castle7 min
- 1.4What still counts - the quilt7 min
- 1.5What it looks like when AI recommends you2 min
- 1.6Content is a commodity - liberate it6 min
- 1.7First technical steps - install the tools, let AI read the data5 min
- 1.7aAsk Google to index your new and updated pages · supplement3 min
- 1.8Two checks to try - site search and Common Crawl2 min
- 1.9AIs learn far more from websites than from content platforms2 min
- 1.9aSubstack posts and your website · supplement1 min
- 1.9bLet AI publish to every place at once · supplement1 min
- 1.9cIs video worth it for SEO and AI search? · supplement4 min
- 1.10Your feedback, and how the course works → the invitation4 min
1.0 · Welcome
Video · 2 min
Key points from the video
- A new name: George has taught this course several times over his 17 years of teaching. Its old name was Authentic SEO, and it's now SEO with AI Support, because for anyone willing to use AI, it can now handle the technical parts of SEO, the data analysis, and even updating content and researching the market.
- Still heart-based: before any technique, this module reframes SEO from a more soulful, authentic perspective.
- Check-in question: given what you understand (or don't) about SEO right now, how does it feel to you? Answer in a word or a few. Live viewers answered in the chat; replay viewers can add theirs in the Student Discussion doc.
- From the live chat, attendees' first feelings about SEO:
- overwhelming, confusing, intimidating
- mysterious, and evolving fast
- one more thing to do in an already busy business
- helpful and fairly simple (a student who had taken George's earlier SEO course), and exciting (another student)
- George's response: he agreed with many of those words, and hopes to make SEO feel simple and exciting for everyone as the course goes on.
George's answers to students' questions (click to show)
From the chat: besides the words George read out, students wrote daunting, exhausting, complicated, "a nuisance" and "constant change". Two saw AI changing how SEO feels: "uninteresting before, now with AI - more interesting" and "It’s one thing I feel totally ethical turning over to AI".
A student shared: the environmental and human impacts of AI are very hard for her to reconcile with her work, especially earth-healing work.
George replied (right after this video): he holds a master's degree in sustainable business development and cares about sustainability and environmental justice more than most people do. Two things: AI is helping sustainability experts do more of their work, and data centers and AI do have environmental impacts, though much of the mainstream news coverage exaggerates them or gets them wrong. He wrote a long blog post on AI and the environment. Another student had found it helpful, and a third posted the link: https://georgekao.substack.com/p/can-we-use-ai-and-still-love-the
Talk about this lesson with other students in the Student Discussion doc.
1.1 · The deeper purpose of SEO - a flywheel
Video · 10 min
Key points from the video
- Why many people avoid SEO: the technical side is intimidating (how to become visible on search engines, and now on ChatGPT, Claude and the like), and mainstream SEO is aggressively competitive: track keyword and ranking data, then keep trying to beat the top pages for each keyword. To George it felt like a constant battle, and nothing about it felt soulful.
- What it's called now: generative engine optimization (GEO, which he enjoys because it's his own nickname), answer engine optimization (AEO), LLMO, AIO and more. They all mean becoming visible beyond Google, as fewer people use Google and more ask ChatGPT, Claude, Gemini, Grok, DeepSeek or Meta's AI. George treats it all as one body of work.
- Authentic, as he uses the word: true to your higher self, meaning what your truest, most connected self would choose to give your limited energy to in this life. For him that rules out battling bots and aggressively competitive people for rankings. The good news on his slide: doing the three steps below well also does SEO well and gets you found by AI.
- Step 1, serve your people: start from the intention to serve; the number-one spot in Google or in ChatGPT's answers stops being the goal. George frames business as service, collaboration, even ministry: the work exists to benefit the people who pay for it, or it would be a hobby. Caring leads you to ask what your ideal clients are seeking, which is what they type into Google and ChatGPT: their confusion, their frustration, the solution they want. Understanding their pain and yearning brings empathy naturally. George calls this the golden rule applied to marketing: "If we don't care, they won't care."
- Step 2, clarify your message: once you know what they're seeking, describe your work and your message in their language, and write it down. Kept in your head, it comes out differently every time, which is fine while you explore; eventually it belongs on your website and social media.
- Step 3, update what spreads: notice what spreads, starting with your website data. AI now reads Google Search Console and Google Analytics for you, gives you the bottom line, and answers your questions about it: which pages get the most traffic, from which sources, and which searches in Google or ChatGPT bring people to which pages. Notice what people say about you online too, which AI can help analyze. If 3 of your 20 pages are the ones people keep visiting, sharing and referring to, especially the right people, focus your energy there. You don't have to keep 100 pages or the whole blog current.
- Why it keeps turning: what spreads shows you what serves your people. That prompts conversations with the people themselves about why they search for those things, which sharpens your message and shows you what to update next. Each pass around the flywheel makes you more skilled at the work.
- His question to you: has this reframed SEO for you, and how do you feel about SEO now? He hopes it's now work you'd want to do, since it amounts to serving your people, clarifying your message and keeping what helps them up to date.
From the live chat (click to show)
When George asked whether the flywheel had reframed SEO for them, students wrote:
- "Feels a bit empowering"
- "helps reduce overwhelm feeling and narrow focus for what I can do"
- "it is like, oh, i want to do this rather than run from it"
- "Already sounding warmer and inviting."
- "Such a great reframe from “overwhelm” to “a more effective way to serve our people”"
- "Sounds simpler"
- "It’s like dating!"
Talk about this lesson with other students in the Student Discussion doc.
1.2 · Update what spreads
Video · 5 min
Key points from the video
- Why this step comes first: of the three flywheel steps, update what spreads confuses people most, and it's also the most time-sensitive. Many clients come by word of mouth, which usually means someone shares a page of your site, or the site itself, and people then look at a few particular pages to decide whether to work with you. Those pages need to be current soon.
- Homework question: which three pages on your site most need updating soon? Important here means what spreads:
- The most visited pages: ask your AI how to analyze your website data and find the pages people visit most. You can ask this today, whether or not you continue the course. Update those first so they're true to what you offer now, your current message and your framework, and update any date shown on the page.
- Pages visited for the wrong reason: your data may show a random page drawing visitors through an unrelated keyword. Update it so you draw your own people and help them decide well.
- Pages you share yourself: the blog posts or pages you keep sending to clients and your audience. Give yourself (or your AI) a recurring task to refresh them every quarter or half-year.
- Decision pages: your work-with-me and service pages, and your About page, which changes as you evolve. George recently updated his.
- A prompt for finding your most visited pages:
How can I analyze my website data to find the pages people visit most?
- In the chat: name the pages that already come to mind. Reading other people's answers jogs ideas (that page does get traffic, and I haven't updated it in a while), and you leave with useful homework.
George's answers to students' questions (click to show)
A student asked: do you mean the pages we think are most important, or the ones we'll find by analyzing our data?
From this lesson: both. Start with what you already know (the pages you share, and decision pages such as Work With Me and About), and check it against your site data, which your AI can analyze for you.
A student asked: how do I find out which pages of my website are visited most?
From the lessons: ask your AI to analyze your website data (this lesson); install Google Search Console and Google Analytics so there is data to analyze (1.7); Google Analytics shows which pages get the most visits and where people go next (1.7a).
From the chat: the pages students named as most in need of an update included the home page, About, services or offerings, FAQ, contact, an events page, and a specific workshop or course page. One student added: "testimonials are dusty".
A student shared: she doesn't know whether anyone visits her website and is thinking of letting it go. (Not answered live. Lesson 1.7 shows how to find out who visits, and 1.9 shows why AIs learn far more from websites than from content platforms.)
Talk about this lesson with other students in the Student Discussion doc.
1.3 · Three doors, one castle
Video · 7 min
Key points from the video
- The castle: George's doorways analogy, applied to SEO. Your business is a castle with many rooms: different offerings, kinds of clients, problems you solve, goals you help with, modalities you bring. People approach from different directions and come in through the door that draws them. For SEO he keeps it to three doors: by name, by kind and by question.
- Door 1, by name. This is what happens when people hear about you and consider working with you. Try it on George first: open the AI you use most (ChatGPT, Claude, Gemini) and ask:
Who is George Kao, and what kind of work does he do?
- Then search for yourself, logged out: use an incognito or private window (in Chrome, File → New Incognito Window, or Command-Shift-N; Safari calls it a private window) and go to ChatGPT, which works without logging in. Logged in, your AI already knows all about you, so the test tells you nothing. Does AI name you, and is what it says accurate? If not, there's website work to do. Ask, with your own name:
Who is [your name], and what kind of work do they do?
- Shared names, and your title: if many people share your name, especially well-known people, it's harder. That's branding homework: choose a short title and keep using it wherever you share your work, so it stays consistent across your site, your social media, and the podcasts and interviews where others talk about you. George's own case: for a long time he called himself an authentic business coach, which he thinks is still how ChatGPT describes him. He's now shifting to soulpreneur coach and isn't sure AI knows that yet.
- Door 2, by kind: someone asks for a type of provider (for example, a business coach for solopreneurs who want more soulful, authentic work) or a type of course, retreat or workshop, and hopes the AI recommends someone. Whether it names you depends on the SEO work you do.
- Door 3, by question: the hardest to reach, and it makes your site more valuable to AI and to search. Which questions do your ideal clients ask Google and ChatGPT? Some are broad (how can a life coach become more visible?), some ask about a term (what is a solopreneur?). Understand them, then answer them from your own perspective. The rest of the course works on finding these questions, through online searches with your AI and through conversations with people, even asking people to share their screen while they type what they'd search.
George's answers to students' questions (click to show)
From the chat, asking AI who George Kao is: the answers students pasted mostly called him an "authentic business coach" (his older title) who works with heart-centered solopreneurs and independent professionals. None used his newer title, soulpreneur coach, which fits what he said about AI lagging behind a change of title. One answer also noted other George Kaos, a historical translator and a food-industry executive, so even he has the shared-name problem. Two students' AIs tied the answer to their own work, a sign they were logged in, which is why the self-search belongs in a private window.
From the chat, searching for themselves: results varied. Some were found and described accurately. One was found but described vaguely, mostly through other websites than her own. Two got the right bio next to a photo of a stranger. One saw a job from 20 years ago listed first. One shares her name with a TV character, and links about her still came up. One found many people with the same name, and one wasn't found at all, only a business with the same name. One student concluded they needed a unique title that describes them; another said she had wrestled with that for a long time.
A student asked (in the Q&A at the end): she does many kinds of work, and to make it less confusing she has started splitting her one big castle into two: a mainstream-friendly practice (holistic hypnosis for things like confidence, sleep and anxiety) and a broader, more esoteric community. She keeps wondering whether that's wise, since then there are two of her. And when someone asks AI about her by name, will two sites confuse the AI?
George replied: his answer has changed in the last six months. He used to tell solopreneurs they lacked the time and energy for two castles; now, if you're willing to use AI, you can run three or five, because AI can manage them for you. He thinks it's a smart idea: search engines and AI try to understand each website as a whole, so a site holding everything is harder to understand, while a focused site makes it clear who it's for, almost like a stage the client is at. Give each its own domain name. AI might at first take you for two separate people, but he expects it to be (or already be) smart enough to see one person with two websites. On a name shared with someone else: if the other person works in a very different field, people can tell you apart; if they're in a similar field, it's very hard, and he now thinks some people should change their names or use a different brand name.
Talk about this lesson with other students in the Student Discussion doc.
1.4 · What still counts - the quilt
Video · 7 min
Key points from the video
- The quilt: the pieces that still lead search engines and AI to respect someone enough to recommend them and refer people to their website. Before AI, George would have said to pick a few so you don't get overwhelmed. Now, the more skilled you get with AI, the more your AI can keep working on all of them for you.
- Your website: still the number one place ChatGPT, Claude, Gemini and the others train on when deciding whether to refer you, ahead of social media. (He shows his own study of this in 1.9.)
- Podcasts: your own podcast gets picked up by search, though less powerfully than your appearances on other people's podcasts, which also feed the next three pieces.
- Mentions: other websites mentioning your business.
- Reviews: someone explaining in depth why you do good work. Clients can write these, but they often come more easily from a colleague who interviewed you in depth.
- Genuine links: for example, a host who interviews you, then writes a blog post about it that links to your site. Swapping interviews is fine, but search engines can see the swap and give it less weight. When your content is good enough, or you do enough netcaring, people will ask to interview you without asking for anything back. (You can ask AI what separates a genuine link from a fake one.)
- Directories: most people use too few. List your business in as many credible online directories as you can. For work related to mental health and emotional regulation, George named Psychology Today: his wife, a therapist, has been listed there for years and gets plenty of clients through it, and he guessed it costs around $25 a month (it's $29.95). He wasn't sure whether it requires a license (see the note below). Directories exist for all kinds of professions, coaches and healers included, so ask your AI which are most credible and most used by search engines, where your work would stand out, and to help you get listed.
- Your clients' words: beyond testimonials, how do your clients describe your kind of work, the problems you help solve, the yearnings and goals you help with, the experiences they want? What do they search for? Find out through keyword research with AI and through your conversations. AI can also draw these words out of your session transcripts or social media comments; the term is voice of customer (VOC). George's AI built him a detailed VOC document from his webinar chat logs: how his audience talks, the words they use, how they describe his work and their own problems.
- Protect privacy first: before having AI analyze client transcripts, turn off the setting that lets your conversations be used for training, so private data stays private. He teaches this in his course on setting up AI.
- What's next: the rest of the course goes deep on this whole picture. On your own, have your AI help you work on each piece.
From the live chat (click to show)
A student shared: "Yes you have to be licensed." (Checked after the workshop: Psychology Today's terms encourage licensed or certified members, so don't assume it will accept an unlicensed coach or healer. As he says, there are directories for other professions too, coaches and healers included.)
Talk about this lesson with other students in the Student Discussion doc.
1.5 · What it looks like when AI recommends you
Video · 2 min
Key points from the video
- AI Overviews: most Google searches now show an AI-written overview before any links. So AI search includes people Googling as usual, who get an AI answer first, with a few source links to look at further down.
- A colleague's example, shared with his permission from George's SEO mastermind a couple of months earlier: a search for the phrase he uses to describe his work, "what is a transformational relationship", brings up an AI overview that recommends his YouTube videos. YouTube is a highly trusted source for search engines and AI (Google owns it), partly because a video takes more work to make, and a video that gets seen, liked and shared is especially powerful.
- George's own example: ask whether coaches should charge what they're worth, and AI recommends one of his blog posts right away, a post that has drawn readers for years.
- What to take from it: once you know your clients' words and the name or phrases you want to be found for, create content for them and put it on your website and on places like YouTube.
- George welcomes more examples of people finding you through AI; add yours in the Student Discussion doc.
Talk about this lesson with other students in the Student Discussion doc.
1.6 · Content is a commodity - liberate it
Video · 6 min
Key points from the video
- Content got harder: George has made content for over twelve years, and the last three have been the hardest for getting traction. AI makes it easy for anyone to write and to make images and video, so the flood keeps growing: more creators as corporate jobs disappear and people go solo, and for years content creator (a YouTuber, famous on TikTok) has been one of the top answers when young people are asked what they want to be.
- The outdated fear: if I share the answers to my clients' questions on my site, on social media or on YouTube, won't people stop needing to work with me?
- George changed his mind: he used to teach that free content should cover the why and the what and keep the how for paid courses. He has reversed that over the last couple of years, so longtime followers still using the old rule can let it go. Content is now a free commodity, and the difficulty is getting anyone to look at your content at all; he sometimes feels he's begging one person to read a post. George would sooner have people take in your most in-depth content, learn about you and come to trust you.
- Liberate your content, the how included, and make it digestible. Digestibility is what separates free content from a paid course or working with you: a whole course posted as it is would overwhelm people and go unread. So he still makes webinars and paid courses, while gradually turning his paid course material into blog posts and other pieces people can take in, so AI finds it all and tells people he's trustworthy, worth learning from, and what his distinctive perspectives are.
- Two pages written mainly for AI: www.georgekao.com/best, his best ideas, chosen from audience surveys he ran over the past year or two about which ideas helped people most; and www.georgekao.com/perspectives, his unique perspectives, which he had his AI find by looking across his website, YouTube videos and courses for answers AI wouldn't usually give. Pages like these show AI a unique, useful body of work, which makes you more credible and more likely to be recommended.
- Write for both readers: make content easy for humans to take in, and think about how AI reads it too. Half in jest, he told everyone to stop protecting their content.
From the live chat (click to show)
A student asked: what does "getting traction" mean? (Not answered live. In this lesson George uses the phrase for content getting noticed at all, seen and shared.)
A student asked: are georgekao.com/best and georgekao.com/perspectives separate websites? (Not answered live. They're two pages on George's own website.)
From the chat: several students said content is hard for them, and one noticed "A lot more content, not sure about a lot more quality". Liberating your content drew "mind blowing" and "Liberate your content WOWZA!!!!"
Talk about this lesson with other students in the Student Discussion doc.
1.7 · First technical steps - install the tools, let AI read the data
Video · 5 min
Key points from the video
- You can start these before Module 2. Every step of the flywheel relies on data you can study: what people are looking for, the words they use, what gets shared. Go to where your people are: your biggest fans may eventually do the hard work of climbing your mountain to understand you, but most people need clients sooner than five or ten years from now, so they come down to the riverbank where their people are.
- Step 1, install Google Search Console and Google Analytics. Both tools cost nothing, and Google gathers your data only once they're installed. George used to teach a whole course on the many clicks this took; now your AI can install them with you, and can first check whether you already have them (some people have had them for years without knowing). Do it today if you can, since every day without them is data you won't get back. To start, ask your AI:
Please check whether my website already has Google Search Console and Google Analytics set up. If either one is missing, walk me through installing it one step at a time, and do as much of it for me as you can. Thank you for your help with this.
- Step 2, let AI make sense of the data. Search Console and especially Analytics used to overwhelm George too, with their charts and tables. Now he asks his AI to study his Search Console data and give him the bottom line of what needs attention, and a scheduled task does this every two months and brings him a report.
- From one of those reports: a blog post about his concentric circles business model was drawing people who were searching for something else, so the AI suggested updating it to attract the right readers. And because most visitors land on his home page, it suggested adding a where-to-begin section there, so people who feel overwhelmed can start sooner. The AI reads the data and also recommends what to do about it: which questions to answer, what's working, which page is spreading and worth refreshing.
- The new way: install the tools, then let AI do the sense-making. That's what makes SEO so much more pleasant in the AI age.
Talk about this lesson with other students in the Student Discussion doc.
1.7a supplement · Ask Google to index your new and updated pages
Video · 3 min
Key points from the video
- Recorded during the Q&A at the end of the workshop.
- Part of George's publishing process: whenever a page or blog post is new or updated, his AI submits it in Google Search Console, which asks Google to index the page. Google might find the page eventually; a request from you (or your AI) asks Google to look sooner. Google guarantees nothing, and it can take a few days to a few weeks.
- When a
site:search surprises you (few of your pages show, or important ones are missing or low in the list; 1.8 shows how to run one), use Search Console to request indexing of those important pages. Treat it as basic website hygiene, and ask your AI to do it every time you publish or update a page or post:Each time I publish or update a page or blog post, please request indexing for it in Google Search Console.
- A checkbox is different: a student asked whether ticking "let Google index this" in her website's SEO settings does the same. George's answer was no: go to Search Console yourself, or have your AI go there, and make the request, because the checkbox only allows indexing, while the request asks Google to come and look.
- A good first step after this workshop? The same student asked whether to have her AI look at her Google data and tell her which pages people look at most. Yes, and your AI would use both tools:
- Google Search Console shows the exact words people type into Google that make your pages appear in the results: how your site interacts with Google's search.
- Google Analytics goes deeper into what happens on your site: which pages get the most visits, where people go next, how long they stay on each page and what they click.
George's answers to students' questions (click to show)
A student asked (just before this video starts): what's the purpose of searching site: plus her web address?
George replied: to see which of your pages Google has indexed. He believes the ones listed higher are the ones Google prioritizes. In practice, that means making sure your important pages are indexed, using Google Search Console, which is where the video picks up.
Talk about this lesson with other students in the Student Discussion doc.
1.8 · Two checks to try - site search and Common Crawl
Video · 2 min
Key points from the video
- Check 1, a
site:search: in Google, typesite:followed by a web address, starting with George's (site:georgekao.com). The search shows which of that site's pages Google has in its database to bring up for various searches. Then try your own site. If few pages come up, you're not alone: many sites aren't properly indexed by Google, and your AI can fix that (see 1.7a). If many come up, congratulations on a very basic step done. George's, to type into Google:site:georgekao.com
- Check 2, Common Crawl (commoncrawl.org): George describes it as a database many AIs train on, and names ChatGPT, Claude and Google among its users. (Checked after the workshop: Grok and DeepSeek couldn't be confirmed, and the big AI companies also run their own crawlers.) You want all your useful pages in that archive. The course goes into Common Crawl in depth as one of the key resources for AI SEO. If you want to start now, ask your AI:
Please check whether my website's pages are in Common Crawl, and if some aren't, help me get them there.
George's answers to students' questions (click to show)
A student asked: how do you search Common Crawl for your own website or name?
From this lesson: the question wasn't answered live, and in the video George suggests asking your AI whether particular pages are in Common Crawl, or to help you get them there.
A student shared: clicking the site: link George posted in the chat didn't search anything, while typing site:georgekao.com straight into Google's search box worked. Type the search into Google's search box yourself; a link won't run it as a search.
Talk about this lesson with other students in the Student Discussion doc.
1.9 · AIs learn far more from websites than from content platforms
Video · 2 min
Key points from the video
- The finding: AIs learn far more from websites than from platforms such as Substack, and much less again from social media.
- Why: AI treats Substack, LinkedIn articles or Medium as one giant website and samples from that one site. AI doesn't treat your publication as a site of its own and crawl all your posts.
- George's own numbers: when he searched Common Crawl, it held about 100 of the almost 500 posts in his main Substack, and none from his newer publications (AI for Soulpreneurs, about six months old, Soulpreneur Money and SpiritPlan). Far more of his website's blog posts were there.
- The bottom line: put your energy into keeping your own website updated, more than into posting on Substack or social media, since AI isn't training on most of those posts.
Talk about this lesson with other students in the Student Discussion doc.
1.9a supplement · Substack posts and your website
Video · 1 min
Key points from the video
- The question (from the Q&A at the end): a student has a Substack, plus a link from their website where visitors can read the articles without going to Substack. Will AI train on those posts as part of the website?
- George's answer: he wouldn't count on AI learning from those posts. Have your AI copy the Substack posts onto your own website as separate blog posts, with their dates. With the same text in two places, Google picks which copy to show, and it doesn't penalize duplicates unless they're meant to deceive. Copying them takes AI about an hour, or overnight if you have hundreds, and you do no admin. Then ChatGPT, Gemini, Claude, Grok and the others can learn from them the next time they train, which he believes happens several times a year. The reason, again: AI learns from website content far more than from Substack posts.
George's answers to students' questions (click to show)
A student asked (later in the Q&A): so the articles should be brought over onto the website itself?
George replied: yes, 100%.
Talk about this lesson with other students in the Student Discussion doc.
1.9b supplement · Let AI publish to every place at once
Video · 1 min
Key points from the video
- AI can publish one post to many places at once. George's AI knows his blog publishing process: he writes each post together with his AI, and once the text is final and he says go ahead, it publishes to his Substack, his website and all his social media.
- He barely touches social media now. His AI also reads the comments and brings him only the relevant ones. The AI blocks people who are mean and passes along any useful nugget without telling him who wrote the comment.
- For you: if you're open to using AI for publishing, let it post to Substack, your blog, LinkedIn and the rest, so publishing in several places stops being extra work.
George's answers to students' questions (click to show)
A student asked (in the Q&A at the end): she has a blog on her website and planned to move everything to Substack and close the blog. Does she need to keep both?
George replied: don't close the blog, for sure. He then recorded this video as the rest of his answer.
A student asked: how does AI "do it all for you"? (Not answered live. Earlier in the Q&A, George pointed people who want their AI to work this way to his short course How I Set Up My AI: www.georgekao.com/aisetup)
Talk about this lesson with other students in the Student Discussion doc.
1.9c supplement · Is video worth it for SEO and AI search?
Video · 4 min
Key points from the video
- The question (from the Q&A at the end): how much is video worth for SEO and AI search, and where should the videos go?
- YouTube first: it's George's top recommendation. Google is still among the top search engines, and ChatGPT and the other AIs also train on YouTube content. Keep your channel's automatic transcription on (he believes it's on by default); without it, search engines and AI would have to watch the video to understand its content.
- Then put it everywhere, with your AI's help: YouTube, and Substack too, where an uploaded video can become a podcast with one more setting, sent out to iTunes, Spotify and elsewhere for free. Substack charges for some things, though not for podcasts, blog posts or newsletters.
- Video hosted on your own site: George isn't 100% sure whether that helps, and asked anyone who knows to comment. He doesn't do it, partly out of a perhaps old-school worry that streaming would use up his site's resources.
- What he does instead: he embeds relevant YouTube videos on the related pages and blog posts. Visitors press play and watch without leaving your page, while the video streams from YouTube's servers. He says embedding YouTube videos on your site does help with search, and recommends it to everyone with a channel. (Checked after the workshop: an embed lets your page show in Google's video results, but Google doesn't treat it as a ranking boost on its own.)
- Add
?rel=0to the end of the YouTube address in the embed code. When the video is paused or ends, YouTube then suggests only videos from your own channel instead of random other channels. If that's unfamiliar, ask your AI:How do I add ?rel=0 to the end of the YouTube address in my embed code?
George's answers to students' questions (click to show)
A student asked (later in the Q&A): does ?rel=0 go on the video in your YouTube channel, or on the embed?
George replied: on the embed code that goes on your website. Visitors never see the code, only the video.
From the chat: a student asked for the command in writing, and another student supplied it: ?rel=0, added to the end of the video's address in the embed code.
Talk about this lesson with other students in the Student Discussion doc.
1.10 · Your feedback, and how the course works → the invitation
Video · 4 min
Key points from the video
- Your feedback: if you have 2 minutes, George would love your feedback on the workshop: the feedback form.
- The full course: SEO with AI Support, where Modules 2 and 3 continue.
- For enrolled students, a preview of how Modules 2 and 3 run. (Live, George closed by asking for feedback on the workshop and inviting people to join the rest of the course.)
- Module 2, the playbook: George delivers a comprehensive playbook, several dozen pages, on the SEO process he now recommends to his clients, researched with care and updated for what currently works with AI search and AI tools such as ChatGPT and Claude. You can read and follow it yourself, but it's meant to be fed to your AI, and he shows you how. Your AI then walks you through the SEO step by step and does as much of it as it can, always asking your permission first, for example before updating pages based on the research you've done together. The AI interviews you to get your framework and how you work now, and helps you reach out to clients, past clients and audience members for conversations, so you hear what they're looking for in their own words. Those human-to-human conversations are yours to have.
- In the Module 2 session: you receive the playbook, install it into your AI together, and then work with your AI on the first couple of stages.
- Module 3, a couple of weeks later: troubleshooting and Q&A, once you've had time to set things up with your AI and start optimizing, and going deeper into the playbook.
- Two live sessions, both recorded, since some people who sign up can't attend. You can use the playbook installed in your AI for life, to keep making your work easier for search and AI to find and recommend.
- Access and updates: enrolling brings lifetime access. In the video George also says lifetime updates; to be precise, he keeps the playbook current while he's actively teaching this topic, researching as SEO changes, and the playbook tells your AI to check for a refreshed version every six months.
George's answers to students' questions (click to show)
A student asked: when are the next sessions?
George replied: Module 2 is Tuesday, November 17, and Module 3 is Tuesday, December 8. Both are recorded for anyone who can't come.
A student asked: she isn't ready to do everything yet. Could she take the course, gather the information, and apply it later?
George replied: yes, either way works.
Talk about this lesson with other students in the Student Discussion doc.
After you watch
If you have 2 minutes, I’d love your feedback: the feedback form.
Thoughts or questions as you watch? Leave them in the comments doc (sign in to Google to comment), where you can also read what others noticed.
If you’d like more after watching: The SEO with AI Support course. If you buy the course, this module comes with it, along with all the others, and you keep them for life.
This replay has closed
Thank you for signing up to watch.
If you’d like to keep going, the complete SEO with AI Support course is available for enrollment: The SEO with AI Support course. If you buy the course, this module comes with it, along with all the others, and you keep them for life.