Where to actually start with AI search training
Every week someone asks me the same question, usually in a DM and usually slightly apologetically: Where do I start with AI search? They have read the LinkedIn takes, they have seen the acronyms and what they want now is a structured hour with a beginning and an end.
I wrote AI Search 101 as my own answer to that and it turned into 10,000 words. Courses do a different job. They give you a syllabus somebody else has argued about, in an order somebody else has tested on real learners, which is worth a lot when a field is 3 years old and nobody agrees on the curriculum.
Three free ones are worth your time right now. I have been through the syllabus of each, checked who is teaching and what gets covered and none of them cost anything beyond an email address. What follows is what each one contains, who it suits and where it stops.
TL;DR: Which AI search course to take first
- Start with Ahrefs’ AEO Course if you want the fundamentals. 12 lessons, 1 hour 26 minutes and the only free syllabus that spends its opening module on how retrieval works before it reaches a tactic.
- Take Semrush’s AI Search Operating System if your problem is your leadership team and not your website. 39 minutes and it is built around getting buy-in.
- Take Sitebulb’s Search Matrix if you are technical and want depth on retrieval, crawler behaviour and log analysis.
- Do all 3 in that order and you have roughly 4 hours covering fundamentals, organisational strategy and technical execution, with no overlap worth complaining about.
- Paid options exist and the independent evidence behind most of them is thin. I have listed the ones that hold up at the bottom, with the caveat that I have not taken any of them.
Best free AI search courses compared
| Course | Provider | Length | Format | Certificate | Best for |
|---|---|---|---|---|---|
| AEO Course | Ahrefs Academy | 1h 26m, 12 lessons, 4 modules | Video | No | Fundamentals and execution |
| AI Search Operating System | Semrush Academy | ~39 min, 9 lessons | Video | Yes, with an exam | Strategy and stakeholder buy-in |
| The Search Matrix | Sitebulb, with iPullRank and DemandSphere | 3 modules plus a recap | On-demand video with downloadable slides | Yes | Technical depth and measurement |
Ahrefs AEO Course: Best for AI search fundamentals
Free. Sign-up required. ahrefs.com/academy/aeo-course
Taught by Sam Oh, VP of Marketing at Ahrefs, this is 12 lessons across 4 modules, at 1 hour 26 minutes in total. The modules run: How AI Search Works (21 minutes), AEO Strategy (16 minutes), AEO Execution (27 minutes) and Measuring AEO (15 minutes).
Module 1 is the part I would send to a colleague who has been asked to “do something about ChatGPT” and does not know where to begin. It covers how different platforms decide what to cite, which is the thing most people skip straight past on their way to tactics. The execution module gets into keyword and prompt research, creating citeable content, earning brand mentions and YouTube optimisation for AI visibility. That last one is unusual in a free course and it matters, because YouTube is now the most cited domain in Google’s AI Overviews.
The strength here is proportion. Ahrefs has published a serious amount of original research on citation overlap and AI Overview behaviour and the syllabus leans on that work heavily. Sam Oh teaches in the same style he does on YouTube, so if that format works for you, this will.
Where it stops: There is no certificate and the measurement module is 15 minutes, which gets you the concepts and leaves the messy attribution problems alone. Those are covered properly in how to measure AI search traffic and visibility if you want to go deeper.
Take this one if: You are new to AEO, you learn from video, or you want a single course that covers the whole arc from theory to measurement.
Semrush AI Search Operating System: Best for leadership buy-in
Free. Enrolment required. Certificate included. semrush.com/academy/courses/ai-search-os
This is the shortest of the 3 at around 39 minutes across 9 lessons, taught by Leigh McKenzie, Director of Organic Visibility at Semrush and Rita Cidre, who runs Semrush Academy.
It is aimed squarely at in-house SEO leads, digital marketing managers and agency leaders and the lesson titles give the game away: “What’s Changed and How to Tell Leadership”, “An AI Search Audit That Gets Buy-In”, “AI Search Action Plan, KPIs and Reporting”. The framework is 4 layers of brand visibility and the whole thing is organised around turning an audit into something a leadership team will actually fund.
I rate this one for a reason most course round-ups miss. Ask any in-house SEO what actually blocks AI search work and you get the same answer: Explaining to a CMO why traffic is flat, why the old dashboard no longer describes reality and why PR and customer success now sit inside your remit. Semrush has built the whole 39 minutes around that problem and the longest single lesson in the course is an 11-minute one on leading cross-functionally.
There is a certificate at the end, gated behind an exam. My honest read on certificates in this category: None of them carry hiring weight on their own yet, though AEO and GEO are now turning up in job specs, so the terms it puts on your CV may do more for you than the badge does.
Where it stops: 39 minutes cannot teach you implementation and it does not try to. Pair it with one of the other 2.
Take this one if: You need to sell AI search work internally, you are building a reporting framework, or you have a leadership conversation coming up and 40 minutes to prepare.
Sitebulb Search Matrix: Best for technical SEOs
Free. Email registration required and it comes with a 60-day Sitebulb trial. Certificate included. sitebulb.com/technical-seo-ai-training
Built by Sitebulb with iPullRank and DemandSphere and taught by Zach Chahalis, this is a 3-part on-demand series with downloadable slides for each module and a certificate you can personalise at the end. It is the most technical of the 3 by a wide margin and Sitebulb pitches it at intermediate SEOs who already have the technical foundations.
Module 1 covers the shift from keyword matching to semantic, vector-based retrieval, RAG fundamentals, passage-level retrieval and hybrid retrieval combining lexical and semantic search. Module 2 is measurement: Input, channel and performance metrics, synthetic fan-out queries, AI crawler behaviour and how to benchmark AI discovery. Module 3 moves to optimisation, covering AI index pollution, passage-level content strategy, evidence density, entity-rich language, log file analysis for AI crawler behaviour and preparation for agentic search and Machine Context Protocols.
That log file section is the one I would flag. Most AI visibility reporting collapses training crawls, indexing visits and real-time fetches into a single “AI bots” line and those are 3 completely different activities telling you 3 completely different things. Very little free training goes near it, which on its own justifies the sign-up.
Where it stops: The course page does not publish individual lesson runtimes, so you are signing up slightly blind on how long it will take you. The 60-day trial is a commercial hook, though the training itself stands up without the tool.
Take this one if: You are a technical SEO, you want the retrieval mechanics sitting under the frameworks, or your next piece of work is a log file audit.
What these AI search courses leave out
Worth knowing before you sign up for all 3:
- Agentic search gets a mention and no more: The move from assistants answering questions to assistants completing tasks is the next thing to plan for and the free curriculum has not caught up.
- Nobody teaches you the attribution mess properly: Every course tells you to measure. None of them spend enough time on why the same platform lands in Referral, Organic, Unassigned or Direct depending on the device.
- Platform-by-platform behaviour dates fast: Anything taught about citation patterns in a recorded video is a snapshot, so check the underlying research yourself.
- Each one is made by a company selling a tool that solves the problem the course just described. That is fine as long as you notice it while you watch.
More free AI search and GEO courses to explore
I have not been through these, so treat the list as a starting point for your own research. All of them advertise themselves as free and I want to confirm that for each one before I vouch for it the way I have for the 3 above.
| Course | Provider | Length | Certificate | Notes |
|---|---|---|---|---|
| AEO Fundamentals | HubSpot Academy | 2h 20m, 7 lessons | Yes, full certification | The most recognised brand offering a free certification and it stays product-agnostic |
| AI Search Optimization Masterclass | Surfer Academy | ~2h, 20 lessons | Yes, exam-gated | Weighted towards team structure, workflows and reporting, with templates included |
| AEO from Webflow | Webflow University | 1h 10m, 7 lessons | No | The most implementation-focused, with an AEO maturity model. CMS-specific in places |
| Learn AI Search | Aleyda Solis | Self-paced roadmap | No | Completely open with no sign-up, vendor-neutral and continuously updated |
| Introduction to GEO | The Open University | ~1 hour, 12 sections | Yes, OU Statement of Participation | The only university-hosted free option I found. Beginner level |
| AEO Academy | Conductor | 10+ guides by email | No | Enterprise framing: prompt-index building, AEO maturity, benchmarking |
Paid AI search courses worth considering
Two caveats before the table and both are load-bearing.
First, I have not taken any of these. They are here as additional consideration, based on what independent evidence exists.
Second, that evidence is thinner than it should be. There is no mature independent review infrastructure for AI search training yet. G2, Trustpilot and Class Central carry almost nothing at course level, the trade press has published no editorial reviews and the “best GEO courses” listicles that dominate the results are almost all written by AI search tool vendors or by the instructors themselves. I have excluded every one of those as evidence.
What is left is a small number of courses with verified-purchase reviews on platforms the vendor does not control, plus a few where the instructor’s track record is the evidence.
| Course | Price | Format | Independent evidence |
|---|---|---|---|
| From SEO to AEO & GEO, Jellyfish Training | £499 + VAT | 1 day, live online or in person | The strongest in this market. Feefo, a verified-purchase platform Jellyfish cannot edit, carries 4.9/5 from 877 reviews, including a course-specific review from January 2026 |
| AEO Masterclass, Mostafa ElBermawy on Maven | $499 | 2 live sessions of 90 minutes | 4.6/5 from 42 verified student reviews on Maven. Worth knowing he also sells an AEO tool |
| Optimize pages for AI search with GEO/AEO, CXL | $299 on demand, $599 cohort | 1h 57m, 14 lessons | Taught by Steve Toth, a verifiable practitioner. CXL as a platform scores 4.7/5 from 169 G2 reviews, though no independent review of this course exists |
| SEO Mastery: From Fundamentals to GenAI and GEO, Coursera | Coursera Plus, $59/month | ~28 hours across 4 courses | 4.7/5 from 73 verified learner reviews with named reviewers. Beginner level. Read the IBM branding as content production, since no university faculty is involved |
| Advanced GEO & AI Optimisation, brightonSEO | £900, includes conference access | 1 day, in person | Taught by Charlie Norledge and Petar Jovetic of Impression. The October 2026 session sold out, which is a market signal the vendor cannot manufacture |
One I would skip: The highest-rated Udemy AEO course carries more than 20,000 ratings, which looks conclusive until you notice the URL slug and archived listings show it was a general SEO course years before AEO existed. The ratings belong to a different curriculum.
How to choose an AI search course
If you have an hour and a half, take the Ahrefs course and stop there, because it is the best single use of the time.
If you have a day, do all 3 free ones in the order at the top, then go and run something. Pick 20 prompts that matter commercially, check what gets cited and pull a week of server logs. A course tells you what to look for, while the work tells you what is actually broken on your site and those are rarely the same thing.
If you are considering paying, my view is that the free curriculum currently covers most of the same ground and what your money buys is peer contact and a trainer who can answer your specific question. Both are worth paying for. Pay for the room and the person in it and take the certificate for the acronyms it puts on your CV.
AI search course FAQs
Are these AI search courses really free?
It depends which section you are reading. The 3 courses I have reviewed above are genuinely free with no card required: Each asks for an email address, Sitebulb’s includes a 60-day trial of its crawler and both Ahrefs and Semrush want a free account so they can track your progress. The 6 under more free AI search and GEO courses all advertise themselves as free and I have kept them in a separate list because I want to confirm that myself before vouching for them. The paid section costs what it says it costs.
Which free AI search course is best for beginners?
Ahrefs’ AEO Course. It spends its first 21 minutes on how AI search actually works before touching a single tactic, which is the right order and the one most courses get wrong.
Do AI search certifications help you get hired?
Not yet, though they can certainly help. No AEO or GEO certification carries hiring weight on its own, because the field is too young and no standard body has emerged. What has changed is the job specs. AEO and GEO have become the acronyms of choice for the people writing them, so having those terms on your CV and your LinkedIn profile can be what gets you past an ATS scanner and into a conversation, which is the part that actually matters. Take the course for the frameworks and the measurement sections and treat the badge as keyword coverage.
How long does it take to learn AI search optimisation?
You can cover the concepts in about 4 hours across these 3 courses. Getting useful at it takes a prompt set you track weekly, a log file you actually read and a few months of watching what moves. The fundamentals of technical SEO still do most of the work, which is why I would not skip them on the way in.
What is the difference between AEO, GEO and AI search?
AI search is the umbrella term for the surfaces themselves: ChatGPT, Perplexity, Claude, Gemini, AI Overviews and AI Mode. AEO (Answer Engine Optimisation) and GEO (Generative Engine Optimisation) are competing labels for the work of getting your brand cited, mentioned or recommended on those surfaces and in practice they describe the same tactics. Those tactics are a mix of old and new. Off-site coverage, digital PR, review sites, Reddit threads and listicle placements all shape what a model reads about you when it goes looking and SEOs have been doing that work for years. The genuinely new pieces are llms.txt and the agentic protocols, because there was nothing to optimise for there until LLMs existed. Underneath all of it sits ordinary technical SEO, which now carries more weight than it did, because a crawler that hits a 403 or an empty JavaScript shell cannot cite you at all. I go through each of those in AI Search 101.
Should I take a paid AI search course instead?
No, take the free courses first. When you have finished them, follow the people publishing on this continuously: Suganthan Mohanadasan, Aleyda Solis and Kevin Indig put out more genuinely useful free material in a month than most paid courses contain and it stays current in a way a recorded video cannot. Work through all of that and you will have covered most of what is knowable right now. If you still want more after that, by all means pay for a course. What your money buys is live access to a trainer and a cohort of peers and both are worth having.
Work with me on AI search
A course will tell you what to look for. If you would rather someone looked with you - at the crawler logs, the prompt set, or the reporting nobody can defend in a board meeting - tell me what you are seeing and I will tell you whether it is worth fixing.