TLDR: Search is changing. Constantly. Learning how to optimize videos for search engines and AI is crucial. Many old practices hamstring the performance of your creative work. I’m going to walk you through the changes in how people are finding your brand and how you can use video to earn business online.
A decade ago, everyone swore by video tags.
They’d place tons of keyword-enriched phrases into their descriptions.
Brian Dean and the Backlinko team set the record straight in 2017, publishing what they found after reviewing over 1.3 million YouTube videos. Unfortunately, many people still don’t get it:
- There is zero correlation between keyword-optimized video descriptions and rankings
- Tags with keywords do little if anything to boost your video’s performance
The algorithm progressed. YouTube, in particular, started getting the context of your videos without all of the old ways of doing search engine optimization (SEO).
These days, generative engine (GEO) and answer engine optimization (AEO) send curveballs past the plate of search optimization that Clayton Kershaw would envy. What’s really happening is that the powers that be want you and me to answer the questions people ask.
Connection. Storytelling. Real and helpful content. These fell to the wayside when people created optimized videos solely for search engines rather than for people.
Does SEO Even Matter Anymore?
Of course. AI won’t kill it off, but it has, and will, change a ton.
The bigger issue I’m seeing in the video production world is that we've gotten away from the reason SEO came into existence in the first place. Google, Bing, and all the others were hoping that we’d create helpful content.
RELATED: Why Your In-House Team Can’t Create The Helpful Content Your Audience Needs
But you know basic human psychology. Most seek ways to game the system.
Doing SEO right is about connecting people with content they can use to improve their lives, learn a new skill, or overcome a problem.
Traditional SEO still delivers results when it’s done right.
It’s also setting the foundation for all the work you need to do to get your videos to drive clicks in the world of AI search.
Have you ever heard of the term “follow the money?”
It’s a term all of us in video production need to keep in mind. Google invests tons into its search engine results pages (SERPs). Plus, Google is a subsidiary of Alphabet Inc., and purchased YouTube for $1.65 billion in 2006. All of that to say, the old ways aren’t gone, but they’re now just one piece of a large puzzle.
Google, Meta, and Microsoft all reported during recent earnings calls that they would spend billions of dollars on capex in 2025. They’re spending most of that money ($320 billion) on all the buildings, materials, machines, and housing to host AI infrastructure. Check your local newspaper. There may even be an AI data center being built near you.
Not too far from our offices in Lexington, bulldozers and excavators prep the ground for a hyperscale data center.
The AI models these centers house are trained on SEO data, so we have to keep the fundamentals in mind. There will be some extra work for you to do with your videos, though, now that people use AI to get answers to their questions.
A Quick Reminder on the Fundamentals of Video SEO
You can still create videos that rank through strategic keyword targeting and smart technical execution. Here is a quick list of the most important items to check off:
- Front-load keywords in your title: Put your primary keyword in the first 60 characters so it doesn't get cut off in search results
- Optimize the first 157 characters of your description: This is what shows up in search previews, so make it count with keywords and value proposition (again, don’t just stuff them in if they don’t make sense with the message of the video)
- Create custom thumbnails: Design eye-catching visuals that drive click-through rates, not just the auto-generated frame YouTube picks
- Add full, accurate transcripts: Not auto-generated captions, but properly formatted transcripts that search engines can crawl
- Implement VideoObject schema markup: Structured data that tells Google exactly what your video is about, its duration, upload date, and thumbnail
- Use strategic file naming: Name your video file "keyword-topic-description.mp4" before uploading, not "IMG_1234.mp4"
- Add chapters and timestamps: Break longer videos into searchable segments that appear as jump links in search results
- Create a video sitemap: Submit this to Google Search Console so all your video content gets indexed properly
- Host on YouTube, embed on your site – YouTube gives you distribution power while embedding builds your own domain authority
- Leverage end screens and cards – Keep viewers engaged with your content by recommending related content strategically
These elements help search engines understand where your video fits in with a user's search query. Get these right before you even think about GEO or AEO.
Video Optimization for GEO and AEO
Optimizing your videos for large language models (LLMs) seems like an overwhelming task. Part of the problem is that there are so many terms floating around. You’ll hear GEO and AEO for sure. You might also hear Generative Search Optimization (GSO) and AIO (AI Search Engine Optimization). It’s the alphabet soup of search. But they all describe the same core shift: optimizing for AI platforms like ChatGPT, Perplexity, and Google's AI Overviews.
Traditional SEO gets you ranked on search engine results pages (SERP). GEO and AEO ensure AI crawlers understand your content well enough to cite it in synthesized answers. The difference? People type keywords into Google. They write complex, conversational prompts into AI engines.
Your videos need to fit both models.
Here’s how the two look side by side:
Understanding How AI Models “See” Your Videos
Large language models rely almost entirely on text-based signals surrounding your video content. The transcript is their primary data source.
AI crawlers also evaluate the page context around your embedded video, structured data markup, and how clearly you signal your topic and expertise.
GEO Optimization Tactics
Getting cited in AI-generated answers requires a different content approach than traditional SEO. Here’s what you need to do to make sure you’re ready for AI search now and in the future:
Create Clear Scripts: AI models love clear transcripts with mentions of brands, people, places, and concepts. While not all current models may “listen” to the audio, it’s important to note that some multimodal LLMs do have this capability (and you can bet this technology will become even more capable). At StoryVid, we double-check our scripts to make sure we’re saying out loud what is important and relevant.
For example, in our SpaceTrek promo for Morehead State University, we included location and program details. Not only are they helpful for the viewer, but those details will be something LLMs pick up as they advance.
Structure Content for Extraction: AI platforms favor content they can easily pull and synthesize. Format your video page with clear bullet points, numbered lists, and FAQ-style sections.
Use Prompt-Friendly Language: People write conversational, long-tail queries into AI engines. Optimize your spoken content and transcripts for how people actually ask questions. Instead of targeting "video SEO tips," create content that answers "how do I get my product videos to show up in ChatGPT searches?"
Publish Comprehensive Summaries: Don't just embed a video and walk away. Add 300-500 words of context that summarize key takeaways, include relevant data points, and link to related resources. This text layer dramatically improves your chances of citation.
Testing Your GEO Performance
You can't optimize what you don't measure.
Start by searching for your target topics in ChatGPT, Claude, and Perplexity. Does your video appear in results? Are you getting cited? Track this manually at first, then monitor referral traffic from AI platforms in your analytics. The signals are there.
There are some limitations. The nature and speed of LLMs mean it’s hard to pinpoint exactly what moved the needle. There are some tools out there, like Clearscope, Rankscale, Peec AI, and Profound, that can help you see if your videos and content are being found when people use AI to search. But it’s not a foolproof process.
Making Your Videos Citation-Worthy
Mentions help. Citations convert.
So what’s the difference?
If you get mentioned by AI, your brand is shared, but there is nothing for a user to click. If you get a citation, an LLM or overview provides a citation in the form of a link.
In your videos, give data points and information that is easy for AI to digest. ChatGPT, Perplexity, and Claude want specifics that are super focused on what a user is hoping to find.
I’ve written about some of the video marketing metrics you should be tracking in a previous post. These are the kind of things to include in your video.
Say "Our clients see an average 34% increase in engagement" instead of "the video is performing well.”
Also, display data visually on screen so it appears in your transcript when you describe it.
AI engines reward depth.
Don’t Forget the Technical Elements
Make sure to add FAQ schema markup on your video pages, especially for how-to content. Add HowTo schema for tutorials. If you don’t know how to do this, go to schema.org to learn how all of these work. Use Clip markup to highlight key segments as well. To make it all work together, add a breadcrumb structure, so AI understands where your video fits in your content hierarchy.
RELATED: The Future of AI in Video Production
Once They Click, SXO Comes Into Play
Getting your video to rank or getting cited by AI is only half the battle. What happens when someone actually clicks through? That's where Search Experience Optimization (SXO) comes in.
Search engines have evolved beyond simple ranking signals. If people click your video and immediately bounce, or if they watch for 10 seconds and then leave, those signals tell search engines that your content didn't deliver on its promise. Going back to traditional SEO, we’d call that pogo sticking, when a viewer or reader has to leave your page because it didn’t give them the answer they were looking for. That’s bad news for your brand and your site.
Key SXO Metrics for Video
The watch time percentage is big. Aim for at least 50% completion rates. If you're creating a five-minute video and people consistently drop off at 45 seconds, you've got a content problem, not a search problem.
Track bounce rates from your landing page and time on page after video completion.
Match video length to search intent. Someone searching "what is video SEO" wants a 90-second overview, not a 20-minute deep dive. Someone searching for "complete video SEO tutorial" expects long-form content.
Include jump links to specific sections so viewers can navigate to what they need. Include clear calls to action, related content recommendations, and opportunities for comment engagement.
The goal is simple: turn every visitor into a satisfied user who got exactly what they came for.
A Unified Video Optimization Framework
Pre-Production: Start with keyword research across all platforms, traditional search engines, and AI platforms. Structure your content plan around questions people actually ask in long-tail queries. Optimize your script to include quotable statements, specific data points, and mentions of entities. Think about what AI will extract before you ever hit record.
Production: Deliver key points clearly and verbally. Don't rely on viewers to make the right inferences. Display important data and statistics on screen.
Post-Production: Create accurate, properly formatted transcripts. Place keywords strategically in your title, description, and on-page content without keyword stuffing. Format your content for multiple platforms.
Distribution & Monitoring: Publish strategically across platforms. Track performance in traditional search, AI platforms, and user behavior metrics. Iterate based on what's working. Test queries in ChatGPT, Perplexity, and Google AI Overviews. Monitor which platforms drive the most valuable traffic, not just the most traffic.
This isn't a one-and-done process. It's a continuous cycle of creating, measuring, and improving based on real performance data across all search channels.
