Last Updated on April 17, 2025
SEO means moving beyond simple keyword optimization. SEO means much more than keywords. Make your content meaningful and connected with the reasons behind what people search. (Nothing new about that). Think about the people, places, things, or ideas related to your topic, and how they connect. (Old news too).
Are you interested in watching the podcast? You can check that out right here: Episode 535 SEO With Warren Laine-Naida Using Ai To Increase Your Search Visibility
Entertain. Engage. Educate. Empower.
Focus on problem-solving instead of chasing keywords; you’ll help search engines and AI systems better understand your content. This should result in your content appearing more often in search results and being used as a source for AI-generated answers.
The basics still matter, even if the tools have changed. “Create content that builds trust and makes connections.” Ross Simmonds
SEO today means connecting the dots: Google, Bing & co., and the many AI search bots. (AI. That is still somewhat new and of growing importance).
If you are new to the topic of AI and SEO and want some context, please read more in my previous articles: https://warrenlainenaida.net/category/artificial-intelligence-online-marketing/
Check it out: Google Search Central Live NYC: Insights On SEO For AI Overviews: https://www.searchenginejournal.com/google-search-central-live-nyc-insights-on-seo-for-ai-overviews/542684/
AI can assist with content generation but should complement your creativity rather than replace it.
AI is reshaping SEO by shifting the focus toward user experience and accessibility. I advocate for optimizing content for people, social media algorithms, traditional search engines, and AI-driven search engines.
AI can aid in keyword research and topic ideation, making content more relevant and discoverable, but you are responsible for the content and the conversation.
Ultimately, a balanced approach is the best. Leverage AI for efficiency while maintaining the human touch that makes your content engaging and valuable.
What are LLMs? What Do They Have to Do with SEO?
Large Language Models (LLMs) are AI systems trained on vast amounts of text to understand and generate human language. They work by predicting what words should come next in a sequence.
AI Chat LLMs (like Claude, perplexity, etc.) can:
- Answer questions
- Generate text
- Summarize documents
- Translate languages
- Write creative content
- Help with coding
Think of LLMs as advanced autocomplete systems that have learned patterns from billions of examples of human writing, allowing them to produce helpful responses.
LLMs and SEO connect in several ways:
- Content creation – LLMs help generate SEO-friendly articles and website content
- Search engines now use LLM technology to produce semantic results
- LLMs can help optimize existing content for better search performance
- They assist with analyzing keywords and understanding search intent
The rise of LLMs is pushing SEO toward more natural content rather than keyword-focused content.
An SEO’s guide to understanding large language models (LLMs) https://searchengineland.com/seo-guide-large-language-models-413227
AI and SEO: How You Can Put Them Together
Here are the core concepts behind AI and SEO and how you can apply them:
1. From Keywords to Semantics and Entities
- This really isn’t your parent’s SEO anymore:
- Old SEO: All about placing the right keywords to match search queries.
- Modern SEO: It’s now about understanding meaning by recognizing entities and how they relate to each other. Both search engines and LLMs are moving beyond just matching keywords.
2. Defining Your Business Domain
- Identify your client’s business (e.g., Adult Education).
- Define related entities:
- Industries served (healthcare, technology, hospitality …).
- Courses required, interested target groups
3. Using Google and Bing for Query Discovery
- Leverage Google and Bing as research tools to find queries related to your entities.
- Gather data from:
- Topic filters
- People Also Ask
- People Also Search For
- Image Search tags
- Google Search Console
- Bing Webmaster Tools
4. Structuring Content for Entity Recognition
- Structured Data (Schema Markup): Use schema to define entities and relationships for search engines and LLMs clearly. For example, use the LocalBusiness schema for a school offering adult training courses. (Schema Markup is included free in both the Yoast and RankMath SEO plugins. RankMath offers more in the free version than does Yoast.)
- Topical Authority: Cover related topics to establish your site as an authority on a given entity. If your topic is “WordPress Development,” write about coding, online shops, websites, etc.
- Internal Linking: Link-related pages to strengthen the connections between entities. A page about WordPress theme development should link to a review of top WordPress themes. (Internal links are backlinks – did you ever think of that? SEO gold.)
- Entity-Rich Language: Use natural language that incorporates relevant entities. Instead of writing “WordPress is popular,” write “WordPress is Open Source, and is known for its ease of use. 50% of the websites currently using a CMS depend on WordPress.”
- Optimize for Featured Snippets, AI Search Engines, and Direct Answers:
- Use Q&A format.
- Use lists and bullet points.
- Provide clear, direct explanations.
- Chunk your content. Why Content Chunking For Ranking AI Overviews Makes Sense
“Optimize for AI and voice search: With the rise of AI-powered search engines, it’s crucial to structure (chunk) content in a way that’s easily digestible for both humans and machines.” Warren Laine-Naida
5. Understanding Embeddings
- What are Embeddings? They’re mathematical representations of words, phrases, and entities that capture their meaning and relationships. Similar meanings are grouped in a multi-dimensional space.
- Understanding Word Embeddings in SEO: A Complete Guide https://jbh.co.uk/blog/seo/understanding-word-embeddings-in-seo/
- Why Embeddings Matter: LLMs use embeddings to understand context and meaning, handle ambiguity, and build a “memory” of conversations.
- Embedding Optimization:
- Use entity-rich language.
- Build semantic relationships through internal and external linking.
- Optimize for entity-based structured data.
- Focus on semantic search, not just keywords.
Embeddings and SEO Simplified
Embeddings are how modern search engines understand meaning, not just keywords. Embeddings are like digital maps that help computers understand how words and ideas relate to each other.
Think of it as giving each word coordinates on a map where similar concepts sit close together (like “coffee” near “latte”) and different concepts sit far apart (like “dog” and “car”).
Search engines use embeddings to understand meaning rather than just matching exact keywords, which is why searching for “caffeinated beverages” can return results about “coffee shops” even when those exact words aren’t used.
Old SEO: Repeat exact keywords like “best coffee shops in Seattle” multiple times.
New SEO: Write naturally about the Seattle coffee scene, quality, atmosphere, etc. Search engines now understand that content about “outstanding cafes in the Emerald City” relates to the same topic.
This matters because:
- You can write for humans, not algorithms
- Content depth beats keyword repetition
- Related topics and natural language work better than forced keywords
Bottom line: Search engines now map related concepts together in a “meaning space,” allowing them to understand what users want even when the exact words don’t match.
6. Retrieval-Augmented Generation (RAG)
What is RAG? https://cloud.google.com/use-cases/retrieval-augmented-generation?hl=en
- Understand how LLMs use RAG:
- Receive a query.
- Search a database of embeddings for relevant information.
- Pull the most relevant pieces and generate an answer.
- If your content is well-structured and entity-rich, it’s more likely to be retrieved and used in AI-generated answers.
Google researchers refine RAG by introducing a sufficient context signal to improve response accuracy: https://www.searchenginejournal.com/google-researchers-improve-rag-with-sufficient-context-signal/542320
7. Content Formats
You probably don’t need to hear, yet again, how important content is to your SEO. What is new is how you format your content for AI citations.
Content chunking is a technique used to break down large amounts of information into smaller, digestible pieces (or “chunks”) to improve readability, comprehension, and user engagement.
You already write for human readers, and for Google’s Featured Snippets – you should write for AI citations too.
Read more here about Why Content Chunking For Ranking AI Overviews Makes Sense.
The Best SEO is Great Content! If your website is already SEO-friendly, and you have the technical stuff taken care of, you should be seeing results. If you’re wondering why you aren’t getting the traffic you had hoped for, perhaps you’re missing the right content.
- Analyze SERP patterns to identify preferred content formats.
- Enrich content with images, videos, and specific text formatting.
- Consider whether a visual format might be more relevant than text.
- Don’t forget to chunk
“If someone is looking for the best running shoes, they’ll be shown blog posts. If they’re looking for the best Nike running shoes, they’ll be shown a product page. What happens with Google also happens with ChatGPT. You (still) need product pages and blog posts.” Bridget Willard
8. Measuring Success
Measure your success with organic traffic, rankings, engagement, backlinks, and conversions. For AI-driven search, track mentions in AI responses, AI-driven traffic, structured data, user intent, and brand visibility.
- Focus on visibility for relevant query sets, not just keyword rankings.
- Track visibility in:
- Classic search results
- Featured snippets
- People Also Ask
- Things to Know
- Image Search box
- Videos
- Branded People Also Search For
- LLM responses (ChatGPT, Perplexity, Gemini, AI Overview)
Wrapping Up
SEO today means shifting from being keyword-centric to a more holistic approach. Instead of researching keywords, focus on understanding and catering to the interests and needs of your target audience.
By focusing on semantics, entities, and relationships, (and by embracing AI tools) you can (even better) optimize your content to shine in both traditional search and the evolving landscape of LLM-powered search and content generation.
AI LLM SEO Frequently Asked Questions
LLMs (Large Language Models) are AI systems that have been trained on huge amounts of text to understand and generate human language. They work by predicting the next word in a sequence. They are relevant for SEO because search engines use LLM technology to deliver more semantic search results. In addition, LLMs can help with content creation and optimization, keyword analysis and understanding search intent. This leads to a shift from purely keyword-focused optimization to more natural and content-rich texts.
Traditional SEO focused primarily on the placement of relevant keywords to match search queries. Modern SEO goes beyond this and aims to understand meaning by recognizing entities (people, places, things, ideas) and their relationships to each other. Both search engines and LLMs strive to capture the semantic meaning of content rather than just looking for exact keyword matches. The focus is now on creating content that solves problems, builds trust and creates connections.
In order to define their own topic area, companies should first identify their core business and then determine related entities. This includes industries served, courses offered or interested target groups. Google and Bing can serve as research tools for discovering relevant search queries. Data can be obtained from topic filters, “People Also Ask” boxes, “People Also Search For” results, image search tags, Google Search Console and Bing Webmaster Tools.
Structured data (schema markup) helps search engines and LLMs to clearly understand the content of a page and the entities it contains as well as their relationships. They serve as a kind of vocabulary to explicitly characterize the meaning of content. Topical Authority is achieved through the comprehensive treatment of related topics and signals to search engines that a website is an authority in a particular subject area. Both help content to be better understood and used more frequently in search results and by AI systems as a source of answers.
Embeddings are mathematical representations of words, phrases and entities that capture their meaning and relationships to each other. Similar meanings are grouped in a multidimensional space. LLMs use embeddings to understand the context and meaning of texts, manage ambiguity and build a “memory” of conversations. For SEO, this means that using entity-rich language, building semantic relationships through internal and external linking, and optimizing for entity-based structured data will become more important as search engines can now connect concepts in a “meaning space”.
Retrieval Augmented Generation (RAG) is a process in which LLMs access a database of embeddings to find relevant information to a search query and use it to generate an answer. To optimize content for RAG, it is important to structure it well and enrich it with many relevant entities. This makes it more likely that the content will be recognized as relevant by LLMs and used to create AI-generated answers.
Content chunking is a technique in which large amounts of information are divided into smaller, more easily digestible sections (chunks). This improves readability, comprehension and user-friendliness. In terms of AI overviews, this structuring is crucial as it makes it easier for AI systems to quickly extract relevant information and use it for direct answers. The use of Q&A formats, lists, bullet points and clear, direct explanations helps to optimize for both human readers and AI-powered search engines.
In addition to traditional SEO metrics such as organic traffic, rankings, engagement, backlinks and conversions, additional metrics are relevant for measuring success in the age of AI-supported search. These include mentions in AI responses, traffic coming directly from AI search results, correct implementation of structured data, understanding of user intent and overall brand awareness in this new search environment. The focus should be on visibility for relevant search query groups, not just individual keyword rankings, and performance in classic search results, featured snippets, “People Also Ask” boxes, “Things to Know”, image search, videos and LLM answers (such as from ChatGPT, Perplexity, Gemini and AI Overview) should be monitored.
Read More:
- https://autonomoustech.ca/blog/beyond-keywords-llms-changing-seo
- https://www.lumar.io/blog/industry-news/ai-search-seo-for-llms-ai-overviews/
Article image thanks to Freepik: https://www.freepik.com/pikaso/ai-image-generator 21.03.2025.10:34