Hello and welcome to. Seo is not that hard. I'm your host, ed Dawson, the founder of keywordspeopleusecom, the place to find and organise the questions people ask online. I'm an SEO developer, affiliate marketer and entrepreneur. I've been building and monetising websites for over 20 years and I've bought and sold a few along the way. I'm here to share with you the SEO knowledge, hints and tips I've built up over the years. The SEO knowledge, hints and tips I've built up over the years.

Hello and welcome back to SEO is not that hard. It's me, ed Dawson, here hosting, as usual, and today we're going to talk about embeddings and give sort of an introduction to embeddings and what they are, because this is an area, a topic, that's basically becoming more and more important in the world of SEO. So let's sort of dive in. So you might be wondering so what exactly are embeddings? I mean, it may be a term that you've heard in the context of AI and machine learning, maybe even in the context of SEO, but it's not always clear what it means. So today I'm going to go through and unpack it step by step and explore how you can use them to improve your SEO. So by the end of this episode, hopefully, you'll have a solid understanding of what embeddings are, how they work and how they can help you rank better in search engines. So let's start with the basics. At its core, an embedding is a way of representing data so words, sentences, images, even users has vectors in a multi-dimensional space. Now that might sound complicated and a lot of jargon, so let me break it sort of down into pieces for you. So imagine, if you can, a regular 2d graph. So this is the sort you might have seen in school, where you have an x-axis and a y-axis and each point on the graph represents a pair of values like 3 comma 5 or 1 comma minus 2. That's, you know, a point in a two-dimensional space. However, in reality, data is way more complex than just two numbers in most cases. So, for example, you think about words. They aren't just defined by two characteristics, they have many different aspects of context, meaning, usage, frequency and things like that. So instead of having just two axes, the X and Y axes we use many axes to represent different features of data, and this creates a multidimensional space. So you might have hundreds of axes to represent all the different nuances of a word or piece of content, and each word or phrase becomes a point in this space, represented by a vector, which really is just a fancy name for a list of numbers that describe where that word is located in this space. Now, the key thing to understand in this is the closer two points are in this space, so two different vectors, the closer they are, the more two points are in this space, so two different vectors. The closer they are, the more similar they are in meaning. So, for example, if you've got a multi-dimensional space, the words king and queen might be very close to each other because they share sort of many semantic features. So they're both royal titles, for example, um, and you know, they're both chess pieces, for example. But on the other hand, king and car would be much further apart, because a king and a car are very different, they're very different aspects to what they're about. So they will be very far away in this multi-dimensional space.

So you're now probably wondering well, why? Why bother, you know, doing all this? Why bother going to representing words and context and content in multi-dimensional space? Well, the beauty of these embeddings is that they allow machines like google search algorithms to understand relationships between words and concepts far beyond simple sort of keyword matching. So in the old world, best yo yo used to focus heavily on exact keyword matches. So if someone searched for, say, the best pizza in London, you'd make sure to repeat best pizza in London several times on your page. Back to the whole keyword stuffing days, and that was the core of SEO many years ago. But now search engines are much smarter, thanks partly due to embeddings. So instead of just looking for exact matches, search engines they can now understand that top pizza places in London or where to find great pizza in the city are very similar to best pizza in London. So they don't have to be exact matches, because the embeddings capture the underlying sort of meaning and words and phrases and it places them close to each other in this multi-dimensional space. So they're.

So how does this work in practice? Right for seo. So let's say you're writing a blog post, say, about hiking boots. In the past you'd focus on repeating that exact phrase throughout your content. Well now, thanks for embedding, search engines have concerned that hiking shoes, footwear for hiking and even sort of best outdoor boots for trails are related. So these different phrases, they're not sharing the exact same wording, but they'll be close together in that multi-dimensional space. And that was trying to explain, and that's just one example.

This idea of capturing meaning through relationships in multi-dimensional space can be used to, you know, extend to all sorts of content, so not just words. So we can use embeddings to represent an entire document or user behavior patterns. We could compare an entire document to a search query so we can see which document in a database or in a search engine is the most relevant, the closest in multidimensional space to a search query, to determine the most relevant document to a query, for example, and this allows the search engine to deliver better, more relevant results for you know, huge, wide range of queries. So now let's think in practical terms. How can you know we, as seos or site owners, start to use embeddings to our advantage? So first, you know we need to talk about content optimization.

So embeddings allows to move beyond simple keyword, you know stuffing. Essentially, instead of focusing on the exact keywords focus, we can focus on creating content that's semantically rich. So this means covering related concept, answering user questions and using varied language. This goes back to the whole concept for keywords people use making sure that we cover all those answers and all those questions that people are asking around a topic. To get that, you know that, get the richness in our, in our documents, to sort of improve our embeddings. So when you do this, we sort of we naturally feed the search engines embedding models, helping us to rank our content better for a wider variety of queries. So, for instance, say we're writing a um article about the you know the best restaurants in rome which might be useful to me because I'm heading there soon. You can include related terms like top places to eat Italian cuisine in Rome or even family-friendly dining. So Google's embedding models will understand that all these phrases are related and relevant to users' queries.

The key area where embeddings can help is in internal linking and site structure. So by using embeddings you can analyze how closely related your pages are to each other beyond just keywords, and this can sort of help us help guide in sort of making more meaningful internal links. And it doesn't just stop there with content. We can also look at embeddings and how they can help us with personalization and user intent prediction. So imagine being able to predict what a user is likely to search for next or which piece of content they're likely to engage with, based on the semantic relationships between the pages they've already visited. So this is like a new frontier of SEO, really, and embeddings are at the heart of that. So by understanding what content your audience is finding most relevant, you can then start to tailor your future content strategies much more effectively based around the embeddings that you learn, around that kind of personalization and the way your users interact with your content, sort of group those similar user behaviors and then sort of create that content that anticipates needs rather than just reacting to how people are searching at the moment.

We've been spending a lot of time delving into embeddings here at QWI's People Use. Obviously, with our new Google Search Console integration, where we can understand sites and the content on your site, we're now looking at how we can use embeddings to sort of grade content, score content and look at how to improve content, because obviously, when you have an embedding, you've then got a vector, a number, that represents how well that piece of content is performing next to other pieces of content in the context of the semantic area you're operating in and it can help on how to improve content, can help make suggestions on content areas that you're missing. It can help work out whether your content is matching well to the queries that you're trying to target. So there's loads of stuff that we think we can do with embeddings. So that's why I wanted to give this little primer today on what embeddings are, so you get that kind of core concept of right.

I understand what an embedding is.

It's a way of encapsulating the content and the context and the semantic meaning of any piece of content, any data, and then being able to relate and compare to other pieces of content, other pieces of data, seeing how closely matched they are in this vector space, which helps us work out these relationships in a much more specific, repeatable and algorithmic way than just making suggestions based on the old way of doing things, the more keyword stuffing kind of way, where we were just working on principles that have worked before but there was no real maths or logic behind it that was repeatable, whereas with embeddings we can actually tie this up and be approach it much more scientifically. So, yeah, hopefully we'll see much more on this from us soon and, yeah, I hope you found that interesting and I hope you're now much clearer on what embeddings are. They're not really that complex once you understand the basic principle and it's very elegant if you think about it. But hopefully it is now a bit clearer for you and, yeah, hope you've enjoyed this and I look forward to seeing you next time.

Before I go, I just wanted to let you know that if you'd like a personal demo of our tools at Keywords People Use that you can book a free, no obligation one-on-one video call with me where I show you how we can help you level up your content by finding and answering the questions your audience actually have. You can also ask me any SEO questions you have. You just need to go to keywordspeopleusecom slash demo where you can pick a time and date that suits you for us to catch up Once again. That's keywordspeopleusecom slash demo and you can also find that link in the show notes of today's episode. Hope to chat with you soon. Thanks for being a listener. I really appreciate it. Please subscribe and share. It really helps.

Seo is not that hard. It's brought to you by keywordspeopleusecom, the place to find and organize the questions people ask online. See why thousands of people use us every day. Try it today for free at keywordspeopleusecom To get an instant hit of more SEO tips. Then find the link to download a free copy of my 101 quick SEO tips in the show notes of today's episode. If you want to get in touch, have any questions, I'd love to hear from you. I'm at Channel 5 on Twitter. You can email me at podcast at keywordspeopleusecom. Bye for now and see you in the next episode of SEO. Is not that hard.