More Changes In Search: 5 Top Takeaways For 2024 SEO

0:00Crystal WaddellWelcome back to the simple and smart SEO show podcast. I'm your host, Crystal Waddell. So glad that you could join me today.

0:05There are a lot of changes happening in search, right? It has been a big month.

0:10It's May 2024 and there were some rumors recently that chat GPT was going to be coming out with the search engine.

0:17They did come out with a faster version of their product and a bunch of other fun and silly slightly terrifying bells and whistles.

0:24But no search engine.

0:26Now Google, on the other hand, did release changes to their search engine, specifically for those of us in the U S who have been seeing that big paragraph at the top, that feature will become permanent for Google searches.

0:37So over the last few months, in case you missed it. it. This feature has been called search generative experience or S. G. E.

0:45And then this past week, Google renamed S. G. E. A. I. Overviews. So almost immediately I could see through this primarily as an ad play.

0:54And so I was pretty proud of myself. But that's one of. A few hundred reasons why AI overviews didn't really interest me.

1:01But sure enough, just yesterday, people in the SEO community on LinkedIn started posting about ads showing up in the AI overview.

1:08So again, I felt pretty proud of myself and said, Hey, I think I called that one.

1:12But.

1:13What does this mean for search and for SEO and for small businesses? I think it means that it's much more important to huddle up with your customers.

1:21And listen closely. And solve their problems. Stay close to your customers.

1:25And it's also important to continue to align those efforts with the technical expectations that search engines have.

1:33Now. More than ever. It's important for business owners to connect with technical expertise. It's going to keep you agile during these times of great innovation.

1:41And that's exactly what I do is small business owners every week who joined my SEO squad. So we meet weekly to talk about search engine optimization.

1:48We talk about branding, how to show up online and so much more.

1:51I'd love to invite you to sign up for my special three day challenge, Supercharge Your SEO, where I'll be giving you the foundational steps to future proof your optimization strategy.

2:00And I'll also be answering any questions about your website.

2:03Every day, live for three days. In a private Facebook group.

2:06You can find the registration link in the show notes and the three day challenge starts June 4th, 2024.

2:12But back to the essential questions!

2:15What do all of these changes in search mean and how can we align our optimization efforts with the technical expectations that search engines have?

2:23I believe those answers can be found within an article that I discovered written by Michael King.

2:29He wrote it on his website, Ipullrank.com, back in February of 2023.

2:34And i'll link to that in the show notes.

2:35Okay, so I have five top takeaways for you, and a few of these are going to get a little technical, a little nerdy, I love some nerdy.

2:43The first thing is that Google has been utilizing what is called a semantic model for almost a decade.

2:51Although, common SEO tactics focus on a lexical model.

2:56And in my opinion, looking over this information, the lexical model is where this obsession with keywords and keyword density and keyword stuffing began.

3:05And then with the semantic model, Google began looking for relationships between words.

3:10Like, why did certain combinations of words mean one thing, but then you change a word or two and the meaning of the sentence completely changes?

3:18So that's number one. Google has been utilizing a semantic model and it's not like brand new news.

3:23They've been utilizing this for 10 years. So why are we still talking about a lot of the same SEO tactics and components that deal more with a lexical model?

3:33That's been a struggle for me since I started learning about SEO, because I was like, some of this stuff just doesn't align, right?

3:39It just doesn't make sense or just doesn't fit.

3:42So that was my first top takeaway.

3:44The second one is that optimization is no longer two dimensional.

3:48So what I mean by that is in order to understand these relationships between words, somebody really smart created something called a vector space model.

3:59And I asked ChatGPT to explain the concept to me like I was a third grader. And so I'm going to read what Chat told me.

4:05" Imagine that you have a big box of crayons and each crayon is a different word. Now, imagine we want to sort those crayons, not just by their color, but by how often you use them and which ones you often use together.

4:20In the vector space model, we're doing something similar, but instead of crayons, we have words. And we turn each word into a special kind of number list, like a magic code. And that magic code tells us a lot about the word. This magic code helps us see how often the word is used and which words it likes to hang out with.

4:39By looking at these number lists, we can understand how words are related to each other. Knowing that blue and green are both cool colors on the color wheel.

4:47Now, one of the things I also liked was that, if you could picture these within a cube. That also helps you understand the three dimensional ness of them.

4:57So my second takeaway is that optimization is no longer two dimensional, which also. Just completely comforts me because like I've told my dad so many times, I'm like this and this are related. And it's not just a straight line. Sometimes it goes up, sometimes it goes over, sometimes it goes down.

5:13And so optimization online takes into it takes into consideration several different factors. So that was takeaway number two. The optimization is no longer two dimensional.

5:24Number three. Okay. This is something that is like far beyond my pay grade.

5:28But I feel like it's an important to understand because it also gives us a little bit of insight into how some of these large language models work. And why they work the way they do.

5:37So if we can understand why they work the way they do, then we can position ourselves within them to show up for results.

5:44So the third takeaway I had was just the discovery of something called word embeddings, which are sparse word embeddings and dense word embeddings.

5:54So again, I go back to my favorite assistant slash tutor. And Chat's example was this.

6:00"Let's think about two ways of building a model of your favorite animal. Let's say a lion. Sparse embeddings are like using Lego bricks to build the lion. You have a specific number of bricks and each one goes in a designated spot, according to the instructions.

6:14The final model is made up of mostly empty space because you only need. use a few bricks for the key parts of the lion. This method is straightforward, and it doesn't use many pieces, which makes it efficient and easy to handle, but it may not show all the small details and character of the lion.

6:32On the other hand, Dense embeddings are like using clay to sculpt the lion. You shape and mold the clay to capture not just the basic shape, but also the texture of the fur, the expression in its eyes, and the curve of its muscles. It's much more detailed. This method uses a lot more material and effort, but it creates a much more detailed and lifelike representation of the lion.

6:56In this analogy, sparse embeddings give you a clear but basic outline with lots of gaps, while dense embeddings provide a fuller, more nuanced depiction that brings the words, or in this case, the lion, to life in a much richer way.

7:09Because of dense embeddings, when you ask Google a question now, it's much better at showing you the exact piece of information you need, like a specific sentence or paragraph.

7:19And this is why when you look something up, you often see those little boxes at the top of the search results, sometimes featured snippets, that seem to perfectly answer your questions. It's like Google has become much smarter at finding the needle in the haystack. Thanks to these advanced methods."

7:34So when you're aware of the advanced methods, you can then align the content that you're creating with the advanced methods.

7:42And the key words that I pulled out of there is that your content needs to provide a fuller, more nuanced depiction. That brings the words to life.

7:53In a much richer way, and this is why expertise is so important and an experience.

7:59Because you when you have direct experience with something, you can provide a little bit more detail, a little bit more of context where people.

8:09When they actually do the thing or they're, learning about the thing, they're like, Oh yeah, I know that person knows what they're talking about because yes, that does happen when you do this.

8:16For someone who's never, done anything with that particular topic. It becomes a lot easier to see right through it, right?

8:22Because they don't know the nuanced things that happen. So number three, word embeddings, sparse and dense.

8:30Those are important to understand because that is how Google's working now. It's no longer a shallow experience. They're looking for a richer, deeper experience.

8:40And number four, where we are at now. And this is what Michael King called leaving SEO behind in his article. I found it Fascinating.

8:49So when I threw it into ChatGPT, ChatGPT explained it back to me like this.

8:55Imagine if you had a magic notebook that could learn and remember not just words, but also what those words mean and how they connect to other words.

9:04Back in 2013, a group of smart people named Thomas McAuliffe, Kai Chen, Greg Corrado, and Jeff Dean came up with a way to make a computer notebook just like that, which they called Word2Vec, which was short for Word2Vector.

9:20Imagine you have a toy robot named Word2Vec that helps you guess what words can come before and after a word that you tell it.

9:28For example, if you say sun, it might guess hot and shine.

9:32Because those words often appear near sun, but this robot has a little problem. It only looks at the few words right next to the word you gave it and it doesn't really understand the whole sentence or story.

9:44So sometimes it might not guess the best words because it's not looking at the bigger picture.

9:49Despite this, Word2vec helped Google start to understand not just the words themselves, but what they mean together, like understanding the whole idea of a story instead of just the words.

9:59And this was a big change from just counting words. Sound familiar? Which is what a lot of tools used to do to help websites show up in searches.

10:07So I just want to say many thanks to Michael King at iPullRank. com for writing this article. I can't believe it came out over a year ago.

10:15But it really does help pull together some of these ideas And if you're like me, where you don't have a strong technical background, I suggest going through it paragraph by paragraph and putting it into chat GPT and asking chat to explain it to you like you are a third grader.

10:29And so coming to the last point of the day for this episode, number five, search is getting deeper. And I think that search is getting deeper because language is deep.

10:39Like the whole study and thought of language, think about it. We have verbal language, we have nonverbal language, we have a variety of languages, we can shout, we can whisper, those are the nuanced things, right?

10:53And there's a phrase that I hear Quite a bit from people when I'm talking about anything that I'm passionate about. Like it can be something big, it can be small, but a lot of times people will say, crystal, it's not that deep. I'm like, you know what?

11:04I think that search is that deep and that it's getting deeper.

11:08And I think that SEO practices over the years were created because the systems around search were fairly shallow. Keywords connected the dots, but that's a two dimensional tactic in a three dimensional ecosystem that we call the internet.

11:21We've got to make sure that the language of our businesses online can translate efficiently to the language of search.

11:28And the good news, we live in that three dimensional world. We use language every day. It's just a mindset shift from shallow SEO. to semantic SEO.

11:39And come to find out my son's fifth grade ELA assignments are probably going to do me more good than they even did for him.

11:46So final point, search is getting deeper because it's learning language.

11:52It's learning how to communicate and communication is multifaceted, right? So it makes sense that search is becoming multifaceted as well.

12:01So thank you for joining me for this episode on the changes in search.

12:06Next week, I'll be giving you another look inside the SEO squad. So you can consider joining for the support. You need to navigate the wild web that is unfolding in front of us.

12:14The link for the three day challenge is in the show notes. So please sign up if you're interested.

12:18And if you have any questions.

12:20Or any comments, feedback, you can now send me a text message. Thank you. Buzzsprout!

12:25You can use the link in the show notes as well. So that's it for today and I'll see you next time.