Speaker 1: 0:54

Hello and welcome to. Seo is not that hard. I'm your host, ed Dawson, the founder of keywordspeopleusecom, the place to find and organize the questions people ask online. I'm an SEO developer, affiliate marketer and entrepreneur. I've been building and monetizing 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.

Speaker 2: 1:18

Hello and welcome back to SEO is not that hard. It's me here at Dawson hosting, as usual, and today I'm going to be talking all about search GPT, which has finally landed from OpenAI, and I'm going to look at what it is and how it works, where it pulls its data from, of course, how it decides to rank information. So let's start with the basics. So, yep, so search gpt. It's a new ai powered search tool developed by openai, who also created, obviously, chat gpt and it's they've designed it to help users retrieve information and answer questions in a sort of nuanced and conversational way. So, unlike traditional search engines like google, which display a list of links obviously, google have now got their ai overviews at the top um, but still you know very much traditional. Here's your list of links um. Search gpt generates a sort of a direct on-the-fly response to questions based on what it understands is the most relevant information that it can find that's available. So how does it do this? Well, it's powered by the same kind of underlying functionality as ChatGPT. You very much have that ChatGPT window if you've used ChatGPT, but they've sort of fine-tuned it to function more like a search engine. So they say it enables them to prioritize accuracy, timeless timeliness and relevancy in its responses. So it's kind of like they're trying to pitch it as a hybrid between a search engine and like what way they will call a knowledgeable assistant. So when you come to it, essentially you will if you've not tried it yet um, go to your chat gpt. You have to be, I think, a plus or a teams user, although they are going to bring it into sort of the lower and free tiers soon. Um, yeah, you basically get the standard um message chat gpt little box at the bottom, but instead of um, there's a, there's like a new little thing you can click and it's like the world symbol and a search. So once you click that, you're basically telling google this is a search, not google, but chat gpt. This is now a search session and you want it to search, search the web as well as use its own knowledge. And then it becomes like yeah, this, this new, this new session. It only works with um chat gpt 440. So it doesn't work with the new 01 and 01 preview and 01 mini. It only works with the 40.

Speaker 2: 3:45

Now let's dig into how it works. So, at its core, search gpt uses a mix of open eyes sort of the proprietary models that we used to when we work with chat gpt, and then integrations with other sources to pull data in. So this means it actually has access to update information through live browsing capabilities. So it means, unlike the previous models, sort of traditional chat GPT where you're limited by a cutoff in their training, data search GPT can actually actively go and pull data from the web in real time and feed that into the answers it gives you. So it's a bit like they say. It's a bit like having a hired skilled researcher. So it looks for sources across the internet and it quickly evaluates that content based on reliability and relevance, and synthesizes this information to develop what they say is a coherent answer. And this capability. This sets apart from the traditional ai which we've seen today, which is, yeah, which has that limit of only answering questions based on static information. It's learned during training.

Speaker 2: 4:48

This is the kind of thing I've talked about when I've said don't use hrgpt for keyword research and things like that, because it doesn't have access to the right data sets. Now, search gpt is limited in its data set, so don't get excited and think all right, ed said last week don't use chat gpt for keyword research, I wouldn't use search gpt for keyword research. Okay, it still hasn't got access to all those data sets that keyword research tools and fugal itself has access to, but it does now have much better access to more live information so you could talk to it about more recent events. You could talk to it about um, for example. I asked the question. You know what's the current bitcoin price and it could go and get that data and bring it straight back. What was the bitcoin price over the past five days, could get that back, that back, and show me the price fluctuations over the past five days. So it could do things like that now that it couldn't do before. So the key thing now is like where does SearchGPT get its data? We need to understand its data sources so we can understand sort of the capabilities and limitations. It's got general sources, which are like news sites, databases, forums and other public web pages, and they say they've partnered with various trusted data providers and sources to make sure they pull a wide array of reliable content.

Speaker 2: 6:15

Now, one of the key things here is the big one in terms of SEO is Bing. It uses the Bing's live index, so a lot of the deciding where to go and search where to go and find sources will come from Bing. So for many years Bing has clung around as the small market share and has generally been ignored by many SEOs, especially if you're in a country where Bing doesn't get used much. But Bing stayed there and it's actually not bad. Its index is comparable, probably, to Google's in size. Is it as sophisticated as Google as a traditional search engine? Maybe not, but it's not bad and I think in many ways it was kept by microsoft for many, many years simply because I don't think microsoft wanted to completely abandon search just to um google. Now obviously bing, and so microsoft and open ai have got um a lot of common ownership. You know microsoft are a big investor into um, open and ChatGPT and all that runs on a lot of Microsoft hardware, microsoft-owned hardware. So they've got that relationship and they're obviously utilizing that to bring the power of Bing's index with the capabilities of OpenAI to synthesize the two together along with the other sources which they are working with to be able to create what isn't we now see as search gpt.

Speaker 2: 7:41

Important thing to note with these results and the data, it doesn't just pull its answers verbatim. Instead it uses its language model to to understand the information that it's pulled and then to summarize that information, which means you're getting a unique, synthesized response based on multiple sources, rather than a direct pull from just one page, from one site. So their aim they're saying this is to reduce the risk of a single source being biased and influencing the answer. It also means that there's not going to be necessarily a number one result. They do give their sources down the side. You can get a sidebar with the sources and obviously they are listed in a particular order, but they're never going to rely on just one data source. So in a way it's a bit like if you were to go to Google and do a search take the top five, six, seven pages, look at them all individually and summarize in your head what they're trying to say between them. This is essentially doing it all that work for you in a matter of seconds and, using that powerful llm ability to summarize information, um present those results to you. So it actually works quite well, I've got to say. Um know it summarizes it in a way that you're still going to want to visit the pages most of the time that are being referenced, those source pages that are there. Clicked they are. You know you are, they are given credit and there's a link to them and you can go to them, so you can go and see where the source of this data is from after seeing the summary of all the data together. So it's quite interesting in that respect and obviously a key one as SEOs for anyone interested in SEO.

Speaker 2: 9:20

We're going to be thinking how does search GPT decide on rankings? And you know relevancy. So this part's quite interesting because, just like a traditional search engine, search GPT has to decide what sources to prioritize, especially when there's conflicted information or a large amount of data on a topic. But rather than relying on keywords, backlinks and other SEO signals, searchgpt says it evaluates data based on factors like credibility, freshness and contextual relevancy. So it's trying to work out, to process, how trustworthy a source is and how current the information is. So, for example, a recent scientific study on a topic would generally, they say, be prioritized over a blog post from a few years ago. So freshness appears to be key to them. And they say they're also looking for consensus across sources. So if multiple credible sources say the same thing, that information is more likely to be included in the response.

Speaker 2: 10:21

Now, obviously, going back to the fact that Bing does factor into this a lot and they're using Bing as a source. There is going to be a situation where, actually, keywords, backlinks and those other traditional SEO signals, they're influencing Bing. They are going to be influencing these results as well. So there's still that traditional side of seo going to be in there on influencing bing. But obviously I think the search gpt is then with that. When it's taking all the data and summarizing the data it's found, it's going to start adding on these things, like you know, like the relevancy, um, like the uh recency, like the relevancy, like the recency, like the credibility, just like you would yourself if you were looking at a set of results. You're going to probably prioritize newer pages over older pages You're going to also look at, you know, you're going to give credibility to sites that you recognize. A particular website, say, it's a well-trusted news source. You're going to trust that over, like swan's homepage blog, you know. So what does this mean for the world of sgo? Okay, what are the implications? What's the impact of this? Going to be.

Speaker 2: 11:26

Well, this is, in my opinion, the first time I've seen and a new approach to search. This actually looks like it could work. That is different from google. It's not. You know, I think the reason google became so prevalent is because it it was a the pinnacle of that type of search engine. Uh, the traditional search engine as was. It was the pinnacle and no one could better it because they've kind of mastered it. This is doing something different. It's not just being a search engine, it's adding in this this whole summarization ability, and I know google are trying with their um, their ai overviews. But this you can have the conversation with search gpt. You can start the initial research, read its initial summarization and you can ask further questions and I'll go and search for more sources that you can delve deeper into that which obviously you can't with traditional Google.

Speaker 2: 12:20

So, thinking about how it affects and will affect SEO in the future, is we now, instead of being, let's say, to rank in a particular position for a particular keyword, what we need to be really is among what SearchGPT considers these credible sources when it's pulling an AI, when the AI is pulling an answer, you want to be one of those sources that get chosen and I think, in the first instance, clearly doing one in Bing is going to work. I've got plenty of sites that do one in Bing. I've gone across and tested it across some of these topics and my sites are being included and some of these sites do better in Bing than they do in Google. So I've got some which are doing well in this, better than they do in Google, purely because Bing likes them that to me at the moment. Okay, so just correlation. I'm not doing a massive, great study, but this appears to be a thing for me that people are going to start. I think paying more attention to being freshness is obviously going to be a key thing. The look, you know, the mention of freshness that they're looking for, although it depends again on your topic. Some places fresh, you know things change very fast, change all the time, so the freshness is going to be dependent upon the topic you're in.

Speaker 2: 13:33

How is this going to affect the consumer search space as a whole? Well, like I said in the podcast just a couple of weeks ago about the actual size of ChatGPT as compared to Google, you know ChatGPT has got such its user base is like an order of magnitude lower than Google's. Most people are still going to be using Google. So this SearchGPT power users, people like ourselves who are interested very into online, are probably going to start using this more. That's without a doubt, but will the rest of the populace, will your average Joe in the street be into using SearchGPT? Well, at the moment it's a paid product. You have to pay for a you know, paid chat GPT plan and you know there are millions of people 10 million people, I think, was the last count, somewhere like that number are paying users of ChatGPT, so are eligible to be able to use SearchGPT. But that's a drop in the issue compared to the you know 5, 6 billion users online worldwide. So you know this is still niche. This isn't going to suddenly destroy your seo efforts if you're focused on google. You know, since this has been launched again, I've seen no impact on any of my sites. I don't think anyone really will have seen an impact because of this, because it is just still so niche. So this may be how things evolve over time and it's still other days. It's still a raw product.

Speaker 2: 15:04

Okay, there's many things it's not great. It's not great at. It's not great at local, for example. So this is things they're going to work on. So you know it's not great for directions. It hasn't got Google, it hasn't got the Maps function. It hasn't got all those functionality pieces that you're used to in Google and lots of other people used to in Google and at the end of creating these ai results compared to the cost of google search results is, you know, it's much more expensive computationally and therefore cost-wise for open ai to do this kind of search gpt, whereas traditional google you know google dialed it down. They've got that system operating so cheap um per query that's why they can offer it free. Obviously the subsides it with ads again, so cheap t.

Speaker 2: 15:52

No ads. It's not monetized beyond you having a subscription. So it's going to be interesting to see how they do come up with a free version. Do they have to put an ad system in all these things? If they have to do that, it's going to take time. So, in the least, in the short to medium term, I don't think this is going to be the google killer for the mass market. It might be the google killer for the sort of specialist market, but google can cope with that. You know it makes its money out of advertising to the masses on a free service. So at this point I wouldn't I wouldn't be thinking this is game over tomorrow.

Speaker 2: 16:27

But obviously we're going to have to keep an eye on this. We're going to want to appear in this as well as we want to appear in Google. But at the focus for you, a few websites should still clearly be Google. But keep an eye on this and use this. You know it's great. I would recommend people to use it. Be aware of its limitations, okay, so I hope you found that useful. I aware of its limitations, okay, so I hope you found that useful. I hope. If you've got any questions, any thoughts, do get in touch, love to hear from people and until next time I'll see you later.

Speaker 1: 16:55

Before I go, I just wanted to let you know that if you'd like a personal demo of our tools that 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.

Speaker 1: 17:37

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 channel5 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.