Marcus Tandler says “So, my number one SEO tip for 2023 is to regularly check your most important SERPs snippets. I know that sounds super basic, but never underestimate the basics when it comes to SEO. We all know Google is rewriting page titles more and more. A recent study showed Google rewriting page titles more than 60% of the time, so it's becoming more important to really be on top of what your SERPs snippets look like. So, what I like to do is to take my most important keywords, which account for, let's say, more than 70% of all my clicks, and then manually check all of those to see if Google is displaying the correct title, and description. And even more, what are the snippet of my competitors? And does my snippet really stand out? To do this at scale, there are multiple tools you can use. For example, I really like this handy little Chrome plugin from a German Stephan Czysch. It's called SERP Snippet Scraper. So, you just go to Google and search for one of your keywords, then just activate the plugin, which will copy the snippet to the clipboard, and then open up Google Sheets or Excel and paste it into the spreadsheet, you now have all the snippets of the entire top 10. You can now not only see if Google is using a title and description, but also the whole top 10. So you can maybe find a more suitable snippet that really stands out from the competition. I also like to save these reports. So, I can check back at a later time to see how my competition has changed their snippet. Or if a search intent shift has occurred, if I see the new snippets targeting a different intent all of a sudden, and checking your most important keywords manually, although it might take quite a bit of time really has lots of benefits. You'll additionally see stuff like what kind of SERP features are being shown. If you for example, see map results, aim for getting into that box as well. Is Google showing Rich Snippet, which of my competitors is using star rating rich snippets or FAQ page schema for extended search real estate? There are multiple ways to create a much better snippet with the help of structured data. What's the main intent of the query? Or is there maybe a mixed intent when people search for this query? Has there been a search intent shift recently. I really liked the concept of happiness in three Google is using. So, by looking at the first three results, and or universal search features, the first three results Google is showing you can easily derive the main intent and or if there are different intents.”

So going back to what you originally said, you originally said that Google was changing about 60% of page titles. Is it possible to force Google to make it more likely to use your page title?

“In a way, yes. I mean, the most titles that Google is changing, and actually too short or way too long titles, where you have maybe your brand in there twice. I mean, like these are like the basics, obviously, that you always should avoid. But sometimes it's always it's other than the nature of the query, which basically dictates how the SERPs snippet it's going to look like. So if I'm searching for a specific keyword combination and this is not included in my meta description, because it may be targeting a different keyword, the main keyword, so to speak and this is like a secondary keyword, where the page is also ranking for but it's not mentioned in the meta description. Google will not use the meta description, but rather use something out of the content, like a text fragment from the content to display as the description. But here, you can also do something because you could go back change the meta description to also include this keyword combination, but this might not be such a good idea because you're targeting a different keyword. And if you're also ranking really well there, you don't want to jeopardize this good ranking by changing your meta description for this secondary keyword. But you can actually look at what text fragment Google is using for the description, and just go into your content that actually changed that text fragment, like gradually, I mean, like, you don't want to change the whole thing, but gradually change it, and Google will pick up on these changes. And so, you can also change your snippet by changing the text fragment on the page.”

So how often should you actually be measuring what's appearing for your snippets?

“Oh, as I'm doing this, like every month and I would do it even more often, it wouldn't be such a lot of work. But I think a month is a very good timeframe here to actually do this and also to ensure yourself everything is going correctly, but I really want to have a good feeling for my most important SERPs. I just want to see what changes. Are there new universal search integrations all of the sudden, like for example, video results, all of the sudden video results are also shown there. And I'm like, okay, that's another opportunity for me, I also need a video result that I can actually put in there, right? So it's no longer for SEO getting to number one. But if I can have 10 benefit, 10 links to my property in the top 10, due to the various Universal Search integrations, that's, that's awesome. So, I would always aim to get as many of these universal search integrations that are shown as well, to just gain more sub real estate.”

So should you be keeping historical records of your snippets, and then comparing that data with your Search Console data? And seeing what has the best impacts in terms of things like click through rates?

“Absolutely, definitely. And even in this case, I would actually do a more sophisticated approach, actually, I do this with Ryte, we do A/B testing. And you can also do this in the past, right? So I can also go in the past and actually look at when I actually changed something, and compare these time periods, and actually can see if there's been an impact on the CTR, due to the sub snippet change. So, I think it's vital to save these results to really see where do I make an impact, is there, like a change that I did was so good that I can also use this specific pattern on other URLs. But this is something I would test with our A/B testing tool to really see is this new pattern in the title suitable for more pages. And if I see this having a positive effect, I'm just going to roll up the change. And then respectively, I'm actually looking at it again, after months to see how it has developed, if it has really helped my hypothesis, or it changed back again, and I have to have to do it over again. But I think this is what SEO is all about, right? It's not about ranking on number one, it's about people clicking through. And even if people are clicking through on my result 25% of the time, I want to get to 26 I want to go to 28, whatever, because this is people going to my site. This is what I want. And so, I think it's always pays off to diligently test your snippets and try new stuff. And you don't even have to do it if you're, like really worried about “oh, my God, I don’t want to change it because I could lose the position”. Try it out with PPC. So just try different variations. And Google will actually automatically find the best version that yields the best CTR, and that you can also use in your organic snippets. And if you say “oh, but that's too expensive, because Google, everybody's bidding on these terms”, then do it with Bing. So, there's, whatever you're saying, there is a way to actually test this and get more out of your clicks. And this is what I really like, because a lot of SEOs advice is always about low hanging fruits. And this is not working, try to make it work. But I also like the approach of I look at something that works really well and make it even better, right? And so, I think this is always a great approach. If you have a good SEO site to get more traffic, more qualified traffic, more good traffic to your site.”

Are there any magical phrases that you can recommend that are quite likely to result in a higher click through rate?

“Even if I could, it would probably be stupid, because it's always dependent on the target audience. I mean, obviously the topic itself, but also, what target audience am I marketing towards? Am I marketing towards b2b people? Or b2c crowd? It's going to be completely different. And they're going to react to totally different stuff. I mean, for an ecommerce shop having something like a big discount. This is not really something where in software as a service, b2b, it's like, what is really going to make an impact, because I'm leading directly with a discount, right? I’ve got to lead with the value, that software as a service actually brings us something. So, I think there isn’t this one phrase if you include this it is just going to work. There's no magic bullet here. And also, it dependent on the geolocation you're actually targeting, for example, in Germany, what was really well for snippet optimization is ASCII characters. So you have these hearts there, and the telephone and whatever kind of things. And what I actually learned from my colleague from the UK was that stuff like this, like with ASCII codes, in this snippet, it's actually holding off people from clicking in the UK because they find it spammy, right? So, in Germany, people love it. You're now seeing these ASCII codes and it really entices them to click more, but in the UK, it looks like a spammer. So, it's always hard to give this bulletproof recommendation. But that's it. It's not about me or other SEO expert, you know, “I've got so much experience, this is what you should do” no, just test it, just test it out, do A/B testing and really see what works best for your target audience in your geolocation. And, yeah, just let the data help because data doesn't lie, I might.”

And obviously, from what you're saying as well, don't assume that what works well in one country is going to work well in another country as well. So completely test differently for each countries as well. And also in relation to obtaining traffic is to get a high click through rate, you want to get the right traffic as well. So do you also try to measure things like maybe dwell time on a page or actually likelihood of making a purchase or actually purchases and tie that back to what's included in the snippet?

“Yeah, that's actually a great way. But obviously, this goes way beyond the snippet. I mean, for all my informational content I will always determine what is like if I write like a 10,000 character informational article, right? So, I would assume it takes about four and a half minutes for somebody to read that content. And so now, I have this value, and now we'll go back to Google Analytics and actually look at all the content, where there's a big disparity between this is how long you should spend on the page, and the actual time on the page. And if I see something like this page that should have four and a half minutes has an average dwell time of 20 seconds, you know, and then I go back to Google Search Console, and actually see people are just looking for one excerpt for one answer, which is in there, but I go way beyond. And this is like, “Oh, my God, I'm going way overboard with the thing”. This is just too much content for this specific case, right? I definitely got to do something about this. And so, I will always try to tie these together and really see why do people just not consume my content? And why is that? How can you make it more prominent, or split it apart into multiple parts, because it has different intent, and Google has really a hard time figuring out, I can't rank you for this intent, because it's more informational, but then it's like, on the commercial side it doesn't work. So, it doesn't work for both sides. I'm like, Okay, let's de-bundle this at the small informational site with the quick info you're getting, and then the transaction site where people actually stay longer. It is a bad example, I'm sorry, but I really love applying sort of the expectation I would have for people how to interact with the content, if there's a disparity, this is how I can optimize it.”

It is interesting to say that you would expect someone to stay about for four and a half minutes if you have a 10,000 word article on there. And obviously, people can really read that volume of article in that time, but maybe it's consistent with people maybe scanning a little bit more online, instead of actually reading a book and spending time reading every single word on the page. So, bearing that in mind, would you actually apply that stat to shorter posts? For instance, if you have a post of about 1000 words, would you expect people to spend about 30 seconds on your page?

“Oh, I'm sorry, for the example that has had four and a half minutes, 10,000 characters, 10,000 words I was just making up numbers, I struggled with the characters because I wanted to say words. It's a bad example. But, I mean, you could try it out yourself, right? Like, There's the article, and I'm going to read it myself. And this is really what I'm doing, I'm actually reading every article that goes live, I'm actually reading myself, still doing this after 10 years. And I am like how long does it take it for me? And I might also be scanning something, but it should at least be one and a half minutes. And this person has interacted with the article, and if it's like 20 seconds, I know I lost him somewhere. He wasn't even enticed to scan further. He was stuck in the beginning. Oh, look at this, there's a big image slider up there. And he was like, immediately upheld by the slider and never really scrolled down. Maybe he didn't even see that you could scroll down, that kind of stuff. So, I think this holds a lot of value to really try to match your own expectations on something with data and really see where this expectation wasn't met.”

You certainly shared a lot about what SEOs should be doing now in 2023. But now let's talk about what SEOs shouldn't be doing. So, what's something that's seductive in terms of time, but ultimately counterproductive? What's something that SEOs shouldn't be doing in 2023?

“Yeah, ditching your content team and solely relying on ChatGPT to write your content. I just had to mention this with the ChatGPT craze going on at the moment.”

Is there a place for GPT content?

“Oh, obviously, I mean, you see what bank writers are doing. And I think they have a great best practice. Like “this content is generated by AI”. Super transparent, and “this is the guy that fact checked, this is the editor”, Right? So, they have multiple people, and they show very transparently what's going on there and even Google can’t say anything against it, right? Because this is how it should be done. And what's not to like? You have AI, basically creating, like a draft with all the stuff that should be included, which is interesting, and then you have somebody go over this, and the typical EEAT stuff, your own expertise, the experience with a certain thing added to this article. Then have another editor review it if it is it really good for the user? What's not to like? Google can't say anything against it. So, I think it's a great best practice. But there's also a lot of people who don't see this, because this still creates a lot of work, and people want to, you know, ditch work. So they're like, “No, I'm just gonna have it write it and just put it on there, oh, look, it works”. And this is great. You know, in Germany, obviously, it works a little bit better, and probably a little bit longer than English content. But still, you know, this is not a sustainable strategy. So I think it's a great way to be more productive, but you have to tie it in into your content process and not just let it run by itself.”

Marcus Tandler is a Co-Founder and Chief Evangelist at Ryte, you can find them over at ryte.com.