Carmen says: “Focus more on CTR and focus less on rankings. CTR has always been very important, but the reason why I think it's more important now than it has been in previous years is because the SERPs are getting more crowded.
The number one ranking doesn't necessarily mean having the highest CTR anymore. So for you to rank number one, number two, number three doesn't even mean that you're going to get a click. On the SERPs now you have all these snippets, you’ve got SGE soon, you have ads, and all of these are going to make the SERPs go further down.
If you always focus on a keyword ranking number one and you don't check the CTR, you won't be able to spot the problem. Being able to monitor that, at least for the keyword groupings should be something that you do normally from now on.”
You were talking about monitoring CTR by keyword grouping. So you monitor CTR, according to the keyword group, and I guess match that keyword group to a URL and then look at the CTR for that URL. Is that what you do?
“Exactly, I tend to group them normally for URLs but for clusters. For example, a SaaS client, let's say they do e-learning, I’ll get my clusters on onboarding topics and then I tag the type of content that it is, and then keywords they rank for.
Google is getting smarter and smarter because there’s also the topic of spam, but also because Google is starting to show things beyond the keyword that you type, and more around the intent and the history of your search. You can no longer assume that your one URL talking about onboarding is the one that is going to be popping up all the time. Depending on the intent and the search history, it might be your blog. But if you monitor the CTR of all this group of keywords that you have, and as well as the URL, you can actually check if the intent behind and how people search and you can also monitor how crowded the SERPs are.
In my opinion the SERPs won't be as crowded depending on where the user is coming from and depending again on the history of the user. I think there is no longer one SERP per keyword. There will be SERPs per user, and that’ll make keyword ranking way more difficult, and less usefull to track when you're trying to check if your strategy is actually working.”
I understand that you can get SERPs per user and different users will experience your pages in different contexts depending on what they're wanting to do which will obviously alter the CTR. So how do you know what's altering the CTR and how do you optimize for the intent that gives you the maximum CTR?
“It’s been a lot of having to go back to the basics of SEO to try to understand how to properly get to the user. I’ve been checking the SERPs constantly to see that my content is actually what the users are looking for and is giving something to the users that they cannot find anywhere else.
So optimizing for CTR, aside of obviously the titles, which is really difficult because Google changes the meta title so you can no longer optimize for meta titles, then you need to know your SERP by heart. Obviously, always try to optimize it anyway in case Google doesn't change it. But the most important bit is to check out your SERP. What are users checking in the SERP?
You need to constantly check it because it's changing so fast. It's no longer a new check once a year, you need to be constantly checking every month or every two months to see how the SERPS are. Where is it coming across? Is there a new thing that is video? Do you need to make sure that you have video content in there? Is your content good enough? Are you actually providing enough for the users? Are the users clicking on my site or my page? Can I get my users to link to the page from different platforms?
Getting the basics is the first thing that you need to do. Making sure that your content is good for the user, not for the SERP. Good for the user is technically sound, so does it have a schema? Does it have authors? Does it have everything that you need, and then from there, is it competing well enough with the rest of the content? I think we have to just get into rankings and do things in the background to make sure that we get ranking. We’ve forgotten the basics, which is convincing the user to read my content throughout and staining on my page and engaging with my content.
I know that I just said a lot of words, and I didn't give proper action to do. But I think the action is to understand your competitors very well in terms of what they're doing, and make sure that you convince your users to stay with you with proper convincing facts, with proper content and with a page that is really sound. This is what we should have always done as SEOs. We should have always focused on the user and not on Google, because at the end of the day, it’s clicks that are going to make us rank higher.”
You said Google changed 73% of titles. So if that's the case, then why not just use AI to automatically generate titles from your content, if Google is going to change it most of the time?
“Maybe if you do it in a way that AI is smart enough and the content is being produced in a way that is not dodgy. I'm not against using AI to actually produce your titles. If Google is going to be changing your titles, they normally change them using the h1 or h2, so if your h1 or your h2 are good enough, it doesn't matter if they change the titles.
If you create titles with AI, as long as it's done properly, then the importance in my opinion is that it doesn't look dodgy and doesn't look like it's not natural. So if AI is working for you to achieve the objective, then AI is good. The important bit in the end is that the user isn't enjoying it less and it doesn't look spammy.”
Do you try and split test different styles of titles to optimize CTR?
“It’s quite common. But again, they come with this problem that Google very often changes page titles. That's why I said that is a bit of a nightmare. I think it has been doing it less recently, probably because they are basically doing all the things, but I remember many A/B tests that we have done that didn't work out just because Google was always changing our titles.
The worst part is that you won't necessarily see it because if I'm checking myself, and I see my title, and I have a colleague check it and they see a different title, then it's really difficult to actually get proper results from the meta titles in particular. So if you use one page that actually gets a lot of clicks, in general, they will change it very often to see if they get more clicks.”
So if you're doing a split test on that, don't assume that Google is going to be showing the same page title all the time. It could actually be changing one version of the title more often which could have a tremendous influence on the result and skew everything that you're trying to test.
“Frankly, in many cases, the things that they have changed, if you have optimized your h1 and your title in a way that they are similar, then the title won't change that much. The problem is when the title and the h1 don't match, which then is a big issue because from the SEO perspective. Ideally they shouldn't be too much different, because your h1 is almost saying the same thing that the title. But it's very common for SEOs to change it to try to rank for more keywords. Sometimes though Google doesn't necessarily always get the h1, sometimes it also gets it from a paragraph, which doesn't help.
This thing of changing the titles, I haven't seen it as much this year as I saw it last year, for example. This year I’ve seen most of all the other weird stuff like getting in the UK, or in Spain, for example, sometimes I get translated articles from the BBC, even if I'm searching in Spanish in Spain, rather than getting published results. For international SEO this has a big impact because then it makes the SERPs even more crowded than what it used to be before.
These are more of the tests I'm seeing this year rather than changing all the titles. But obviously, it does have an impact on CTR, because if you are getting way more competition now that you were having before, there are more factors that you need to consider as well.”
Do you have an example to share?
“I’ve been doing two separate experiments, one over time, and another one only in the last year. The first one was using just one URL, checking out the rankings that have been constant and checking out the CTR to see how it had changed. So in the last three years, the CTR of this blog post has really changed and reduced constantly, year after year, update after update, even though ranking is always number one. This then made me check all the rankings for all the page ones of the client of mine, and that is when I realized that the correlation between ranking position one was very low versus getting all the clicks and all the traffic. In fact, I even got a graph where I could see that 80% of the clicks were coming from pages that were not even on page one.
In my opinion, this is crazy, because if you ask SEOs, we've always been told to check our rankings. But now rankings don't necessarily mean that you are getting the clicks. That's where I thought, we're getting everything wrong. In order to check that I was not freaking out, I even did the math behind it. So I got a financial guy who does all the data for our agency to look at the proper numbers mathematically to prove that the correlation didn't exist, and even with a formula, the correlation didn't exist, which made me a bit concerned.”
So is the primary reason that you're getting high rankings but no click through rate that you're hitting the wrong intent, so people aren't looking for that particular page at that point in their user journey?
“There are a few different reasons. So in the ones that we did get the intent of the page, the SERP was really crowded. So we had paid on the top, we had all the snippets, including we had snippets underneath. This means that the SERP was so confusing that people were jumping around the content to see whatever was more interesting. It didn’t matter if you were in position one.
Another thing is that you might be ranking number one, but if it's not attractive enough in the results, so people might not click on you. We know that in e-commerce in particular, you can get all your products shown in the SERPs, so does the user even need to click into your page to see what they're gonna find? Or if you get them to click, then you need to make sure that it looks perfectly with all the different snippets. I think the issue is there is not just one service.
Previously we have always monitored everything with ‘this keyword equals this SERP’, but there is no longer ‘this keyword that equals this SERP, because this keyword can be 10 Different SERPs, which is making it very difficult to actually track data and make decisions.
The second thing is that even if you have the intent, you can no longer guarantee that the user is going to click on you because they can find the answers to whatever they're looking for within the SERP without the need of clicking. In my opinion, that is going to get even worse when SGE is there and they can find all the information in front of them without the need of actually reading. That’s not to say SGE is going to break SEO, because the users still need to go to the website to buy and to engage with the content, but we cannot assume that ranking position ‘x’ is going to bring us ‘x’ amount of traffic.”
What's something that's seductive in terms of time, but ultimately counterproductive or something that SEOs shouldn't be doing in 2024?
“I think SEOs spend a lot of time trying to figure out what keywords to go after and which keywords are going to be the most productive. But actually, we should be focusing more on what are the topics that are going to bring our users to the site. We do use keywords as a way of understanding their usage, but that shouldn't be our main metric.
Instead, I think we always need to put the user above everything. Then from there, moving into topics, technical stuff, schema, internal linking, etc, but always keeping the user as the main objective and not the keywords.”
Carmen Dominguez is Head of Organic at Hallam, and you can find her over at HallamInternet.com.