How closely do you keep an eye on how the SERP is changing? And how is the changing SERP changing the way that SEO should be done?
Those are just two of the topics that we're going to be covering today with a man whose favorite cuisines include Indian, Indonesian, and Vietnamese. He’s presented at Brighton SEO three times and he works with various stakeholders to create unique SEO functionality over at Similarweb. A warm welcome to the In Search SEO podcast, Gerald Murphy.
The nine ways are:
Complex SEO behaviors
SERP feature evolution
Quality importance for ranking
Integration of different engines on each engine
Nine Google SERP Changes
Gerald: Thank you very much for having me today, David.
D: Thanks so much for coming on. You can find Gerald over at similarweb.com. So Gerald, how closely would you say that SEOs actually need to keep an eye on the SERP?
G: Well, first of all, you definitely need to do it on a weekly basis, particularly for keywords that matter to your website. So if you're in eCommerce retail, they're going to matter to your most important lines of business where that website makes the most money. Essentially, when it comes to the SERP, it's also thinking about how people may discover your website. So looking at it as part of a weekly report is critical, because it’s a very volatile world out there. There are SERP features that are released literally this year in 2022 that just haven't even been released globally, that are being tested, for example, in the United States.
D: I remember when I started out with SEO, it was just ten lines and ten links and that was it. It was slightly different from what queries look like nowadays. So today you're sharing nine ways that the SERP is changing. Starting off with number one, SERP-less. So what does SERP-less mean?
G: It’s actually how search itself started out back in the day. Remarkably, search goes back to 1945 by a US scientist called Vannevar Bush. And he created information retrieval using electronic equipment. And back then, when you think about things like searching for information you were asking a neighbor for a suggestion on a babysitter. That was a SERP-less world before web search engines came about. What happened was when web search engines came originally, they had really basic SERPs, the 10 blue links, as you have mentioned. And over the years, they've started to add other features that in essence, help us find information as quickly as possible. They're the SERP features. And there's been a massive big change in the number of those features over the years across all markets. And as I mentioned before, in the United States, it typically is for the engines the main market where tests occur for those features. What we're now moving towards is things like voice-activated searches, for example, Alexa or Google Home, that's bringing us into a SERP-less world again. So we're moving away towards looking at a screen and just asking someone and they're talking back this time. But this time it’s electronic rather than the neighbor on the street.
D: Going back to my own experience, starting in about 1999, one of the common search engines in the UK was Ask Jeeves. We were always going back there because Ask Jeeves was the answer to this particular question. So number two is complex SEO behaviors.
2. Complex SEO Behaviours
G: Indeed. As I mentioned before, it's like asking a neighbor and it actually has a lot of complexity behind it, who you're going to ask, and have they got experience in this thing. In essence, in the world of SERP, it's making sure that the right things are ranking at the right moment, those little micro-moments. And that's what SERP features aim to do. But online, when you add our really complex psychologies and how we find information, people can ask for the same thing in very different ways. Essentially, what we've also developed online is really complex behaviors of finding information.
The best person to learn about this is a guy called Peter Morville. And what he did is broke down six core online behaviors when it comes to looking for information on a SERP, an incredibly complex piece of work. But he simplified it really beautifully. And he started to showcase some examples. So for example, there's one called Parallel Growing. And what that means is we maybe come across a topic whether it's through a billboard in the street, we do a search for that, but we're not actually experts in that particular topic. Therefore, we don't have the keywords to formulate. And what we're doing is using the SERP to better understand other related terms or little snippets of information, for example, using a few lines of the description on the SERP to better understand that keyword. And then what we would do is start to do a search for other keywords and we parallel grow out and it really helps us to develop our knowledge of that topic.
And of course, one of the most famous ones I believe in the industry is pogo-sticking. The idea here, when you are on the SERP, you go to your results, you go back to it, and you pogo stick to and from. That's what he describes as search at its best as a conversation. Because your pogo-sticking on other results, you're getting other ideas, you’re going back to and from, and fully utilizing the system. It's also a really good indication that those results are very high quality.
D: I'm sure a lot of SEOs will be looking into that further. What you were initially seeing there about websites trying to sometimes rank for keywords that weren't absolutely core to what they do is, learning from that that SEOs probably need to encourage content writers and themselves to stop ranking for any terms that aren't quite so core to the business and may result in poorer click through rates, and rather focus much more on terms that are more close to exactly what they do as a business.
G: Absolutely. And the other thing that SERP features do is maximize your digital assets. For example, there's video as a SERP feature and what it's doing within a company is helping a company to promote videos as a digital asset on a search engine like Google or Bing. Essentially, it's making sure that we're thinking about what our keyword strategy is, but more strategically thinking about those as digital assets. And you're absolutely bang on. Relevancy is the absolute key, not just for the SERP feature, but for the core keywords. So it's not just about having a scattergun approach where you're randomly doing various things. It's really thinking about all the keywords that I care about and adding value to my performance online, what's the data showing. What are the biggest potential SERP features that I could be optimizing for? And that's where you really start to open up that conversation.
D: And number three is SERP feature evolution.
3. SERP Feature Evolution
G: Absolutely. We actually did an analysis from January 2020 right up to March of this year. And what we noticed is that SERP features themselves are growing hugely, i.e. the number of keywords that actually have SERP features associated with them. And as I mentioned before, there are actually new SERPs that are constantly occurring that are right there. So the best guy I would recommend on this is Brody Clark. Brody Clark's constantly doing stuff on SERPs and he's doing things like Things to Know is a new SERP feature. For instance, if you're doing a search for coffee grinding, it's actually looking at things like tips for coffee grinding, and a bit like what you have with People Also Ask and Related Questions where you've got these little dropdowns of their features. And what it's doing there is feeding you with more information directly on the SERP. But crucially, what our data also showed, what we did when we analyzed it, is that SERP features are becoming a lot more real-time. On Google, what we saw when we overlaid SERP features for Google Maps, and we overlaid it with the three national lockdowns in the United Kingdom, we actually saw that it took Google about two months to reduce the number of Google Maps as an actual SERP feature because it realized in the real world, very few people go to physical locations. In the second and third lockdown, it took them only a few weeks. They actually started to feed the system and acknowledge what's going on in the real world in line with SERP features. So the future of SERP features isn't just about them getting bigger, it's that they’re becoming more clever, and using a lot of different data sources to better serve searchers.
D: Yeah, I love that the Also Asked section comes in just over 50% of queries. In episode 112 of the In Search SEO podcast I interviewed Mark Williams-Cook, he's the founder of alsoasked.com. It's a nice bit of software to zone in on those types of keyword phrases. Moving on to number four, quality importance for ranking.
4. Quality Importance for Ranking
G: So search engines are now using multiple different data sets. And going back to that example I mentioned earlier about asking the neighbor and finding information literally through conversations, search engines equally take that massive big broad approach to this. This whole concept is information retrieval, informatics, and those types of disciplines in the academia world. But essentially, what it's also doing is taking that really broad approach. Right now, we're both in a room, but there's a lot of information in this room. It's the speed, the temperature, the direction of the very air that's in it. And what search engines are also doing is taking that massive broad approach in terms of qualifying if this is the best page for this particular keyword. And they're constantly reviewing that information. Not that alone, leaving aside things like updates that they do. Essentially, quality importance is all around E-A-T, particularly for the Core Web search engines of Google, Yahoo, and Bing. And it's not just about having this title with this keyword and this paragraph of text is what’s going to make me rank. It's about content strategy. It's about screaming that I know about this keyword and here's how I know it. Because I have, for example, an author that actually specializes in this topic. I actually have other areas of expertise that can showcase this particular topic.
D: I love that. And I think that's certainly a wonderful opportunity to identify keyword phrases that are more optimum opportunities to try to rank for. If you see that the SERP for a keyword phrase that you want to target contains an answer that isn't ideal, that isn't of high quality, and you could do a better job than what's currently there, that's probably a better opportunity than a keyword phrase with more search volume is. Is that something you would agree with?
G: Absolutely. So search volume is a good indication to look in terms of relevancy, but also to think about it in terms of the buyer or consumer journey. And what keywords when they're grouped. What ones are really effective, particularly for SERP features is when you split them up into your awareness, consideration, decision, and intent. Let's say, for example, we work for a stationery company. Paper would be an awareness term. It's too broad. Do they mean wallpaper? Do they mean a newspaper? Whereas when someone does a search, for example, for A4 printer paper that would maybe be more of a consideration keyword so the second phase. And typically that will have a lot fewer searches than the word paper. Equally so it may be that someone has actually searched for things like A4 printer paper. That's a decision example of a keyword where the searcher knows exactly what they want. And equally, on the flip side, you've got inspiration, which is, for example, making a plane out of A4 printer paper where you can fly it in the air. And in essence, when you've got that setup, particularly at least for your most core business or topic area of your website, you can then start to see one of the most prominent SERP features for each of those categories. And that can be incredibly insightful.
D: And number five is integration of different engines on each engine.
5. Integration of Different Engines on Each Engine
G: Indeed, it sounds a bit weird at the start. But essentially, what it means is that different search engines can rank on their results. For example, on Google, someone could do a search for Bing news and, of course, Bing ranks on Google. But crucially, when you start looking at other engines that aren't really web search engines as such, for instance, YouTube…When you start looking at YouTube, you'll actually see that there are millions of YouTube results in the United States, for instance, on Google. And what actually has happened there is yes, of course, they're part of the same company, Alphabet. But what it actually does indirectly is, potentially use that data to start building out your integrated search strategy. That's to say, if you've got a lot of keywords that are showing videos, you might want to start thinking about having a YouTube-specific strategy, and not necessarily ranking for videos on Google. So the trick here with SERP features is not to always get bogged down with one source. It's really to utilize that data and see what the trends are to maximize your website's performance.
D: Love it. So dig into and have a think about the likely style of content that the user is expecting and would like to see for a particular keyword phrase. That takes us up to number six, keyword/query understanding.
6. Keyword/Query Understanding
G: Yeah, and this is the future of search, what they're trying to do even with things like Google Discover is trying to feed us information. But when we're in that pool phase, when we're actually going to a search engine and asking for a particular query they're not just trying to understand what the searcher is searching for now, but what are they likely to search for next. That’s why Google has Bert, which is all-around natural language processing type stuff. What it's trying to do is to work out linguistics and conversations that may actually be searched for. For instance, for you and I, it's very logical to think, who is the prime minister? What age are they? We already know that's associated with an entity, in this case, a prime minister of a particular country, but actually in the world of search, they're trying to get into that mindset of trying to think about what would this searcher wants next can better serve the searcher with relevant results and suggestions of SERP features, like people also ask, autocomplete, auto suggestion, and all of those types of stuff.
D: Which takes us up to number seven, devices/mobile.
G: Indeed. Mobile is really important and search engines are now prioritizing it in terms of how they index websites and so on. But actually, when you look at SERP features, they're vastly different on both devices. Of course, it makes sense, as mobile has a much smaller screen. You can’t put certain features on because they would be the entire screen. So the SERP features themselves not only differ in terms of quantity but also in terms of the SERP features themselves. They change enormously, for instance, apps as a SERP feature are designed for mobile and not for desktop. So when we're looking at even basic things like tracking keywords and looking at SERP features, we have to be very cautious if a website is getting even 20 - 30% of the traffic from desktop, the SERP features are vastly different. So we need to be looking at both devices and comparing the SERP features too because it may be that a part of our content strategy is actually designed towards getting those desktop users into websites through really good digital assets via SERP features.
D: Obviously, that means that if you've got an app that's available on, for example, Google Play, then you can indicate to Google that that app exists and get them to potentially even feature that within the SERP. But I guess you need to be a little bit careful and clever about that. Because there are certain keyword phrases that don't want an app, they just want a very specific answer to a question. So should website managers get very specific about the keyword phrase that brings up the app?
G: Absolutely. And that's a really good point, actually. And this is why SERP features themselves exist because what they're trying to do is feed a medium that they feel is most applicable to a particular keyword. But again, search engines still have a long way to go. For instance, there was a professor in the United States who did a big analysis on image search, I think it was last year. And essentially what you find out is when you search for a professor in the United States, most of the image results, about 70% of them, were of white male professors. So when you looked at images alone, when the market has over 50% of professors that are female, and a substantial amount of them are people of color, SERP features themselves individually need to really grow and develop. There are actually a lot of biases. It's going back on your point of sometimes there are keywords that are very biased toward certain SERP features. And engines today are trying to work out what those little tweak things are so that they can tweak their algorithm.
D: Absolutely. And that relates to potential biases in AI, as they're only learning from things that have happened in the past, and previous biases from humans. And that takes us to number eight, which is zero clicks.
8. Zero Clicks
G: It's a metric we have on the platform where on the keyword level you can get the average zero click, soon to be trended over time. And essentially, it's allowing you to see things like when people do a search for a keyword, how many people click on nothing. For example, what age is Joe Biden? In the US, 95% of people who do that search, a substantial amount of them click on nothing. Zero clicks is a really powerful metric, not just in terms of picking the right keywords, but it showcases that SERP features are doing their job. Because in this case, when you're asking for questions explicit like that, the SERP feature answers it, and in that case, an instant answer is showcasing that. And also what it does, down below of that 5% of people who do click, they may click on other SERP features, like, for example, related people, and they're really understanding various entities to that query.
D: In relation to zero clicks, is it actually possible for an SEO to measure the positive impact of your brand being featured on a search result but not being clicked on?
G: Absolutely. You can definitely look at the keyword level or keyword list level and look at the zero clicks. That's actually one of the reasons why you want to trend the metric over time. But when you start overlaying that with a SERP feature, like for example, instant answers, for instance, it could be product listing ads or shopping feeds from a paid search perspective, and in essence that can really help you to better analyze and understand what is going on with your performance. Because unfortunately, when you're looking at some like console tools, like Bing Webmaster Tools, for instance, it doesn't actually give you that really deep layer of data. And that's actually why we do need to be looking at tracking keywords and so on.
D: And taking us up to number nine, personalization.
G: I've read an article and I think it was mostly a conversation on Twitter last year in which people were asking if my personal Gmail, for example, would be a SERP feature. Are we moving towards a much more personalized world? This is more to the future of SERP features. I don't believe so. I think the search engines are probably going to be a bit more cautious with things like GDPR here in Europe, and what sort of things we're doing with the information. But essentially, SERP features do have the potential and they are personalized to a topic, whereby they're saying things like there should be a lot of paid search type of features for retail websites which makes sense in many ways. But definitely, the future is to look around personalization, but also to think about what the next thing is, and that's why looking at test data is incredibly important today.
D: Yeah, it's challenging for everyone, because I think the general public isn't ready for a hyper level of personalized SERP. Because, for example, when I book a flight, and I get a receipt in my Gmail, and then maybe I search for the flight in Google, then I see the flight results. You're flying on this date to this location. And even for me, I think, how did you know that? But I think generally someone that doesn't have a background in SEO might be a little bit perturbed by the fact that there's so much information in the SERP.
G: Absolutely. And it's always about striking that balance just like SERP features, they don’t want to overload. And remarkably, I think this year, there's been tests where Brody Clark highlighted this on Twitter. And he mentioned that he found two People Also Ask boxes on the search results. So rather than just looking at one SERP and thinking, there's only one of that entity or space on the SERP, Google may potentially be trying to better understand and provide multiple SERP features that are the same to help us get to what we need.
The Pareto Pickle - Internal Linking
D: And let's finish off with the Pareto Pickle. Pareto says you can get 80% of your results from 20% of your efforts. What's one SEO activity that you would recommend that provides incredible results for modest levels of effort?
G: The simple answer is internal links. Really know what your authoritative pages are, know which pages are doing really well from a SERP feature/keyword rank perspective, and use all the other data we've discussed on today, for instance, zero clicks. But make sure that when they're relevant that you’re linking to each of those pages to flow that link juice around the website. And that really helps spread your authority across the domain, which has a nice little lift, especially when they're all connected to really relevant URLs that are really relevant to each topic.
D: Wonderful stuff. I've been your host, David Bain. Gerald, thanks so much for being on the In Search SEO podcast.