In this Episode
- [03:04] – Marcus responds to Stephan’s question about needing to base content creation on data (rather than just writing whatever you think your audience may want to hear). In his response, Marcus gives some great insight into creating content, all while explaining how Searchmetrics works.
- [08:17] – Marcus talks about the differentiation between coming up with a topic and a list of keywords. He discusses the relevance of intentions by using the example of hairstyles.
- [11:49] – We hear a bit about how Google’s changes mean that marketers should now focus on having happy users instead of good rankings.
- [13:04] – A lot of people still view content the old way in terms of creating throwaway content based around a specific keyword.
- [15:48] – Where is the future in SEO if we’re moving away from checklists? How should we be preparing today?
- [21:13] – Marcus tells us the steps we need to keep in mind for mobile searches and gives the concept of “mobile-first design.”
- [23:33] – Marcus clarifies what he meant in his reference to “AI first.”
- [25:12] – After Stephan’s suggestion that interactive voice commands (rather than the keyboard) are the interface to the web of the future, Marcus agrees and elaborates.
- [28:53] – Marcus talks a bit about the new ranking factors. He clarifies by giving an example about the length of pages by word count. We also learn about the extensive research his company has done through ranking-factor analysis.
- [33:28] – Stephan asks Marcus which SEO rumors he can squash.
- [35:20] – One of the myths out there is that Google is using your own analytics data against you, Stephan says. Marcus gives his thoughts, which is that Google doesn’t need to do that because they have enough data from Chrome and search behavior.
- [37:07] – Marcus talks about dwell time versus bounce rate. He points out that users in some areas are more likely to look for more results and points out how this can make the bounce rate a poor indicator in these fields.
- [39:02] – After every major Google update, Marcus’s company makes a report on the biggest winners and losers. He talks more about how that’s calculated and how they’re able to see patterns across pages.
- [42:29] – Marcus explains his thoughts on RankBrain and explains how it works.
- [45:00] – What are the next steps for someone who wants to work with Marcus’s toolset? The answer: if you go to searchmetrics.com/content, you can apply for a demo.
- [47:22] – We learn why Searchmetrics is being released this way, rather than simply giving it a large release.
Hello and welcome to Marketing Speak. I’m your host, Stephan Spencer. Today we have Marcus Tober joining us. Marcus is the founder of Search Metrics, he is well known in SEO circles as a real innovator. Prior to Search Metrics, he founded Ecards and More which was one of the largest online greeting card companies in Germany and he sold that in 2008, was at 180,000 unique visitors per day, generating 30 million impressions. Sizeable property there, it’s now called BonnyPrints and his focus is Search Metrics. He started that company in 2005, changed the name in 2007 to Search Metrics and it’s been a real leader in the space of search analytics and enterprise software for SEO. He is a leading thinker there and focusing on driving the technology forward. He’s the CTO, Chief Technology Officer, at Search Metrics as well as founder and he was also named as the Search Personality of the Year at the European Search Awards this year in 2016, that’s pretty awesome. Marcus is a frequent commenter on trends in the industry, he’s a speaker at many of the conferences, he gets quoted in different journals, and magazines, and so forth. Welcome Marcus, it’s great to have you.
Thanks for having me on this podcast with you, Stephan. It’s really a pleasure.
We’ve been friends for a while now. I remember when we first met, I think it was an SMX Conference I believe.
Yes, and Rand Fishkin introduced the two of us. That was a good day. We’ve been friends ever since. I know that you guys have really made some big leaps forward in innovation, just this year with the topic explorer and the content experience suite which is mind blowingly amazing. I got to demonstrate that live at the PubCon Conference just a few weeks ago, I was speaking on the advanced keyword research session and everybody was ooh-ing and aah-ing in the audience in a packed room, it was pretty amazing. The idea of taking data on what people are searching for, what is popular, turning that into content instead of just sitting down in a blank screen and starting to type something just because you want to or you’re thinking that that might be appealing to the audience. Those days are gone, you really need to base your decisions and content creation on data. Let’s talk about that.
That’s a pretty big topic. Actually it’s not new but many people still use Microsoft Word as their foundation to create content. Many people nowadays, they collect and dig through a lot of data that they export maybe not just hundreds even thousands of keywords with search volume data from sources like The Keyword Planner. They analyze themselves and competitors. In the end, for many days they build pivot tables and spreadsheets and they try to find topics. The goal always is to find opportunities, to find good topics where you can improve. When you have content already, you shouldn’t just publish and forget the content, you should repurpose the content, you should change it, update it, you should look at what does the user actually want. When you write new content, it’s the same. The problem is when you waste many days of your time without knowing what is actually the outcome because you don’t know the use and tension when you dig through spreadsheets, so you try to build classes all with the gut feeling and we saw this happening since many years so we started over three years ago to work on the software that’s where the topic explorer that you just mentioned is part of it where we wanted to give people a visual way to dig through topics but not based on gut feeling, based on data and it’s pretty cool so we changed the visualization many times, we talked to many writers, we got their feedback. What we have now is a very easy and simple way to do the work where you normally need days, you can now do it in minutes. It’s hard to explain because when people hear that, “Oh wow, you needed days before now you need minutes? How’s that possible?” The reason is because the data, because when you build clusters based on existing historical data, when you analyze landing pages historically over time, you can analyze how these pages are changing and you can also correlate this to their search performance, for example. I know that there are many signals, there’s a lot of noise and you can’t always tie everything back to content, but if you have enough data, you see very quickly that content is a huge part of the success of our website. We took this historical data and we built a very large graph around search queries and cluster to search queries into multiple classes. One class is what we call use intention, so in the software it’s called semantic proximities, that’s what you’re using but actually we internally call it use intention because when you see a cluster we want to give the user everything that’s in relationship. For example, when you search for short hairstyles, when you do it because you have short hair. Maybe other people search for what’s the best hairstyle at the moment and maybe it’s the short hairstyle for men but it’s a different search query but actually they mean the same because the content can answer both questions, and this is what we did. We took hundreds of hundreds of millions of these search queries and we clustered it. When you now do a research in the topic explorer and real time, you get these clusters and that’s really cool because normally you try to mimic the same thing in Excel and it’s not that easy because you need to have gut feeling and good experience around SEO for it. The same we did for example in the topic explorer for other classifications, for example the buying intent, we built a buying funnel from awareness to purchase and then retention. We are now able to also classify all these different search queries according to the stage in the sales funnel. Especially when the companies want to create content that’s appealing for retention because they want to serve a good service to all the buyers they already have, they can focus on all the search queries that people do especially in the retention phase. An example would be how to charge a car battery, because when people search for how to charge a car battery they already have a cary battery, that’s the assumption. When they search for how much is a car battery or when they just search for car battery, they’re more in the awareness or consideration phase. This is how you can think about it, you have an enormous amount of these search queries and you can drill down and see in a very visual way how these cluster. It’s pretty cool.
The old way of doing this is to think about big piles of keywords and then group them together, give them keyword markets or add groups if you’re thinking AdWords terminology. Then you make decisions on what you’re going to write about based on those list of keywords but you’re not really getting a sense for the topic. Let’s quickly differentiate the power of thinking about a topic versus thinking about a list of keywords.
One of the reasons why we came up with this idea, if you have a conversation with very smart people, people who actually really know how to do this topic research, most of these people are always very proud of having, “I did this research and I came back with 10,000 keywords.” Wow. One of their valued propositions is that they can aggregate large amounts of keywords. We thought okay, why is this value propositions to have a lots of amounts of keywords because when you want to become relevant for certain topic, it’s not about that you write a piece of content for 10,000 keywords, right? You don’t do that. You want to become relevant for only stuff that matters in this kind of relationship. It means that the keyword research is based on finding as many keywords as possible. For example, keywords like a short hair keyword then you find all variations according to this short hair keyword, but there is no meaning behind the relationship of all long hair keywords to the short hair keywords so that’s the problem. When you have hairstyles, you have all combinations and you have long hairstyles, short hairstyles, curly hairstyles, brown hairstyles, blonde hairstyles. But the thing is you have a large list but when someone searches for blonde and for brown or brunette or long or short, they have different intentions. Our topic explorer would give you the borders between all of them and this is what you then call a topic. Topic is if you still have a large list of keywords, you can build clusters of these keyword that belong together and this then becomes a topic with a name. When you have for example short hairstyles, then maybe best short hairstyles in 2016, best men hairstyles, because most of men have short hair or maybe hairstyles that are popular in Berlin for example. We cluster them because actually these keywords are becoming one topic because the user actually is looking for the same even if the user is articulating a different search query within search. Because search is not specific, especially now when you have conversation search, you start asking a question. It’s even more complicated for a search engine now to understand what does the user really want. That’s why this topic research is so important because SEO in the very early days was write one page or one piece of content for each search query that you found and now it’s the opposite, now less is more, right? You write less content but you make it more realistic, you make it longer, you have a better structure and then you become more relevant. This is the new SEO.
Long form content, the kind of stuff that you see on medium.com is the future and not these short throw away articles, 500 words, how to boil an egg, take a pot of water, turn on the heat and then step three is put the egg in, that sort of stuff that got hammered by Panda because it was content farming, just creating low-quality, thin content. Those day are over, you got to really put the effort in and you got to write to a topic now instead of just an individual keyword.
Exactly. We became so clever that they are able to understand the structure and the meaning of the content. When content marketers do not adopt this new philosophy, they would never have a long term success and in the end they don’t do anything in favor for themselves because when they only try to have thin content because they have it, they don’t do a good job for the user in the end. That’s why marketer’s stage would change their goal in having happy users instead of having good rankings, the rankings come later anyway.
Right. Essentially if you’re writing for the users, then the different metrics that show that you’re on the right track like the bounce rate and the dwell time, and that sort of stuff, will improve but you’re not just writing content for particular metric. You’re not trying to write it to hit a certain keyword density or some nonsense like that, you’re just writing to be helpful, useful, informative, solve user problems and then the benefits of the rankings come after that.
Yes. You mentioned how many people are still like this because there’s a big delay between what we talk about, what SEOs and marketers talk about, and then what’s actually happening in the companies when you talk to content writers. Because it took us many, many years to tell content writers hey you should utilize this keyword more often. Okay. Now I get it. Now I’ll utilize the keyword. There are other SEO platforms in the market and when you look at a couple of them, you still find it off these checklist approaches, right? There’s some content and then you get a check mark when the keyword is found in the content and then you get another check when the keyword is found in the title. And then when you maybe use the keyword three times in the content then you get another point. And then at the end you get a lot of points and it’s interesting to see that we already, since years, we talk about creativity and content and all that stuff. But in the companies it takes a very, very long time to change. It’s really all about change management too because habits don’t change that quick. That’s the problem.
Basically there’s this tectonic shift in the SEO industry from the tactical checklist of make sure that the title includes the keyword, make sure it’s at the beginning of the title and not at the end, make sure that the keyword is also present in the H1 tag, make sure that it’s in the first paragraph, etc. That’s very tactical and it doesn’t really work anymore, not that it really was that effective but the way I like to think about it is strategy will always win over tactics. If you’re a tactician, you’re going to get just decimated. If you’re a strategist on the other hand, you can always hire a tactician to help you clean up the little messes, the details that you missed. We made a syntax error in the robots that takes tea, get that cleaned up, of course. But the strategy is always going to win, it’s like in The Art of War, it was said that tactics without strategy is the noise before defeat.
Yes. Where is the future in SEO if we’re moving from these tactical checklists and it’s very siloed and granular and short sided? What is SEO going to be like in just a few years time and what shall we be preparing for today?
That’s more like an epic question. The thing is it’s changing dramatically, I believe we are in a very big transition from what SEO was into what SEO is going to become. I will give you a couple of ideas, you gave me a couple of ideas already. For example, when you look at the diversity of devices we already have, it’s not just that we have mobile devices and we have laptops, now we also have tablets and convertibles and we will have VR glasses, we have AR glasses, we are talking to devices like the Amazon Eco or Google Home and the diversity of devices where you can have an interaction and then you have a result out of search is incredibly increasing. It means that the SEO people or the company need to be aware that whatever they use when it comes to their technology needs to be accessible and analyzable, that the content is understood and they can be considered when someone makes a search query that they are the most relevant source. Because when you do a search on your mobile phone, the screen equity is very small so you only see the first couple of results. If you do it on your desktop, you have more equity normally. If you have a conversation search, you normally only get one answer. Then there’s almost no equity anymore because only the best wins. It means that SEO is going to have a tremendous shift in knowing what you do. For example now it’s very popular to use a single page app framework like ReactJS or AngularJS but they also bring in a higher complexity that these single page apps are crawlable now, it’s possible so that’s not the point. If you use the out of the box solution, you have very many problems very quick or if you for example have an international site and you just make some small, little, minor mistakes in the HREF link tag, you can’t be found in many countries for example because it’s a small, little confirmation that means the burden that SEOs have on technical SEO is going to be increased which is good for us because it is very good for SEOs if they follow this kind of profile because it’s getting more complicated. I believe the other side is, and I also talked a little bit about this, how can you create the content that the content is appealing to the user and the most relevant one that the search engine chooses you over the others by not looking at just simple factors? If you don’t have the most backlinks, if you don’t have the best structure but you have the best answer for the user, very likely you are the chosen one, it means we have these two worlds, we have the content world where you need to understand what is your audience, what do they want, do you need to structure the data that the audience can consume the content quicker. Imagine you want to become relevant for hairstyles, what I mentioned a couple of times, the people don’t want to read content like text for hairstyles, they want to see a lot of images, maybe an images gallery. Or if you search for something like XBox Unboxing, people want to see video and no text at all. You as an SEO, you need to understand not just how to make topic research nowadays, you also need to understand what does the user actually want in terms of the content itself and this is what I believe is this big transition into a whole new SEO paradigm where for example in the last years, SEOs, they used to be generalists, they knew about the content and the topic research and technical SEO stuff. I believe this is going to be split because of the diversity of the devices. Content needs to be prepared differently, this is a pretty important and complicated job and on the technical side, it’s the same. This is what I believe is the future of SEO and it’s great, it’s a bright future because it’s becoming more complicated and people have fun also kind of think about business because SEO always needs to understand the business KPIs. What each kind of change pays onto the business, what each change onto the page means for crawlability or the potential to have a higher visibility and so on. It’s pretty exciting how SEO is going to be changing in the next years.
Yes. One of the things that’s going to term that’s bandied about is mobile first. For a marketer, what should they be doing differently because more searches happen on mobile devices than on desktop and it’s never going back. People are spending more time than ever on their mobile phones searching the internet, the majority of interactions with the web will be happening on mobile devices. What’s the practical implementation steps that we need to take as marketers for a mobile-first world?
I think one of the biggest differences when it comes to mobile first is when people do a mobile search, it’s all about their current situations. The first factor for example is their location. When I’m in Berlin and I do a query for hairstyles, very likely I mean something locally. When I do the same in New York, very likely I have something I’m looking for in New York. Even if I’m looking for pictures, but maybe it’s more like New York or American style or German style. This is important when it comes to mobile first as the first factor. The next one is that if you do a search on the mobile phone, you have more information about the person. When you use the phone, the phone gives a search engine much more data so they know you, they know your history, I don’t mean the search history, I mean the history also of where you’ve been for example when you traveled, that’s what I mean. They know your habits, for example Google automatically understands where you live and where you work and all this has a big impact on how the content or how the service, the products need to be presented that it reflects the user when he makes a search. Mobile first, from my perspective, will play a big role when Google’s going to roll out a separate mobile index next year, what they announced just a couple of weeks ago. Because then it’s really all about understanding that if your business has a higher spread and searches on mobile phones, you need to think about having different content that you present. It’s true because at the moment when you look at stuff like our topic explorer for example, it’s mainly one topic and there’s no spread between mobile and desktop but I believe it needs to change because when Google proceeds this very quick in the good of the mobile user obviously, then we will have different variations also on the content and on the mobile site. That’s very exciting, and there’s another paradigm that Google for example, a picture you said that’s not only mobile first anymore, it’s now AI-first, which is really interesting.One of the biggest differences when it comes to mobile first is when people do a mobile search, it’s all about their current situations. Click To Tweet
What does that mean? AI-first?
It’s more like an internal thing for the Google people, AI-first, but if you look at the devices they just brought out with Google Home or their daydream devices for Virtual Reality or Google’s new products like Allo, their messenger where AI is built in where you can just ask randomly in the messenger book me U2 tickets for tonight. Because you use a mobile phone, the product, the messenger knows where you are, the messenger can also access your calendar, the messenger will know where you will be, the messenger can create an entry in your calendar and then book the tickets if you have the credit card details saved in it. That’s what I believe what he meant with AI-first, having a lot of convenience for the end user especially mobile devices is going to be significantly important for a company like Google.
We’ll be talking to our personal agent, our smart agent which will be our app or maybe it’s an Amazon Eco and we’re talking to, I’m not going to say the word, the wake up word because I have an echo right here that’s going to start talking and it’s going to break into our interview, that happened before, I learned my lessons. Begins with an ‘a’ and ends with an ‘a’ and that’s the wake up word.
I know the word.
It’s always listening and listening for that wake up word and who knows what other things it’s listening for. But that smart agent is our interface to the web of the future, it’s not the keyboard and typing on that, I think.
Exactly, that’s a dramatic shift. Users now have diversity, they like to choose, but you as a user of this device, you experience already the convenience of getting an answer very quick. I believe that you don’t miss the diversity, maybe it happens for a couple of things that you miss the diversity. For example when you want to compare prices because when you’re price conscious then of course you’re looking for price and quality and that’s something maybe these devices can’t do at the moment but normally the convenience factor is so high that when you just make a search like what is this and this and that, then you get a response and that’s it. This is now something the SEOs need to understand because this is something where we can become relevant. Not because these devices pick us as the source of trust and then answers to the user because this is also not how we make money, right? When they just use our content to give people the answer, but when we are wise and we hook into the systems and we use the structure data to also make it happen that you can book on our pages that you can have for example chain searches, like when you do a search like maybe what’s the distance from New York to Paris and then how’s the weather there and maybe book a nice restaurant and we are the source of this, then we can get the conversion in the end. This is a dramatic shift. We will see how fast it will become our habit that we use these devices, you know how it is, every trend always starts with a big hype like VR two years ago and now we have Oculus Rift and all these devices and nobody’s using them, me too. I got sick when I used my Oculus VR and since then I’m not using it anymore but we will see how quick it will influence our daily life. Then, marketers, they need to change their habits very quick because when they are there, and then they start and it’s too late.
Yeah. Part of it I think AI-first is to learn to optimize websites for the AI and not just for a checklist or a bunch of “200 signals” or factors that used to be the factors that were considered. Now it’s a much larger universe of potential things that the Google ranking algorithm is going to look at since machine learning is going to come up with things that just seem to work better, right?
Who knows what sort of signals we’re going to have to optimize for. It’s not going to be just the typical ones that we’ve gotten used to and even Google engineers won’t know because they won’t be able to understand what the AI is up to, it’s kind of scary actually.
No. Absolutely, that’s the point, no machine is perfect so it also means that we can reverse engineer what kind of signals the search engines are using or the AI is using that we become the most relevant source.
Great. That brings up a good point that you have this regular thing that you guys at Search Metrics do where you analyze the ranking factors that have positive correlation and seem to actually be moving the needle for different websites. Let’s talk a bit about that. How do you base that research, that’s a huge data set that you’re basing that on, tell us a bit more about it.
That’s a good point. I just read today the last draft of the new ranking factors that are going to come out in a couple of weeks. This year, we made it different so when we started beginning of this year to work on the new ranking factors, we thought okay, it’s now the fifth year of the ranking factors, it’s again very similar than the years before, we will make comparisons, we will see the change bla bla bla. But because we worked on this content product already since years, we thought okay we can’t just follow our own old habits and make the same, we decided to build a prototype where we compared the ranking factors for different industries. We took health, finance and e-commerce because we saw that there’s almost no overlap between all of them. Finance is transactional driven and also a very important topic for Google because they put the finance stuff specific in their quality writer guidelines. The health, that’s also something people care about a lot, I mean I do, you do as well Stephan, at least two. E-commerce because it’s heavily transactional driven, when we build this prototype, we found out very quickly that the ranking factors for these different industries, there’s a big difference between all of them. When you just look at the correlations of general ranking factors, then the assumptions for your industry, they could be wrong. Just give me one example, okay? When we look for example at the word length, the word count on pages, there was always this assumption that if you have long form content, the chances that you have better rankings are going to be higher. Very easy assumption, everybody would sign this thing. When we made this analysis, the difference between for example, finance and e-commerce and health have been so large, it’s amazing. For example in health we found a very large positive correlation that longer texts is going to have better rankings. In finance it was the opposite, the better ranking pages in finance, they had less text than the pages that ranked from position 6th then to 20. When we checked the pages, get some more insights into it, we found that it made sense because really good pages in health like WebMD and Mayo Clinic they have really, really long text, very good structures, they go into the pros and cons and they really describe things in detail. That’s why it make sense that longer form content performs better because people do care about their health. And then finance, we found more stuff like forms and conversion elements on better ranking pages because they really focus on what method is the fastest, not like long form text. That’s just one example. When you look at hundreds of ranking factors that Google uses to a computer ranking in real time, you can imagine that actually for each topic, Google is using different ranking factors. It happened that we now took five industries plus the general ranking factors plus the devices of mobile and desktop. We’re going to come out with at least 12 ranking factors in the next couple of weeks. That’s amazing because now you can just pick your industry and then see how you defer to the general ranking factors or how you defer another different industry and it’s pretty cool. I can promise that’s the best study we’ve brought out and that’s also the most expensive one because we started working on it in end of February, so beginning of March, and we just finished it. It’s a pretty long study and a couple of people worked on it.
Pretty cool, I can’t wait to see it. I imagine that in that research you probably can debunk quite a number of SEO myths, like for example that the H1 tag is a really important ranking signal and so forth. That’s probably been debunked in previous versions of your ranking factor analysis. What would be some of these sorts of myths that you can squash through your research?
There are some amazing data points on it, we also used some user engagement metrics, we took data from Google Analytics and Google Search console and bagged it into the study too. It’s also really interesting to see that successful content or content with a lot of rankings and high rankings, that this content also has better user metrics. Do they have a lower bounce rate, do they have a higher time on site, do they have better user metrics than others? It’s interesting to see because there’s a correlation and we are going to explain this also in the reigning factors, but it’s nice to see it in a study because often people have the feeling like, yes when I just have better user data or better user metrics, I should have a better performance. That’s why we put this into the study where we use a couple of million different URLs, that’s not just small assumptions, it’s based on a large data set. It’s pretty cool and it makes sense because when you have better content, people do engage with your content more than with content that’s not that relevant.
Right. Let’s take the idea of Google Analytics and one of the myths out there is that Google is using your own analytics data against you even though they publicly swear up and down that we’re not using your Google Analytics data in the rankings algorithm, that’s your private data, that’s off limits. People gets this idea that my bounce rate, my time on site and those other things are being watched through Google Analytics and that’s why it’s free because Google wants to siphon that data and use it against you. What are your thoughts?
This is a discussion that started in that moment when Google bought Urchin, the company that built Google Analytics before. I loved Urchin because I did my lock file analysis with it. This is when the whole discussion around Google is using the data started but there was never someone who proved it and there have been studies but I would never say specifically that Google is using Google Analytics data because you don’t need to because the browser with the highest market share in the Western World, you have to say Western World because except China. In the Western World is Chrome and Google has, I don’t know if data with Chrome, and even if they wouldn’t use Chrome, they have enough data of the user behavior on the search site itself. When someone does a search and then he bounces back to search, and then he clicks the next result and Google is tracking everything as we know because Google has all their little parameters behind each URLs, Google knows exactly what you click, when you click it, how long you stay, when you come back and click the next URL. Google has enough evidence without claiming that okay of course we use Google Analytics and we never told anyone, sorry. They don’t need to do it because they have enough data anyway, they don’t need Google analytics in the end.
The metric is talked about in SEO is dwell time because if you go click on a Google search result and then hit the back button and choose another search result, Google can track the time that was spent between those two clicks because now you’re back on Google and all that’s being clicked tracked. That’s dwell time versus bounce rate which they’re not looking at, probably.
Yes. But the bounce rate in that moment is this strong signal of is the user happy with your site. The thing is also that’s something you have to keep in mind too, when you are in an area where people normally do more searches. Let’s take health because we just talked about this. When people search for serious stuff like do I have Skin Cancer, they don’t know, they look at their skin and they see something weird, it’s very likely that if they search on Google for Skin Cancer symptoms that they click many, many results. Not just the first one, they read it and then they look for evidence on many, many pages. The bounce rate very likely for these topics are not a good indicator. But if you take enough topics, then it’s statistically irrelevant. For some topics like health, the bounce rate doesn’t matter because for other topics like e-commerce or finance or travel or events, for them it matters. That’s why we took the bounce rate into our ranking factor study not by saying it’s a direct ranking factor but by saying it is measurable that obviously more successful pages have a lower bounce rate than pages that are not that successful.
Yup, makes sense. There is some other research that you guys do that beyond the ranking factor studies, like looking at top winners and losers after Google makes an algorithmic update. Can you share a bit of insight from that because I always love reading what you have to say after each Panda update and Penguin and of course none of these are real time but each of these major updates. People just don’t know what the shakeout was until they read your report saying these are the top winners and losers.
That’s true, they can’t see it from the outside. If they don’t have enough data, they can only make guesses. The reason why we are doing it is of course to get some PR around it, sure. One of the big philosophies of search metrics, from the early beginning, that’s also the reason why I founded the company is being curious. I was an SEO in a company in a very competitive space in Germany and we looked at competitors and our own site and the whole market all the time and that’s also part of the winners and losers studies because if you don’t know what you don’t know, you can’t have a good judgment for yourself. That’s why we built our visibility index and this large database to have a better idea of what’s going on in general, this is the foundation for the winners and the losers analysis that we can see patterns in not just Google Updates, also patterns in pages that are all kind of lost visibilities. When you find this pattern, you can just do the opposite for your own business. When you look at the winners, you try to mimic what may be the pattern for the winner is. That’s why we publish that stuff to give more knowledge into the community that people can just take this as ideas to improve their own business. It works because when you look at, for example, all the penguin updates and now we have just the last one a couple of weeks ago and it’s real time that Google just set. When penguin came out, there have been many, many, many winners, almost no loser. Not that you saw a pattern but many winners, and for me it means when I look at the clusters of these winners is that obviously with other penguin updates that happened before, they got a penalty and they never were released all of this penalty. What Google did now, the last Penguin 4.0 update, obviously they fixed a lot of things that for example all these different links that got completely devalued by Google. All these pages that got sucked down by all this penalized blings, they now got free and got back a lot of visibility and that’s interesting. We will see what’s going to happen next, now Panda is in real time, Penguin is in real time, Rank Brain really takes care about all search queries that are maybe unusual, I believe that Ranks Brain is really the most important ranking factor even if Google is only saying it’s just the third most important ranking factor. We will see what’s going to happen next because Google is that large and there would be new increments where we’ll see winners and losers because not everything is smooth and fluid and you just see slight changes, I don’t believe in that.
Rank Brain, it’s being taunted as being third largest ranking signal. The party line over at Google is that Rank Brain is only a query analysis machine learning algorithm, it’s not analyzing the pages for quality and things like that, it’s not a rankings algorithm, it’s a query understanding algorithm. What are your thoughts on that?
Rank Brain does analyzes the query and looks for the proximity of different queries to each other, because then it’s easier for Google to just give back the same result for queries that have low proximity to each other that makes a lot of sense. Google is that smart and has so many signals that it would be waste of intelligence when they are not using their machine learning implementation, these other signals too, if we just talked about users signals. The user signals, confidence signals for Google that they made the right decision, right? If they continuously change the algorithm and then use these confidence signals that they see okay we made the right decision here and there and okay maybe on this topic we put a higher weight on backlinks because it’s a query, it’s more competitive. On the other query topic they put a higher weight on structure because it’s a query where it’s all about recipes for example, where structure data is more important. I believe that Ranks Brain is continuously analyzing every change and the way things that Google gives for all these factors and that Rank Brain in the end is a combination of many factors. For each topic, it gives them a different weight. When they do this implementation, the algorithm can make autonomous decisions. When you have all these different factors, then you use the user as a confidence signal for example, you know okay I made a change in the algorithm so maybe Ranks Brain and engineer and then you see okay, it makes sense, it was successful obviously, the quality improved, we have now better user signal engagement. That’s definitely something but I believe Ranks Brain also does and not just this stemming of search queries.
Yeah. That makes sense. I know we’re out of time but how would people who are interested in all the technology we’ve talked about, how would they take the next step and perhaps sign up for a trial with Search Metrics. I know that the topic explorer which is part of the content experience suite is still in beta and it’s not available to the public at large yet. What’ll be the next step for somebody who’s intrigued and wants to work with you guys and your tool set?
We want to have a very good release of our software so we just don’t want to send a demo login to everyone and then say here, this is the login and do whatever you want. I believe in writing good content, you need to have a good training and there needs to be a change process. We did many interviews with companies, with writers, with managers, we asked about the expectations and how they at the moment create content and it’s very scary because if you ask 10 companies, you get at least 11 different answers, at least, normally you get more. The larger the companies are, the more answers you get. What we want is we want to help and enable companies to create more relevant content in a short amount of time. It means that we have a form, if you go on searchmetrics.com you’d see directly on the homepage or if you go and searchmetrics.com/content you’ll see also the information about content experience, you can apply for a demo. One of our client success salespeople will reach out and schedule a demo. We have a very long cue of many companies who are signed up already but you shouldn’t be afraid of it because there will be a point of time in the next weeks when you get a demo and then we are also going to help these companies to train the trainers internally because you, Stephan, you wrote books and you worked with many companies, so you know how hard it is to create content. We want to help these companies to become more powerful with it, that’s why we’re not just sending out hundreds of these invites to the demo because I believe you need to have a good education around how you should leverage the software because it’s really new. We’ve invented it actually, so there’s nothing like our software on the market yet.
The key thing here is that a tool without a skilled operator is pretty much useless. You want to make sure that the tool is going to do something important in the company and that’s why you have this process where you give them the demo, you empower them to actually use it wisely.
One analogy. Imagine you sent the login to a CRM to sales people, so what do you think would happen? You know what I mean? They would say oh that’s great, thank you. And then they would just follow their daily routine, right? You need to have the guy, the champion inside the company that kind of keeps all of them together and then reminds them every day, hey look, this is how you should use it. That’s why we pick this way and not just make a very large release and really want to make our customer successful.
Yup, perfect. Alright. Thank you Marcus, thank you listeners. Go check out searchmetrics.com, apply for a demo they have the Research Suite as well as the Content Experience Suite and both are amazing. I’m a happy, happy user of both and also follow Marcus and all the company’s research and the stuff that they’re doing, it’s pretty amazing as you can tell. Be sure to check out the show notes from this episode at marketingspeak.com and the transcript will be available as well and also a checklist of things that you can do to take action and apply what you’ve learned in the past hour to your business or website. Thanks, this is Stephan Spencer signing off and we’ll catch you on the next episode of Marketing Speak.
- Marcus.tober on Facebook
- @marcustober on Twitter
- Marcus Tober on LinkedIn
- Mobile-first design
- AI first
- Searchmetrics Winners and Losers
Your Checklist of Actions to Take
☑ Examine the content you already have. If any of it is based around keywords, completely rewrite it to
focus on providing genuine value to readers instead.
☑ Rethink how you come up with your content ideas. Don’t just list a bunch of vaguely related keywords; instead, think about genuinely related information and provide value around that.
☑ Are you a strategist or a tactician? If it’s the latter, reframe your thinking and work on becoming a
strategist; you can always hire a tactician if necessary.
☑ Work on understanding your audience by figuring out what drives them to your site. Create content
relevant to the motivation behind what brought them there in the first place.
☑ Create content for your users in the format that fits the topic. Users searching for hairstyles want lots of images with relatively little text; users searching for unboxing videos may not want text at all.
☑ Pay attention to your bounce rate. Spend some time writing down how it could be lowered in terms of
user behavior instead of simply quality, and see if that gives you insight into improving it.
☑ If you haven’t already done so, view each page of your site from several different mobile devices (both
Android and iOS). If they don’t display beautifully, adjust them until they do.
☑ Look at the Searchmetrics Winners and Losers lists, and pay attention to what traits the losing sites
have in common. Use that insight to guide you in improving your site.
☑ Do the same thing, but this time look at the winning sites. Observe their similar traits, and incorporate
what you find there into your own site.
☑ Go to searchmetrics.com/content and sign up for a demo with Searchmetrics. It may take a while, but
you’ll eventually get a guided demo of the content tool.
About Marcus Tober
Well known in SEO circles as an innovative thinker and the founder of Searchmetrics, Marcus Tober also founded Ecards and More, one of the largest German online greeting card companies. When Marcus sold a majority share of the company in 2008 to Rocket Internet, it had as many as 180,000 unique visitors per day and generated up to 30 million impressions—results driven primarily by SEO. Now called Bonnyprints, it is one of Europe’s fastest growing online destinations for print products.
When Marcus began in earnest to develop search analytics software, search engine optimization was a little like panning for gold in the Hudson River – an unruly race between a handful of pioneers. In order to take his work to the next level, he devised a number of tools with which he assembled his own and others’ SEO projects into a dynamic, up-to-date database. With his comprehensive approach and an eye for detail, he quickly became one of the leading minds in the nascent field of search engine optimization. In 2005, Marcus founded his first SEO company, which became Searchmetrics in 2007, and rapidly evolved the company to become a market leader in search analytics software.
A self-described Big Data and stats geek, Marcus is frequently asked to comment on trends in the industry. Every time Google makes a significant change to its algorithm—think Panda—or upsets the apple cart with ideas like secure search, Marcus is in demand. And he doesn’t mind ruffling some feathers with his opinions.