Episode 1 of our “SEO in 2020” podcast interviews Rand Fishkin from Moz; getting his views on how he thinks SEO is likely to evolve over the coming few years.
DAVID BAIN: So Rand, will we be calling it SEO in the year 2020?
RAND FISHKIN: Yeah, probably. 2020, you know I was thinking about it, it’s only four years away. If I think about what SEO was in 2012 versus 2016, yeah, there are a lot more opportunities, and there have been some tactical changes, a few things that don’t work, a few things that sort of are working well. But honestly, this is not a practice that has changed dramatically, at least the fundamentals haven’t changed dramatically in the last four years, and I don’t think we’re going to see them change that dramatically in the next four. I think that search has reached a more maturing stage. You know, you could argue that there’s going to be growth in voice search. There’s certainly going to be a lot of the volume growth in search worldwide is going to come from developing countries rather than the developed world. We’re going to continue to see mobile be a big trend. I suspect the engines, Google primarily, are going to continue to make some technological advances in how they understand queries, and how they return quality results, and those kinds of things.
But we’re in a maturing stage. This is not 2005 or 2006 anymore, so I think we can more effectively predict at least three or four years out with some certainty, barring the fundamental paradigm change, right? Maybe VR means that we strap on helmets, and go into VR world to discover information, and search volume drops off dramatically in traditional search. I don’t see that happening quite yet.
DAVID BAIN: Intriguing that you use the time frame four years, because four years ago as well, you say it hasn’t changed a lot in four years. Was there a particular event four years ago that you’re thinking of in particular that led to a significant change in SEO?
RAND FISHKIN: Well, certainly if I think about Panda and Penguin, which were both in that time frame, that sort of 2011/2012/2013 time frame, those were both relatively big shakeups for SEO professionals in that they changed a lot of things tactically for us. But to be honest, no. If you were doing SEO the right way, long term way, in 2012, that would really be quite similar to how you’d do things in 2016. And in fact, the craziest part is it wouldn’t be all that dissimilar to how you did it in 2008. So Google’s had this goal for a long time of delivering the highest quality, best experience, most relevant information to searchers as quickly as they possibly can in a way that provides a great user experience. And if you were doing that in ’08, certainly, like I said, some of those tactics have evolved, but if you were following that strategy rather than sort of a hey, here’s this one exploit in Google that works today that might not work in the future, if you were following that long term methodology, you probably had a great run because the fundamentals behind SEO haven’t shifted all that much.
DAVID BAIN: It seems to me that SEO is splitting a little bit into different specialisms; you’ve got things like UX, website performance, content marketing, maybe to name a few. Do you think it’s reasonable for an SEO to be able to do everything? And looking forward a few years, should they not look to become a specialist in one of those areas rather than a generic SEO?
RAND FISHKIN: Yeah, I think any time you have an industry like ours, where it’s growing and maturing, it’s going from a few tens of thousands of practitioners to hundreds of thousands, and now many millions of practitioners, it’s becoming a standard for every business in every organisation. You know, it’s something you just have to deal with and take care of. When that happens, you start to see more and more specialisation, and the importance of specialisation growing. So today if you are a technical SEO who deals with crawl issues, and that is a specialisation of yours, you can probably make a great living either as a consultant or an in house SEO for the right brand or company working on that. Likewise, if you’re an SEO who specialises in cured research and targeting, you’ve probably got a great pathway there. If you’re an SEO who specialises in content marketing, in creating content that tends to perform well in search engines, and the written word, and other types of visual or interactive or audio content are your specialties, you can probably do great there. If you’re a video SEO who does stuff for YouTube, awesome, you’ve probably got a great living there.
So yes, I see a lot of segmentation. I think that classic diagram I drew a few years ago of the T shaped marketer – this idea that you have essentially knowledge of many, many disciplines in web marketing. Right, so I have SEO knowledge, but I also have PPC knowledge, I have email marketing knowledge, and I have content knowledge, and I’ve got knowledge of paid campaigns on Facebook, and I understand social media, right, but then I’m a specialist. I go very, very deep into one of those. It’s almost happening again in microcosm for each of those.
So if you’re in the email marketing world, building up an email subscription list for a new start up, versus email outreach, versus optimising an existing email list, or working on email migrations, or setting up recurring emails for SAS, and subscription based businesses, you can find the specialisations within the specialisation, and the T shape inside each of those eyes of the T.
DAVID BAIN: One thing that I certainly didn’t hear you mention there as a specialism was links. If we can possibly look forward a few years, do you think that links are still going to be an integral part of Google’s algorithm in the future?
RAND FISHKIN: Let’s see. I think they will still be in the top five most important things that you have to think about for SEO, but I think it’s possible they won’t be one or two, which they still are today. And I think that will promote a very tactical shift for a lot of marketers. They’ll be thinking much more about content and user experience, and social signals, and user engagement, and usage signals, and possibly new signals that haven’t emerged yet that we will learn about in the future, and we’ll be thinking about those a little more than we think about links.
One of the things though, I think for every year that I’ve been in the SEO field, which is fourteen years now that I’ve been doing SEO in some form or another, there have been people predicting the death of links as a ranking factor, I think because they’re gameable, and the engines have complained about them a lot, and they continue to get harder and harder to earn the right ones. You know, you have to go through a lot of effort to do those, and manipulation of the link graph gets harder and harder.
So I think there’s this natural prediction that someday Google’s going to get rid of them as a ranking factor. I’m not sure. I think that they are still useful. I think in fact, over the last four or five years, they’ve become more useful than they were four or five years ago as many marketers, particularly in places where Google’s doing a great job with spam fighting – English language countries, Western Europe, they’ve really done a great job. And then a lot of other places they’ve done a less high quality job, in my opinion, fighting link spam so far. And that could happen in the future, but so far.
And so as a result of that, the link graph gets higher quality. If we’re all scared to manipulate the link graph, well guess what? That graph becomes a higher point of signal versus noise in the ranking algorithm. It almost gets better. I think the social web in some ways has helped the link graph to become more high quality. Not just links from social sites themselves, but the links that are created because people find things through social media, and through apps, and then link to them in the content that they create, that sort of gives Google more of those kinds of signals.
So I’m a little bit of a sceptic when it comes to getting rid of links entirely, but I wouldn’t put it out of the realm of possibility that it drops in importance in the next few years.
DAVID BAIN: That’s intriguing that Google’s scaring the living daylights out of SEOs with regards to buying links, but has helped them actually build a higher quality network of links between different websites out there.
RAND FISHKIN: Yeah, yeah.
DAVID BAIN: But it’s not going to be number one, so what’s going to be number one? Rand brain?
RAND FISHKIN: Oh god. It’s so weird. You know, there’s this odd thing where I sort of feel like I live apart from this weird character of Rand Fishkin. There’s like the character, and caricature of Rand, and then there’s me. And I just walk to work, and try and do my job, and then create good content, and help Moz be successful, and try to lean on our big data team to do their jobs better, and lean on the explorer team to do their jobs better. And then out in the web, it’s like oh, you know, I don’t understand. There’s a whole weird religion of making me into something that I’m not. I cannot emphasize to you enough how unimportant I am in the grand scheme of things.
DAVID BAIN: I’m sure your wife helps you. Certainly my wife helps me in that sense.
RAND FISHKIN: Oh yeah. No, she makes sure that I keep humble. So back to your question though, with regards to the things that I think are going to increase in importance, look I think it is already very important, I think it will continue to be important, but Google doesn’t talk about it much. I think they are concerned that there’s still a lot of manipulation potential, and I want to be very respectful and empathetic to that. So I understand why they haven’t talked about it. But user and usage data signals, in my opinion, are on the rise, they’re here today, they’re powerful, they’re influential, we can test them and observe them with relative consistency – not perfect consistency, but relative consistency – and we can see that things like lots of people searching for an unbranded term in combination with brand tends to associate that brand with that non branded term, and tends to mean that that brand, that website, is going to rank higher. That just is what it is.
In terms of other things, I think that content evaluation is something Google is doing a better and better job with, and we are seeing them grow the number and types and diversity of signals that they look at from content. I think they’re getting better at parsing the meaning, not just from queries, but from content itself. And that means they can look at a sentence, a phrase, a fragment of a sentence, a paragraph, an entire document, and say this is what this is about, and this is how relevant it is to these particular queries versus relevant to these other queries. The days of raw keyword matching are really at an end. I think today we almost do keyword matching as SEOs more for searchers than for the engines themselves. Engines are getting smarter and smarter about being able to connect up synonyms and phrase meanings, but searchers, when you search for best restaurants in Rome, I still want to visit a page that has as its headline “Best Restaurants in Rome” or “Rome’s Best Restaurants”. I’m not looking for that sort of broad synonym fragment.
So I think that content evaluation is another one. I do think Google is going to get more into some measurement of social signals, and that could be exclusively through user and usage data, and visitor data, but I think that they may have designs on getting more seriously into social – being able to say these are influential accounts, and this is what these accounts are sharing and saying, and we want to try and understand the meaning behind them, and whether they’re endorsing things, and maybe we want to give some of those endorsements link like powers. I would definitely think that way.
DAVID BAIN: Do you think that we would still see Google+ in 2020 as some kind of social network? Is there any chance that Google might have purchased something like a Twitter in that time frame?
RAND FISHKIN: Yeah. I mean, if I were them, I would be thinking about buying, or I would be thinking about doing API deals like what they’ve done with Twitter right now, where they’re getting access to the full fire hose, and they can put that in their results, and they can use it however they want. That is really how I’d be playing it if I were them. I think Google+ - I don’t know. Maybe they have one more big push in them to try and build a Google centric social network, but it feels like maybe they haven’t hit that yet, and maybe they should give up, and try some other strategies. I don’t know.
DAVID BAIN: Okay. And I know it’s tough to imagine what might happen, but it’s also intriguing that there are so many social hubs that people go to now. Obviously Facebook, you’ve got Snapchat, Twitter to a certain degree, and you can’t really say that about Google at the moment, and you have to think well, that’s part of the movement with the modern web, and Google has to involve itself quite significantly with that to be successful in the future you would think.
RAND FISHKIN: I mean, I would imagine that if you were sitting in the Google boardroom talking about long term strategy and long term risk, their primary value chain, and how they’re generating 70, 80%+ of their revenue, the addiction to social, and the potential that social becomes the way that people find and discover information has got to be at least discussed as a major threat point, and I can’t imagine that they’re not thinking about how do we mitigate that threat? One of those mitigations obviously is the relationship with Twitter. I think that’s pretty smart. Another mitigation is they appear to be doing a better and better job of crawling Facebook, and potentially using, I don’t know, at least link discovery from Facebook, but maybe some content and sentiment analysis, maybe some traffic analysis, those kinds of things. You know, I don’t think they’re looking at likes and shares. Well, let me rephrase – I don’t think getting a bunch of likes and a bunch of shares is going to boost your rankings. But I will say this, and we’ve observed this, which has been really interesting. I don’t have enough data to prove it yet, but I’m going to speculate here. I think if you put a piece of content up on Facebook that nobody links to, but many, many people like, and share, and comment, and many people visit and engage with through their browsers of whatever type, Google finds ways to rank those. And I don’t know whether that’s directly through them crawling Facebook, and indexing that content, and those links, and all that kind of stuff. I don’t know whether it’s them looking at some of the social share signals, and then looking at who was sharing, and whether they were real people and influencers. I don’t know whether it’s them doing traffic and visitor analysis, but it is happening. It would be hard for me to prove, like I said I don’t have the data to prove it, but that is my strong feeling. You want a really interesting one?
DAVID BAIN: Go for it, yeah, absolutely.
RAND FISHKIN: Yeah, so alright this is very, very early, but I was at a search love conference, I think it was in London last year, and someone mentioned hey, we noticed this odd phenomena where we had written an article about web marketing of some kind. It wasn’t ranking particularly well, like it didn’t blow anyone away, but it was picked up by the Moz top 10 email lists. And if you’re not familiar by the Moz top 10, it’s subscribed to by a little over 300,000 marketers, it gets very high engagement, the top few links on there often get 10, 20, even 30,000 visits sometimes when the newsletter comes out. And they said when the Moz top 10 came out our rankings shot up. Yesterday, I got an email from someone that the same thing had happened to them, and that’s actually the third email I’ve gotten from someone who’s been in the Moz top 10. And so a few months ago, we started tracking some keyword searches around things that we were putting in the Moz top 10 to see if the rankings went up after the newsletter came out, and there were a few testing problems, but early signals were damn, that’s weird. It sure looks like when something goes in the Moz top 10 email, its rankings go up, especially the ones that get more clicks.
DAVID BAIN: And might your hypothesis be that the fact that that site had suddenly got a surge in traffic resulted in the increasing rankings? And if so, is it possible, potentially, with pay per click advertising to buy traffic, and have that result?
RAND FISHKIN: I mean, so if I were Google, here’s how I would be doing it. I would be accumulating all of the click stream data that I can get, and they have full click stream data from Chrome, from Android, from all the desktop devices where they have plugins installed, from Google search itself. They have just tons and tons of data. They have Google hybrid data, all the web infrastructure that they’re part of where they can see all these clicks and stuff, so they’re collecting a tremendous amount of data just where are people going on the web, and what are they doing while they’re on the web?
If I were them, I would take all those signals, and then I would cut out anything that is paid, so I can see anytime traffic comes from a paid source. Anything that’s organic, a type in, a bookmark, direct, referred from an organic email list, coming through social media like Facebook, or through the Moz Top 10, whatever it is, lots of people clicking on a link on the homepage of Yahoo, lots of people clicking on a link in Hacker News. If I saw a lot of those signals, they were coming from an organic source, I might be very tempted to give a ranking’s boost, because probably we’re seeing a highly relevant, new, useful piece of content that searchers would want to see. That is total speculation on my part. Google has never confirmed that they do this, but they definitely are collecting that data.
DAVID BAIN: I think that would certainly make sense, but potentially only if it’s relevant traffic that spends a decent length of time on the website. Obviously if you just drive loads of traffic at a website, stay for five seconds, surely that’s not an authoritative signal.
RAND FISHKIN: Totally agree. You want to see traffic, you want to see diversity of people visiting, so lots of people from lots of different places with lots of different Google accounts that have real history that are clearly not just bots or people who are paid to visit a site through Mechanical Turk or something like that. You want to see diversity of IP addresses and cookies. You want to see people who haven’t ever visited that site before visiting that site along with people who’ve been loyal to it. Right, all the things that would make it look like hey, this is a natural boost in organic traffic, and then you want to see behaviour. Positive behaviours are things like huh, you know, a good amount of people after visiting this piece went on to drive traffic to it in some other way. They shared it on their social media accounts, they emailed somebody about it. They chatted it to somebody. They saved it in some app of theirs. They copy and pasted the URL. These are all indications to Google that like huh, you know, I think maybe people really like this. And that is a good sign that it should rank.
I mean, it would be crazy to me if there haven’t been a bunch of Google engineers sitting around like so, see all this data that we get from Chrome and Android? What do you guys think about using some of it?
DAVID BAIN: Makes perfect sense. Just finally in relation to this actually, we’ve seen a big change in the SERP recently with the disappearance of ads from the right hand side, and many commercial queries now having four ads on top. How do you think the SERP may evolve over the next couple of years? Are we going to see some significant differences there?
RAND FISHKIN: Yeah, well Google said look to the mobile device. They’re going mobile first, desktop second. So I think a lot of the evolution that we’re going to see is going to be mobile centric, which I’ll be totally honest, I think it kind of gives short shrift to desktop users. Making the mobile experience and the desktop experience the same means that you make the desktop experience worse. That’s just my opinion, I’m one guy, whatever, but I don’t totally love it personally. That being said, Google’s going in that direction, that’s what they care about, so I think that we can take our cues directly from mobile. One thing that you see in mobile right now that we have not seen as much in desktop search is if you have a mobile device – let’s see if I can do this for you. I searched yesterday for men’s haircuts, and a Pinterest—well actually you can sort of do it with Google. Let’s see if I can turn this so I’m not so backlit.
DAVID BAIN: So for audio replay viewers…
RAND FISHKIN: Yeah, can you see this? So do you see that scrolling left to right behaviour?
DAVID BAIN: Yeah, so we see left to right image scroll here on the top of a SERP result.
RAND FISHKIN: Yeah, yeah, so that I think is something that has not yet made its way very far into Google’s desktop search, and I think that might be something we see more of. You know, we’ve had that carousel up at the top for some different types of knowledge, but I suspect we’re going to see more of that interactive behaviour inside a search result to get at the core of a query, and the core of what a searcher’s looking for. I certainly think that there are going to be more of those interactive elements.
DAVID BAIN: I’m sure Google will be thinking about bigger screens in the future as well because there are many tablets, many mobile devices, but perhaps people might not use desktop so much in the future, but they’ve got a great big screen in their living room. So surely they’re going to be browsing the web on that.
RAND FISHKIN: Yeah. It’s weird that we haven’t seen as much of that. We’ve seen that with YouTube, and I think that Google is smart to go into the content play world, where they’re essentially saying hey, why don’t we make some content the way Netflix and Amazon, and those folks are? Smart. But yeah, it’ll be interesting to see where that goes. I certainly expect, you know, Google is launching tons of SERP changes all the time to try and get that engagement up, to try and answer people’s questions faster and better. I have no doubt that there will be many, many changes. I think that’s one of the reasons that it’s so important that projects like Moz cast exist. You know, this is Dr. Pete’s monitoring of search results, and if you go to mozcast.com you can see here’s all these SERP features, and here’s their rise and degradation over time, and here’s the prolific-ness of a particular domain in search results. So keeping track of those features I think is really important. That’s something that we’re thinking hard about with regards to software too. I think a lot of software vendors are thinking about that, like hey, we need to tell search marketers, and marketers of all stripes who care about their search traffic what does it look like.