All of us in the digital marketing world have become increasingly fixated with trying to make sense of Google’s enigmatic search engine results pages (SERPs). But there a very few data scientists on the planet who can provide as much insight into the nuance of Google’s SERP rankings as Moz’s Pete Meyers (aka Dr. Pete).
Photo courtesy of Rob Timar Photography
Pete is a cognitive psychologist, computer scientist, founder of the User Effect, and now Marketing Scientist at Moz where he conducts experiments on the SERPs. But most importantly and often neglected, he’s a brilliant author and definitely one of my favourite online writers. I interviewed him during SMX Sydney this year and we talk about Dr. Pete’s start-up days, MozCast, Australian SERPs, international SERPs, Google’s mobile update, content writing and much much more. Here’s how it went down…
Woj: So let’s talk about how it all began, your involvement with a start-up and the success that took you to Moz.
What got you interested in the web?
Pete: Yeah, my undergraduate degree was in Computer Science, which I finished in 1992, and I went to graduate school because in 1992 nothing was going on in computer science other than just writing code. You could write code for ATT or Bell Labs or whoever.
Then during the five years I was in school suddenly the Internet exploded and everything changed, and the kind of jobs I had wanted suddenly appeared. So I decided to give it a shot, and went to work for a start-up. It was the typical late nineties.
It was me and the founder in an abandoned warehouse – his dad owned the warehouse – and we had no business model.
Woj: It’s like the Apple story.
Pete: We built five website that we were hosting for $50 a month. He was gone three days a week, so I started to pick up the phone.
The questions I got were the same kind of questions we get now: “why am I not ranking?” Back then it was Yahoo and AltaVista but the questions were the same, “why am I not appearing on this?” and, “what can I do?” and, “we need to fix our site.”
Fast forward Eight years and we were a $2 million software to service company. I never thought I’d be there for eight years. So I left and started to look around, and decided to go on my own.
One of my clients who I’d worked with for years had always tried to get me to go to SES, but it didn’t really fit our business model at the time. But since I was on my own, I’m like, “Okay let’s go.”
And I saw Rand there, plus I saw how much SEO had evolved between the late nineties and 2006. How much was different… How much search quality had improved…
It wasn’t just this game of keyword stuffing and that kind of stuff anymore. So it renewed my interest and I got back into search and, long story short, I started to do a little work on the side for Moz with Pro Q&A, you know, answering sort of advanced questions for people, and on a lark I had the idea for MozCast.
I built a 50 keyword version of it and it seemed to work. So that turned into this whole thing, and a year and a half later it was a full time job. So I’ve been there two years now.
Woj: It feels like longer.
Pete: I’ve been involved with the community since 2007. But I’ve only really been in this role for two.
Woj: With the Moz community everyone’s so aligned, and it’s so close knit, that you probably did feel like an actual employee.
Pete: Oh yeah, yeah hehe. Especially early on. It’s almost like a graduating class, the people that joined Moz when I did were people like Will Critchlow and Ciarán Norris, who I just saw the other day. He’s down here now, for Yahoo. Now all these people have gone onto other things, or have gone on to build companies, so it’s like following the progress of your college class. It’s funny how it’s evolved.
Woj: I love turning up to Mozcon because you feel like everyone’s of a similar mindset, so you can just walk up to anyone and they’ll give you the time of day to have a conversation.
Pete: Yeah. It’s almost half search conference and half user conference in a way, so you get that.
Woj: So thanks for setting up MozCast, it’s a great resource for tracking the SERPs flux.
Are there any plans to enhance MozCast or introduce country specific versions? Maybe in Celsius for us Aussies?
Pete: We only tend to do small research projects internationally now, like what I did yesterday. Truthfully, it’s just that we don’t have a lot of front-end resources, so I build the back-end and it’s really easy for me to add things like a new feature. But as soon as it turns into building a new layout it becomes a product problem, and we have to juggle it against our product.
Woj: Surely there’s designers in the Moz community that could put their hands up?
Pete: The country stuff is hard because we really need a full crawl for every country and that, resource-wise, that means irritating Google even more. That’s why I’ve been trying to explore country by country just doing these small studies and seeing how people respond. But in a way I’m much more interested in where the feature graph is headed than where core flux is because the algorithm changes so much.
That’s really hard to filter out. The signal-to-noise ratio is really bad. And so I feel like the nicest thing I hear about MozCast is when people just basically say, “Thank you for letting me know I’m not crazy” and that’s really gratifying, but when we try to dig in to say what caused that or what does that break down into, that is an extremely difficult task. So I’ve become more interested in what’s happening around the SERP. But yeah, country by country, we wish we had the resources to do more. I wish it wouldn’t take 200 times the IPs. LOL
Woj: Yeah, if only you could scale it easily. So the presentation yesterday was really good. It was Australian focused.
So how different did you find the Australian SERPs compared to the rest of the world?
Pete: Compared to us, they were less different than I thought.
~90% of results in the AU SERPs contain >=1 non-organic entity (paid, local entities, enhancements, knowledge graph) @dr_pete #SMX
— Woj Kwasi (@WojKwasi) May 11, 2015
So I think on the organic side virtually everything that’s rolled out in the U.S. has rolled out in Australia, feature-wise.
The paid ecosystem is different and I think there’s more regulatory pressure.
We see it in the EU too. There are things that Google’s rolled out in the U.S. that they can’t get away with in the rest of the world, paid-wise. But organic, it’s virtually the same. I did a talk in the Czech Republic last year, and truthfully, it was the same. There were a handful of things they didn’t have yet feature-wise, but in many ways it was very similar. More and more Google’s rolling out things globally.
They have some challenges with things like language sets, where they might have to have a better understanding of the local language for things like news. But in terms of what’s available, it’s everywhere for the most part. And I know that the mobile update was global.
We’re seeing that more and more. Something would roll out and three months later it would hit Europe, and then three months after that it would hit the rest of the English speaking world, and then it would hit the rest of Europe, and then it would hit the Asian languages, you know, a year later hehe.
Woj: Yeah. And…
…Is there a bias to English language search engines?
Pete: I think it’s about expertise. To me it comes down to the fact that that’s where they started, that’s where their money is, and that’s where they have the most expertise.
Woj: Sure. Of course, makes sense.
Pete: And every market they go into, they have to build into. And so, is it a bias?
Yes, but it’s a bias towards where you started from.
And they certainly had trouble in some countries. They made a lot of missteps in China that’s made things difficult. Then they have to get around Baidu and other market forces. It’s just that they have to build a team to do that. And they have a lot of international teams. There’s something like 200 CCTLDs and around 250 countries in the world that I know of!
So Google almost have a CCTLD for something like 80% of the countries in the world.
Pete: So they’re getting there.
Did you find much difference in the SERPs when you compared U.S. English Google versus European ones at Marketing Festival in the Czech Republic recently?
Woj: I know that in some cases there’s suspicion, if that’s the right word, that there’s some sub-algos that get rolled out into different markets.
Pete: Yeah, at the time they didn’t have the Pigeon update yet. That took a little while, because that is so hyper-local. So it really takes understanding.
Pigeon also started to look at local intent and queries differently. So I think they struggled with that in some other languages.
I also saw some differences in things like news and in-depth, where for in-depth you not only have to be able to understand that the query is news related, but you have to have local sources.
So if you don’t have good local sources in the local language, it’s really hard to put out in-depth articles. So until they have a base of trusted sources they can use, or an algorithmic way to build that, it becomes a country by country roll out. They have to say, well if your query is in Czech, you probably don’t want the New York Times showing up. You want a trusted local source. You can’t cheat that. That has to be built.
Woj: Pretty hard to determine.
Pete: What’s interesting is when we ran English queries in Google.cz, we wouldn’t see a lot of features. But when we switched to the Czech query set those features jumped back up.
Woj: Oh wow.
Pete: It was very contextual. So things like news are very contextual. So if you’re typing English queries in the Czech Republic they might just go, “That’s not really a news search here. That’s not a brand here.” But then you run something like a Czech brand in the Czech Republic and that’s big here. That same brand over there might get a knowledge panel that same brand here wouldn’t.
So that contextual understanding of what’s a brand and who’s famous and what sports they care about, and all these things, that’s become really rich. So we see a lot of differences there.
Woj: Seems like everything’s about the contextual component of the query, the user…
Pete: Yeah, and especially with local now, more things have a local flavour and mobile is inherently local.
Woj: Like with voice search, as we saw earlier today.
Google is pushing changes into mobile more, but when you look at the search results above the fold on a mobile device we’re seeing even less organic listings than on desktop. Is this part of Google’s game plan to generate more revenue?
Pete: I don’t think so. I think right now it’s just the reality of real estate on the phone, there’s just less. People are more willing to scroll and so I think the revenue aspect is that when they realised mobile adoption was climbing and CTR was falling, they got scared.
Woj: Had to do something.
Pete: And they had to do something. That’s when mobile first design happened and that was an economic thing. Is the future of search mobile? Maybe. But that’s not why they made the decision.
They made the decision because the click through was falling and they had to reorganise around mobile.
Plus I don’t think they are designing SERPs so that there’s only three results on a page because that works. I think they’re doing that because that’s as big as the phone (screen) is. But they are definitely designing to offset that loss and make sure it doesn’t fall any farther.
So what’s the future of the SERPs likely to look like, given Google’s revenue agenda?
Woj: It seems like more blended paid initiatives and ‘keeping people at Google’ is on the cards.
Pete: I think we’re going to see more on blended paid, and I think we’re going to see more knowledge graph. And voice search. When people were asking today, where there’s just an answer and how do they monetise that? I don’t think Google knows. They know that’s what we need. We need small screens, we need voice.
How do you monetize a voice answer if there’s no SERP? They don’t know and they have to figure that out. So I think they’re struggling with that too.
I think knowledge graph ads are very contextual and they like that. It isn’t a bad thing in the sense that it’s a very well defined information space, and so you can advertise within that your clicks are going to go up. Your conversions are going to go up. You’re more likely to have a targeted answer. So I think they view that as a kind of targeting.
What we’re going to see in the next few years is that familiar ad block on desktop, with three at the top, eight on the side, and three on the bottom being broken up. The EU kind of forced that ad label. So what used to be a block is now really just separate ad entities. So I think you’re going to see an ad and then two or three organic and then an ad and then a few organic and then an ad and they’re going to start to play, to blend that a little bit. We’ll probably see that on phones first. Or it might just become an extension in a sense.
So I guess their changes are arguably focused on usability, given their mission statement. How important do you think it is for brands to focus on the customer as opposed to beating their own chests?
Pete: I think it’s just a revenue issue. You can’t do everything for Google and complain about Google, and then build a site that doesn’t sell a product and doesn’t make money. That’s self-defeating. I think, in a way, what Google wants is to model the world, but then we criticise them for favouring big brands. Well the problem is, if you type in a brand name and that brand isn’t first, that result is wrong. Regardless of signals, that result is probably wrong because in the real world that brand has the most presence.
Woj: It’s known, yeah.
Pete: They have to look at cross signals. So they’re going to start to look at things like your own performance within your site, and how much people talk about you. These signals have to corroborate, so it’s not good enough to have a site that has all these tweets, but has no links, no traffic, no usability, and no conversions. That doesn’t represent anything in the real world. You have to have this cluster of things together.
Woj: It’s like having a pristine bricks and mortar store, but no one coming into the store.
Pete: Right, right. Yeah, or no door. You have a great sign but you have no door!
People tend to be quite risk averse online – What are some simple steps we can take to bridge the gap between what we “can do” and what we “want to do”?
Pete: I think especially for small business right now, you still have to step back and look at your weaknesses. We used to see this with on-page versus link building. Somebody would really get into on-page and then they would just hyper tweak their titles or rewrite all their content and they’d become obsessed with on-page stuff. And you’d be like, “Well you have no links and no social. Nobody’s heard of you, you’re trying to get from 98% to 99% with your on-page factors.” And going from 98 to 99% costs a lot more than going from 60 to 65%.
Every little step is going to cost that much more, plus there’s diminishing returns. I think for a lot of businesses it comes down to saying, “Okay look, we’re not going to obsess about our on page structure, because nobody’s heard of us. We have to get out there and build awareness so that there’s people linking to us and there’s something to link to.” That’s the content story.
I’m not going to sit here and say content is king or sugar coat it and pretend that’s the only thing, but if you have nothing to link to and nothing to talk about then you’re dead. You have to have something.
On the flip side, and we hear this with big enterprises such as news sites, some sites have tremendous content and brand equity, but their on page stuff is a disaster. I wrote a post a little while back about what’s more important, links or on page? It completely depends on where you’re at and where you’re hurting. To focus on the one that you’re already good at it doesn’t work. So you have to have that awareness.
You have to be a real entity in the world, even if you’re an online store. So I think it’s going to get a lot harder for pure affiliates who aren’t . . . I don’t want to say they’re not real, but they’re web only. They’re selling other people’s products, they don’t have a physical presence, they don’t have unique content of any kind, and they don’t have a unique thing to sell. Life’s going to get a lot harder for them.
At MozCon 2012 you said, “if you’re still tracking rankings the way you did five years ago, if you’re sitting on ten year old content and strategies from ten years ago, and just riding your coattails and hoping everything will stay okay, it won’t. Even if the algo doesn’t change, 80% changes every day in the SERPs.” That really resonated with me and seems like pretty evergreen advice. Now it’s 2015, how do you feel about that statement today?
Jump to 40:18 for the quote.
Pete: I think it’s only accelerated. The index is getting updated more frequently. We’re getting into more dynamic entities like news, where it’s changing every day. More is being added to the index, so I don’t think you can stand still.
We’ve seen it with organic when people get upset and they go, “What did Google do? I’ve ranked great for six years and now I don’t.” Well Google didn’t do anything, it’s just that neither did you. You haven’t really changed in six years and you want to ride this forever, but it doesn’t work like that anymore.
Woj: Google has evolved. It used to be like a little primitive mollusc and now it’s this huge artificially intelligent being… and it’s maintained its mission statement consistently! It’s evolved so much that if your business is stagnating then you’re just shooting yourself on the foot.
Pete: Yeah, I think they see that.
I touched on the fact that you’re a great author, with an awesome sense of humour. The content you put together is very engaging. I love the clarity with which you communicate your data and research.
How important are data visualisations for communicating your findings?
Pete: Yeah it’s important, but I think sometimes we get obsessed with trying to be fancy. It’s kind of funny that so much on my decks is just bar graphs and pie charts hehe, because I’ll go to do that new, cool thing – like Avinash loves – and I’ll go to do that new fancy thing like make a chord chart . . .
Woj: Chord charts, yeah, they’re crazy.
Pete: . . . or sunbursts and then I go, “I don’t know what that meant.” So I really like Edward Tufte‘s stuff. I saw him last year at a seminar, which was cool. I think there’s a lot to be said for simplicity. So there are times when a visualisation communicates well, but when we obsess with things like infographics, and try to make everything visual all the time, that gets to be a little self-defeating.
Sometimes the best way to say it is with words. Sometimes the best way to say it is with a table. Sometimes it’s a bar graph and sometimes you need something more complicated, because it’s a complicated problem. But sometimes we like to make things complicated just to seem smart.
Last year at MozCon your talk was about generating content ideas. Who or what inspires you to write?
Pete: I’ve always enjoyed writing, but I’m not the type of person who knows what topic they’re really passionate about. I’m sort of passionate about whatever I’m deep into at the time, so a lot of my Google work is just what’s in my head at the time. I just want to dig into something, and then I want to dig into the next thing, and it comes out of that.
Woj: But I mean like the logo ideas and the stuff on your own blog, what inspires you to look into that area? Or that body of work?
Pete: It’s sad how much just comes down to entertaining myself. Like all the logo stuff, that whole blog came out of my basic love of minimalism. Maybe I love minimalism because I’m a bad designer, I don’t know. I feel like, “well, maybe I can do that,” which isn’t an insult to minimalism because to do it well is really hard. But then I think brands do a lot of weird things and because I’ve been a start-up person I kind of always make fun of that.
That whole blog started as a joke with myself. Sometimes that’s where it comes from. I guess I have a little bit of a class-clown personality. The amount of times I’ll spend 15-20 minutes researching a joke for twitter which has no value at all to anybody, but I’ll still go and look up facts and Photoshop something together.
Woj: That’s cool. One final question – I’ve seen you speak a couple of times, and it’s always insightful and super engaging.
You always end your talks by alleviating fear and focusing on opportunity. Why is that important?
Pete: I get frustrated when people look at what I’m doing and get depressed. I want to give them something, an actionable takeaway. But it can be hard to give a one size fits all takeaway for this stuff because it’s becoming so niche. I think what’s important to me is just the practicality of it.
Google isn’t going to stop doing things just because we don’t like it, they’re not going to stop evolving. They’re a massive company driven by investors and advertising. They have to move forward, and we have to adapt. It’s about pragmatism.
Woj: Well thank you for your time. I really appreciate it. I’m looking forward to MozCon coming up again, should be good.
Pete: Yeah I know, it’s coming up so fast!
There you have it folks! I hope you enjoyed that as much as I did. Please feel free to leave your comments below.
One final thing – at a dinner after the conference, I asked Dr Pete if he had tried any fine native Australian cuisine… you know… the regular stuff: kangaroo, crocodile etc. He asked if we eat wombats and I was shocked but also quite amused. It lead to a lot of laughs and even Gary Illyes got involved:
Please don’t eat wombats, @dr_pete #smx pic.twitter.com/0vIUY2GxhL — Gary Illyes (@methode) May 14, 2015
You may also like my other interviews:
- Rand Fishkin (Moz) – How Authentic Wizardry Nudged Moz to the Top
- Nathalie Nahai (The Web Psychologist) – Things That Make You Go Click
- Jon Cooper (PointBlankSEO) – Don’t Link Like You’re Not Impressed
- Paddy Moogan (Distilled) – Link Building Lessons
- Mike King (iPullRank) – Inbound Marketer Without a Pause