What Goes Into Tracking Google’s Algorithm? – SEO Podcast
The In Search SEO Podcast
You can keep up with the In Search SEO Podcast by subscribing on iTunes, by following the podcast on SoundCloud, or by subscribing to the In Search Newsletter!
The Benefits & Complexities of Tracking Google’s Updates & Trends: Summary of Episode #79
The great Dr. Pete joins the podcast to share his experiences and thoughts on analyzing Google’s updates and SERP feature trends:
- Why analyzing Google’s algorithm updates are even more complex than many think
- Why SERP feature trends play an important role in SEO
- How to understand Google by looking at what’s changing on the SERP and in the algorithm
Plus, we look at how effective Search as Journey really is and how it can improve dramatically!
Mordy Oberstein (Host)
Sapir Karabello (Co-Host)
Dr. Pete Meyers of Moz (Special Guest)
In Search SEO [Episode 78]
What’s Google’s Goal? Why Google is All About Authority
Is top of the funnel content under attack? How the SERP is set to change
Google’s May 2020 Update: Winners, Winners, Winlosers, and Why It’s All Probably Crap
How to Handle Zero-Click Searches
Google Rich Results Test Tool Out of Beta
Mobile Image Results Get Knowledge Cards
Google Ads App Now Supports Manager Accounts
Google Ads to Ban Spyware Surveillance
Is Search as Journey Effective? [00:03:53 – 00:19:54]
Last week, we talked about how Google is very likely to go product heavy and how that leaves searchers looking for information prior to buying in the lurch. Today, we’re going to look at this from the perspective of Google needing to take Search as a Journey out of infancy.
Search as a Journey, born in the fall of 2018, was meant to get the user to where they want to go and guide them along the way. And it hasn’t exactly done that yet!
To recap, last week, we discussed how PLAs are showing up before the informational results for informational commerce queries. In this instance, Mordy said the solution is for Google to make sure that the Research Carousel is always, or almost always, above the fold.
The problem is, Mordy believes this sort of problem is not only related to certain commerce queries but that it is systemic.
We talked last week about how Google has conflicting goals at times. In specific, Google’s need to push products on the main SERP to support shopping puts it at odds with what’s best for a certain and prominent demographic of users.
Well, Mordy thinks that that very problem along with Google’s push for answers puts it at odds with Search as a Journey, or at least a vital component of it. The problem puts Google at odds with the accessibility of deeper, more targeted, and overall more substantial content! Why? Because there are not a lot of access points to that content.
For example, the “Interesting Finds” feature you would think would show you deeper content, but not, it just shows the same kind of thing you see in the organic results. It’s literally the same URLs far too often.
In a lot of ways, and here’s the problem, Google is incentivizing the creation of highly-detailed and highly-nuanced content. For example, SERP filters that point the user towards more specific content and the lack of opportunity for your average site to rank for top-level content (because you’re competing with super authorities) all push for the creation of deeper and more specific content. Even answers on the SERP incentive you to write deeper content as they eliminate a lot of the opportunity to drive traffic with top-level content.
But at the same time, you have very little access via the SERP to this deeper sort of content. That’s what we saw with products last week. The access points, even on informational commerce queries, play a strong second fiddle to PLAs, product carousels, etc. Even on SERPs filled with answers, information, and top-level content the access points to deeper content is not there. How is Google supporting Search as a Journey if it’s not giving us access to it?
For example, if you pull up a sports team on the SERP the page is going to be filled with top-level direct “answerish” content. You will get options to extend your journey, your exploration. You get carousels of other sports teams whereupon clicking you get a SERP filled with more top-level direct “answerish” content. Do you know what you don’t get? Deep content!
Google’s mixed messaging is both confusing to the industry and holds back Search as a Journey! Instead of Search as a Journey opening up new vistas to searchers, it helps them pick up where they left off at times. It shows some related topics but it doesn’t really get in there like you would want it to.
Here’s a quick example. Recipes. What do you get when you search for ‘how to make chocolate cake?’ Maybe you’ll get a Featured Snippet or a video carousel. When Mordy bakes, everything goes wrong and it never turns out right. Wouldn’t a carousel of mistakes to avoid be awesome here? That would be perfect as it would be something the user wouldn’t necessarily think of on their own to search, but they would totally look into if it was shown to them.
But as it is now, you get nothing to really extend the journey. Search as a Journey is a great concept, but Google’s multifaceted goals are holding it back both directly and indirectly in so far as Google just doesn’t seem focused on bringing the right opportunities to the SERP.
Chasing and Tracking Google Updates and SERP Features: A Conversation with Dr. Pete Meyers [00:19:54 – 00:54:47]
Mordy: You are listening to another In Search SEO podcast interview. Today we have a magical man. He lives in the ethos that is all that is good, grand, and noble about SEO. He’s on the Mount Rushmore of SEO. He’s one of the most upstanding, uncompromising, irreplaceable people in the industry. He’s Moz’s own marketing scientist. He’s Dr. Pete Meyers.
Pete: Hey, nice to be here. I can’t add anything to that.
M: I always tell my guests that this is the pinnacle of the interview. It’s all downhill from here.
P: At least you got a good start.
M: So I sort of view you as like the Madonna of SEO because it’s just Dr. Pete. How did that come about? How did you end up being Dr. Pete?
P: You know, it’s funny because I was not so loud about my Ph.D. when I finished school, especially since I changed careers, but I think one of my clients realized I had a Ph.D. and for some reason, it was a big deal to them and they started to always call me Dr. Pete on the phone. When I went out on my own and started my consulting firm, my last name is Myers and it’s so common to misspell as it’s spelled like 500 different ways. So I thought I’ll use Dr. Pete and see how it goes and it was the best accidental branding decision I ever made. But now I don’t even hear it. New employees ask if they’re supposed to call me Dr. Pete and I tell them to please just call me Pete.
M: That’s awesome. What’s the Ph.D. in?
P: Human experimental psychology.
M: Well, that kind of relates.
P: Yeah. Truthfully, it was cognitive psychology. It’s just they didn’t call it that yet.
M: So let’s get into the wider world of tracking Google SERP and Google algorithms. But I have to ask you before we get started, so we’re competitors. What the hell are you doing here?
P: Yeah, it’s kind of awkward because before quarantine this was going to be a cage match.
M: Right, I have my ladder still outside.
P: I guess part of it is coming out of academia. I still have that research mindset a little bit. That’s not to say academia is not competitive. So some of it is that research background. I guess I care more about getting the right answer some days than the competition. But I also think, frankly, if you sum us all up, the big players, we’re about half a billion dollars while Google is 100 billion. Some days, we got to play nice with each other. I’ve always been impressed with the generosity of people in the industry and the data science. Even our biggest competitors now, like SEMRush, we talk with. There’s also a little bit of karma too. Sometimes I let somebody know what’s going on with Google and a month later I get something back.
M: Yeah, it’s such an interesting field where there’s so much depth, so much theory, and so much practicality going on that it’s almost stupid not to talk to people.
P: We’re also at a point in the industry where you can’t know everything. I mean, I’m sure there are some people who think they do. But you can’t keep up with it all and you can’t know everything. There’s always going to be some gap.
M: On this note of being open with each other, when there’s a big surge in rank fluctuations or SERP features, will you check out SEMRush or go to JP?
P: Absolutely. I just view it as confirmation sometimes, just so I’m on the right track. If I’ve got a big spike and nobody else does then something’s wrong.
It’s funny. One of the older tools launched two weeks before MozCast went public and it was like somebody else publishing your dissertation. And credit where it’s due, they didn’t rip it off or anything, they were just doing it in parallel. I think it was one of those ideas that it was in the right place at the right time. Frankly, we don’t put a lot of front end resources on MozCast so sometimes there’s just some bits and pieces other people have that are cool to see.
M: Yeah, SEMrush has a whole per niche breakdown. We were thinking about doing that too.
P: Yeah, you guys have a couple of page factors that we don’t look at.
M: I literally do the same thing. Recently, our Knowledge Panel data was off and all of a sudden there’s this giant drop in Knowledge Panel results. So I thought something is probably broken, let’s check MozCast. And to be honest, I think we were both broken.
P: Yeah, that happened.
Some people ask me about data sets and things like that. We bought STAT Search Analytics about a year ago. I work with Eric on the STAT team and they have a much different and larger data set because they work with enterprise folks. So we compare a lot internally too.
M: That’s interesting. Is it vastly different?
P: It is because MozCast was meant to be static forever which has pros and cons. The pro is I have six years of history that I know really well. It’s not changing every day, but it’s also pretty small. With STAT, we have a little better access to cross customer sets, all anonymized of course, but it’s much larger and enterprise-focused. Moz is a little more for mid-market customers. We also, quite frankly, have a data structure that is not built to look across all the data. It’s very compartmentalized by customers. It was designed that way for privacy and various things.
M: Yeah, we did the same thing. Same problem. We should buy a STAT. Where do you get one of those?
P: Find some nice Canadian and ask around.
M: Right, it’s funny because when I look at the different weather tools, I’ll think that they’re looking at a very similar data set that we are or they’re not looking at a very similar data set than we are. It’s interesting because I think people think that this is the data and it’s either right or it’s wrong, but it’s way more nuanced than that.
P: Yeah, there’s no representative sample of SERPs. You could type in anything. Even for Google, they say that 15% of searches are things that Google has never seen. We’ve tried to create that gold standard SERP set and you can do better and worse, but there’s no set. It’s not like a population you’re sampling where you know the characteristics. It’s way more complicated than people think.
M: That’s a really nice pivot into that piece you wrote on the winners and losers on the Google algorithm updates. In case some listeners missed this, you wrote this awesome piece about a winners and losers list. I once did a winners and losers list based on an algorithm update and I will never do it again. One, because I got a lot of flack about it and two because it was a really annoying pain in the ass. Because there are so many data spikes where you can catch a winner but it’s not really a winner, rather it’s a reverse from the day before. It’s one of those things where I feel that we take these winners and losers as if they’re the 10 Commandments. It’s kind of crazy. Do you really think that these are the most winning and losing sites across the entire internet?
P: It depends on the data set. We talk about that a lot. But it also depends on the timespan and volatility of that niche and volatility of the site. I don’t want to bash anybody because there are people who do good work.
M: Right, I don’t want to bash anybody, obviously.
P: Right, but to go in a different direction. When I first started analyzing algorithm updates, I would look at what keywords shifted the most. I even published early on who the big movers are, then I tried to figure out how they were connected, and then things got a little shady. Someone at Google must be just shaking their head.
M: Every time I write a post, I must stick in 400 caveats to please don’t take this as the gospel.
P: I’ll occasionally DM John Mueller asking, “Is this the dumbest thing you’ve ever read?”
I don’t think I realized when I started this tracking how volatile things were. Once I saw ‘Xbox 360’ really moved. And then I go look back 30 days and notice that it’s moving every day. You know, people always say that news sites move a lot. Yeah, the news is all over the place. The top stories change by like 40% over the course of a week.
I don’t think we realize how volatile and real-time the SERPs are. You have to factor that in. This was a great example with LinkedIn where if you look at the right day, like LinkedIn lost 100%. It’s so easy to overgeneralize.
M: Is there a benefit at all in your mind? I know Lilly Ray did 550 different winners and losers for different verticals which was a different take on having just five winners and five losers.
P: I think as long as we get better at it over time than it’s okay. I think we all recognize that there are limitations. I think the next step for us is to categorize these sites and look more in that sense. I did an analysis that was super complicated that I can’t publish trying to correlate the core updates because I think we have an intuition that the same sites keep getting hit or keep going up or down.
M: Yeah, like Healthline or Doctor Axe.
P: Yeah, so we spot those. I did a correlation between the absolute movement and if the same sites were moving all the time. The short answer is they don’t seem to be for the most part.
M: You just saved me hours of time.
P: The only real relationship I found is that the sites that moved in this core update tend to be related to the sites and moved in the most recent core update, which is not the most useful. So I think they’re more different than we think. But I think the trick is that Google’s not targeting individual sites so we have to look at what are the common factors of these sites, is it niche-based, or is it something they’re doing? That’s tough because that pulls in so much more data that we can’t see on the SERP.
M: Right. You have to actually go in there and use this thing called your brain which we don’t.
By the way, I have seen sites that for whatever reason, whether it’s an unconfirmed update or a core update, move every single time.
P: Yeah, there’s definitely a small subset that is being affected.
M: Why is that?
P: I think there’s a common thread to the core update. Truthfully, my hope was that if I could find a relationship between them, I could spot unannounced core updates. That was my dream
M: You can even predict who will get hit in the next quarter update.
P: One thing that really strikes me is that a while back, eBay took a big legitimate SERP hit and I wrote an article about it. I’ve worked with four hedge funds based on that one article.
We have the same problem with the press. I used to have a contact with the Wall Street Journal and I worked with the press a lot over the years. Truthfully, those winners and losers takes are hot. There’s money and press in that. I’m not saying they have ulterior motives or anything else. I think we’ve all kind of grown and evolved as we’ve gone but that’s a rush. That’s a hit when you publish that big news that this mega-site got hit. You can get sucked into that a bit.
There are times when marketing asks, “Hey, are you going to publish about this Google update?” And I tell them, “I don’t have anything to say.” I’ve dug into the data and what I’ve learned is that I don’t know anything. It’s not a very long blog post.
M: I don’t think people realize that. It’s very hit or miss. You can either find something or you can totally not find anything.
P: Yeah, and there’s a little luck in there too.
M: Totally. And most of the time I find something it’s not because I set out to find that. I set out to find something else and I went down this whole rabbit hole.
P: Don’t give away the secret sauce.
M: The secret sauce is that there’s no secret sauce.
We track all the SERP data and all this algorithm data and, at least for myself, I don’t focus day-to-day on practical SEO. It’s a weird space to be in. Because you’re dealing with SEOs who are very practically minded. But you’re looking at, at least for myself, tons of SERP data, algorithm data, update data, etc. And obviously, I’m looking at our SEO at the same time. But I find at least for me that it offers me a very different perspective. I end up giving SERP theory, SERP philosophy, or SERP abstraction, as opposed to giving five tips on how to do this. So how does it impact you to be looking at all that kind of data?
P: Yeah, it is weird. It is almost like academic SEO in a sense. Certainly, there are people who criticize us that they’re doing the work and we’re chasing the algorithm. We’ve heard it all.
I have asked over the years if this is useful. Why do I keep doing this? Originally, I was really frustrated. I think it was that testimony before Congress that Eric Schmidt did in 2011 or 2012 when he was CEO where under oath he said there were 600 or so updates a year. We have all these people, like Barry Schwartz, who are trying to name the updates and you have Eric mentioning 600 while we only named 13. And now it’s closer to 3000 updates a year. And I know that might mean some very small niche features.
So that’s what got me started. I was frustrated with this huge gap where we would just sit around waiting for Google to tell us when an update was coming. It just seemed so counter to our culture as SEOs. But I think over the years, the thing that people said to me is that it really helps to know if it was just me or was it the whole world. Was it just me or was it Google? At least then I know where to start. That’s the thing that’s kept me on it.
It’s funny as a content marketer I have the problem of making this Google update actionable. What does this actually mean? I’m always sending it to the editors with no conclusion.
M: I have the same problem with finding the takeaways.
That’s awesome. The same thing with SERP features. Who the hell cares? Playing devil’s advocate here, who cares how many sitelinks are on page one of the SERP?
P: I have an easy answer to that one. I think if you look at the early days of SERP features we just saw them as add-ons. We got the SERP and it can have up to seven ads, it can have video thumbnails, it can have news, it can have images, and they were all these building blocks that Google could put on the SERP if there was relevant content. Then over time, there was a clear shift where SERP features became fundamentally about intent and it’s not just something where every SERP can have this.
Featured Snippets are a great example. You don’t generally see those on certain commercial SERPs. They tend to be informational (but not always). They tend to be a certain kind of query. And even with ads now, when Google went from three ads on top to four ads on top, there was an immediate shift in the SERPs that got four ads. They weren’t the ones where people were just bidding enough to justify having four ads. They were the most commercial SERPs that I think Google knew were converting. And now we see it with this whole SERP funnel where you’ll see related research carousels and there are no ads, yet when you click down a layer then there’s the ad. And Google is working on this refinement funnel.
So I think the thing that’s so important now is that those features tell you about intent and tell you what you should be targeting and tell you that maybe you should stay the hell away from that SERP. Maybe you shouldn’t be trying to do organic for wedding dresses where there’s 17 paid positions on the top, a Local Pack, and three other things before you get to organic and think differently about how you target. So that’s the fascinating thing to me. I don’t think these are just stats on checking if it went up or went down. I don’t think that’s that important. But if you’re not doing video on a SERP with a video carousel and a giant video at the top, what’s the point? Go somewhere else. You either start doing video or find a greener pasture for what you do best.
M: I saw one today, for a query about COVID-19 and the top of the SERP was a News carousel, a Video carousel, and a People Also Ask box. It’s unbelievable.
P: I think even Google knows it. Why are they not showing ads? Because there was no money there. Nobody was clicking on those ads. They were lousy ads. They realized we need to get people to the point in the funnel where they want to buy. We have to mimic that.
Right now, COVID-19 is a great example. This is a great time to shift to more mid-funnel to top-funnel content. Maybe you’re not actively selling right now. Maybe you got all sorts of challenges and you’re trying to stay afloat. Get that informational content out there, get your brand awareness, and get that mid-funnel stuff so that when you’re ready to open back up the bottom of the funnel, you can make that move. But those are different SERPs. That’s not just organic versus ads that we used to see. Right. Those are completely different SERPs that require a completely different kind of targeting. We have to be a lot more strategic and get away from that whole vanity keyword viewpoint.
M: It’s funny because when you look at Google, the biggest thing for me how they use their SERP features is such a clue into how they’re thinking about things and I don’t think we spend enough time looking at that.
P: Yeah, it’s very segmented. And when you start to see that there are some exceptions like People Also Ask is in 90% of searches.
M: Yeah, but even that you can see what kind of questions they’re asking.
P: Yeah, it’s an interesting feature to mine. But a lot of times you look at certain features they top at 15-20% and you wonder why they don’t keep going up. People ask me, for example, if everything is going to go to voice and the answer is no because it can’t.
I think what’s interesting is thinking about targeting differently and thinking about how different kinds of content are better for different kinds of SERPs. It’s not all one big mess and we don’t like it because it’s hard. It takes a lot more work to do that. But I just think there’s so much where Google is just trying to say that this stuff is effective and this other stuff isn’t. And if we’re still trying to do all that other stuff, or just hit everything with a hammer, because that’s the only tool we have, we’re just not getting the results we used to get. And we see that we see traffic falling in some cases and CTR falling where we don’t feel like we did anything wrong. We didn’t lose ranking and we didn’t get a penalty, but we’re not aligned with that intent and we’re beating searchers over the head.
M: It’s kind of the point where, not to criticize SEO tools, but there’s really no substitute for looking at the actual SERP and seeing what’s there. You can see there’s a Featured Snippet there, but what’s in that Featured Snippet? What kind of Featured Snippet is it? What was Google trying to do with that Featured Snippet, and so forth?
P: I think the challenge for us is that it’s hard to look at every SERP individually and we do want that aggregate data, but it’s hard to aggregate intent. We’ve been working on intent tools. A lot of us I’m sure have been trying to launch and it’s tough because the UI aspect of how we present that to people is probably harder than the modeling of the content itself.
Optimize It or Disavow It
M: If you had to check just one thing, either SERP feature trends or top-level rank fluctuations. Meaning, you wouldn’t know anything about the niche or you wouldn’t know anything about the sites being impacted, just top-level weather rank fluctuations. Which one would you track?
P: SERP features. No question.
M: Explain, please.
P: I think SERP features have a lot more that’s actionable. That’s why I want that shift. The only thing that top-level ranking tells me is if it was just me or am I crazy. It’s useful to a point but if we don’t break that down, it’s not actionable for most people. As we’ve been talking about, SERP features show us intent. They show us how to segment keywords, they show us how to think differently, and they show us how to target content and what to create. At the end of the day, I just think that it’s much more actionable.
M: Dr. Pete, thank you so much for coming on.
P: Thank you. Good to be here.
SEO News [00:55:22 – 00:59:21]
Google Rich Results Test Tool Out of Beta: Big news, Google has taken its Rich Results Test tool out of beta saying all of the supported markups are now supported by the tool. With this, Google will shutter the Structured Data Testing Tool at some point in the future.
Mobile Image Results Get Knowledge Cards: Mobile image results may now get Knowledge Cards that appear in the preview screen. The cards expand and offer snippets of info on the topic represented by the image.
Google Ads App Now Supports Manager Accounts: The Google Ads app now allows manager accounts so that you can overview data across all accounts.
Google Ads to Ban Spyware Surveillance: Google Ads is banning the advertising of anything that provides surveillance without authorization. The change in policy goes into effect on August 11th.
Tune in next Tuesday for a new episode of The In Search SEO Podcast.