Episode 60: Google May 2020 Core Update and SEO myths

Play this episode:

Or get it on:

• Apple iTunes
• Anchor
• Breaker
• Google Podcasts
• Overcast
• Pocket Casts
• RadioPublic
• Spotify
• Sticher

What's in this episode?

In this episode, Mark Williams-Cook will talk about why talking about the Google May 2020 Core update at this point is a waste of time and will go through the most popular user submitted 'SEO myths' including content length, bounce rate, and links not being important!

Show notes

https://twitter.com/searchliaison/status/1257376879172038656

https://twitter.com/crestodina/status/1256223105938264069

Transcription MC: Welcome to episode 60 of the Search with Candour podcast! Recorded on Thursday the 7th of May 2020. My name is Mark Williams-Cook and today we're going to be talking, very briefly, about the Google May 2020 core update, and also we'll be talking about your favourite five SEO myths that are still alive in 2020.

It’s core update time Google, on the 4th of May, so that was this last Monday gone - on Star Wars day - announced they're incredibly creatively named core updates, which they are calling the May 2020, core update - much like the core update we had back in January which I believe they also called the January 2020 core update, so continuing that very exciting naming system. So in very similar fashion, copy and paste fashion to last time the Google search liaison Twitter account tweeted saying, later today we are releasing a broad core algorithm update as we do several times per year, it is called the May 2020 core update. Our guidance about such updates remains as we've covered before. Please see this blog post for more about that and we will link you to the Google webmasters blog about Google core updates as we have done before as well.’

Again the advice hasn't really changed much, this summary is just if you see your site affected negatively by this it's not that your site specifically has had actions taken against it, it's just that the algorithm has now decided that for those queries other sites are of a better fit, so their general advice is normally there isn’t anything specific you can just go and easily fix on your site, it's not the same as getting a penalty. We've gone through that before a few times. I wanted to mention it just at the top of the podcast as I just to make sure everyone was aware that there was a core update, I'm aware not everyone is obviously at work, some of you may be furloughed at the moment and not keeping up to date, dare I say with the SEO news, I mean who wouldn't do that? And so if you do see any big changes in ranking to your sites, it could well be because of this core update. Again as usual with these core updates, they do not just happen overnight, they normally take days or actually weeks to roll out and see the effects and impacts of that, so they're happening as Google's recrawling the web and applying whatever changes they've decided to make. And for that reason, yet because Google hasn't provided any specific information saying, we're looking at these factors or this this algorithm update is going to target these sectors, nobody really knows anything, so it's probably not worth your time reading the ultimate guide to the May core update 2020 just yet, you can hold off maybe a couple of weeks before you start reading the guides that contain everything you need to know about that.

Looking at the data we do have, so tracking tools like SISTRIX, like SEMRush do monitor what they call SERP volatility, so this is just looking at all of the keywords they track across their databases and they give an index of a zero to ten for how much they are shifting. The SEMRush SERP volatility for the last 30 days did actually spike on the 5th and 6th of May so far. so they were averaging below 5 over the last 30 days on their volatility index and it kind of peaks and troughs between 2 and 5, which they kind of classes green in there traffic light system and then on the 5th of May, this jumped up to kind of a seven point nine and eight and then on the 6th of May it was nine point four, so they were showing that across all sectors they were seeing quite a lot of movement up and down - it does look like it calmed down the following day on May the 7th, as I said, it's too early to draw any kind of conclusions yet. If you've been doing quite quote the right thing for SEO then hopefully you shouldn't have anything to worry about and if anything does happen, I would hope that you fall on the positive side of that but again, we will keep our ear to the ground, we'll talk to people, other SEOs and see what results people are seeing and where relevant we'll give you an update in the next few episodes as well.

About a week ago, I put a message out just asking the SEO community what were some of their favourite SEO myths that are still going around in 2020. I thought it'd be fun just to go through some of these and just talk through them. I still see, well looking at this list that I got from everyone, it's nothing new in 2020 - it's the same SEO myths that have been persisting for years, the same ones I saw in 2015, the same ones I saw in 2010, some are even more applicable I think nowadays. But it's good to always discuss these things and talk about them in a bit of depth because I think everyone at some point has been a bit of a victim to algorithm chasing; where you get a little bit too focused on specific ranking factors and lose sight of that bigger picture and actually what search engines are trying to achieve. So I've taken five of the SEO myths that got submitted, I had about twenty or so, and I've taken five that either came up multiple times because quite a few of you did submit the same thing a few times over, or I've taken ones that I think are particularly important.

So in at number one is content length and someone actually submitted a tweet to reference what they mean by this which says, ‘the average blog post in 2019 was 1236 words, that's 53% longer than six years ago. Today, longer content performs better and gets better results. this and other great tips for writing high-performing content from friends at da-da-da’ - I won't go into who that was and they've put a little chart at the bottom, that's showing the amount of words on a page and the performance of that page. And this is the myth that we hear a lot which is basically search engines want long content, the longer the content the better and there's some very artificial guidelines it seems going around to, you know a blog post must be 500 words or if you want it to rank really well it needs to be 2,000 words and we've seen these studies get published time and time again.

Now the studies that are published are obviously showing a correlation in some cases. Actually an interesting exception to that was the study that Dan Taylor and I talked about in episode 59, the last episode, when we discussed the backlinko, I think it was 11.8 million search results study, because one of the conclusions that was drawn from that was that content length didn't actually seem to have too much of an impact, but usually these correlation studies do show - hey look, we've seen that these number one rankings tend to have more words than number 2, 3, 4, 5, etc and sometimes there's a nice correlation there. But as we all know and we're probably all sick of hearing, correlation doesn't necessarily mean causation. And it doesn't make sense to me either, it doesn't even make sense that that would be a thing - why would Google want content that is more wordy versus content that isn't wordy, so I'll talk through that just to give you my thoughts on that.

So one of the things we know Google optimises for as a thing is time to result. so they've spoken about this a lot before, it's one of their kind of internal metrics for judging how their search engine is performing - ie we have a user do a query, how long does it take for them to find the piece of information or do the thing that they want to do - the time to result, so that's important. And we know Google optimises everything around intent as well - so this is why for you know over ten years now Google's been on this mission, where they've made this statement, you know there's said a few ways, they've talked about things not strings, they've talked about becoming an answer engine as opposed to a search engine, and a lot of that is based around this time to result and intent. So really basic example, if you Google what the weather is, the top results aren't weather sites anymore, you just get the featured result that just tells you the answer to the thing you're looking for and of course, we're all aware we've got you know universal search results now that bring in news stories, images, video or we've got features now in the SERPs, like you know recipes and featured snippets, that are all aimed at reducing this time to result for users. We know that Google does run tests on these different SERP features, to find out what they think users prefer - so with all of that based around intent and trying to get the result as quickly as possible to someone, it doesn't make sense to me that Google would want huge wordy bits of content for, at least if we think about specific search terms.

So if the intent of a search is clear, you know I want this one piece of information, why would they prefer to rank something that's 5000 words long? And I've seen some really great examples of this over the years, one of my favourites, which has now been replaced I think by one of Google's own panels, was queries around when the clocks change and when the clocks go back and the example was that, the top result was basically a website that pretty much just said in massive letters the clocks go back on this date and ranked number two - that was an article with hundreds of links, that had a video about people sleep cycles, had a history of you know why the clocks changed, it had images of people sleeping that were marked up about that, it had a poll on it to get people kind of clicking and sharing, and it didn't rank as well as this article, which you know looked like it perfectly matched the intent which is I'm after a date here's the date. And even if you look at really competitive SERPs, so if you look at something in the UK at least if I look at something like, play live blackjack, and that search term is one of the more expensive Google ads terms, so you're looking at around about 120 pounds per click for a term like play live blackjack. So ranking for this term, even with relatively small traffic, is very much worthwhile and when you look at the number one website for this, it's got around about 250 words in their main content and that's again because the intent of someone searching for a ‘play live blackjack’ is fairly straightforward, that's a functional thing that they want to do. It doesn't make sense that Google would prefer to rank a site that's got 2,000 words about the history of playing live blackjack, it's about matching the intent of what that user wants to do. and I think when we get these concrete examples, like look there's a really competitive search term here and look, it doesn't have that many words, it makes sense.

However for terms that have broader intent, then it does make sense naturally, even if we don't think about search engines to have longer content. So I'll give you two examples, if we had a search say for car maintenance, I would expect that to be a long article because the term car maintenance itself is not clear in its intent, so there are lots of things that make up car maintenance - you know like checking tire pressure, changing oil, de-icing your car and maybe all of those things themselves might be shorter articles, but the the intent of the term ‘car maintenance’ is quite broad and this is why you see people implementing these, what they call, ‘hub-and-spoke’ is one way or ‘pillar content’ strategies, where they identify these broader key phrases and they write these really long, very good, high-quality, detailed guides because it's that length of content that's answering those broader queries and then within those articles and bits of content, they'll usually deep link off to the specific shorter articles, which can then then rank themselves. So putting all that together, I think it's fair to say that on these correlation studies; firstly, you're more likely in in these correlation studies to see results from these ahead and mid tail terms, not so much long tail terms, which will naturally be broader intent, so naturally you are going to see longer content rank in these situations. And you know, we're not seeing them rank because Google is saying, oh it's 3,000 words it's likely to be better content just because of the word length, it's because the people have gone to the effort to answer those queries are literally just producing better, more helpful content, which in turn as well and is the advantage of though that kind of pillar content strategies, they'll be gaining links doing that because they're genuinely helpful guides. So I think it is very misleading to say that just generally long content performs better, because especially in the case of these more specific intent terms, it doesn't make sense and to back this logic up, it's something that Google themselves have been very clear about.

Now, we of course always need to be careful or critical or cynical about how we interpret things that Google say, and I'm not saying I think they outright lie when they tell us specific bits of information, but there have been cases where they haven't necessarily been as clear as they might have been or mistakes have been made. But content length is something they have been crystal clear on and they've been very specific and say that we don't use word count as part of our decision for where we're ranking these pages. Again, there are side benefits of if you've got a long piece of content, you've got more content, so there's more potential to rank there for different longtail terms, but when we're talking about individual rankings for these broad intent terms or any term, I don't think it's its right to say that they're just looking at this pure word count.

Number two myth was, Google uses bounce rate for ranking and off the bat I would say again, this is something that Google has been super clear about and super specific and they say look, we do not use bounce rate/ Google Analytics data for our ranking algorithm and they've said this multiple times, from many different Googlers. And I think there's just actually a lot of confusion about what people mean when they say bounce rates, because you know bounce rate as hopefully we all know, has nothing to do with engagement in terms of how long a user is spending on a page. To give you the Google Analytics definition they say ‘a bounce is a single page session on your site. In analytics a bounce is calculated specifically as a session that triggers only a single request to the analytics server, such as when a user opens a single page on your site and then exits without triggering any other requests to the analytics server during that session, so that's our definition of bounce.’ And we can, with that, challenge the assumption that goes along with bounce rates which is that, a high bounce rate equals bad, because again I don't think that's necessarily the case.

So if we take this example of going back to our car maintenance example, we've written an article about how to change a car battery and someone's Googled, how to change a car battery and we've ranked number two, they've clicked on our result and the page has loaded and they've spent ten or fifteen minutes reading that article, thought okay this is brilliant, I'm perfectly happy doing this now and then they leave the page, and then twenty other people have the same experience they come, think this is amazing, brilliant content, read through it and then they leave. So this then has got us pretty much a hundred percent bounce rate. The same if you're Googling for a phone number or an address or some contact details from a company and Google does its job and sends you to the correct page, you go there, you jot down the information you need, and you leave again - that's a hundred percent or very high, if you've got other visitors, bounce rate. But those experiences those people have had are what you would say, a perfect user experience. They've searched for a piece of information, they've landed precisely where they want to land, so actually in terms of engagement, clicking around, not bouncing, it means they haven't had to navigate anywhere else to find any further information that they've needed, everything they needed has been on the page that the website and served that Google's found for them.

So I think it's very difficult to build a general rule for a search engine, if they were to use bounce rate to say when is contextually a high bounce rate bad and when is a high bounce rate good, so it doesn't make much sense as a metric for them to use. And of course Google to do this would need access to Google Analytics and we have to be aware, although it's very popular, not all sites use Google Analytics, so they would need a way to reliably measure the bounce rate on all websites if they want to use that as a metric. Of course with our definition of what a bounce rate or bounce is, it's super easy to gain bounce rate. We can make it so any visitor on our page within Google Analytics finds off an event, which will bring our bounce rate down to, if not very close to down to zero, so it really doesn't make much sense again as a usable metric.

And where some of the other discussions have happened, have actually been more around user behavior on the SERPs, so we've seen people talking about things like dwell time and pogo-sticking, so how long users are spending on the page and then going back to the search results maybe and then looking at other search results and Google have been a little bit less clear about this. So when people have mentioned dwell time, Google's replied sometimes on Twitter saying things like, what's dwell time?

Pogo-sticking is interesting when you think about it on the surface, yeah that might be really helpful so someone does a query, they click on a site but then they come back to the search results and then they look at page, the second or third result and then they stick there, so maybe if people are frequently bouncing off the first result and by bounce, I mean going back to the search page not bouncing off the site, so they're going back to the search page and then looking at other search results, maybe that's something they could look at. And again, there's a few problems with this, if you think about the whole user journey, and how Google can reliably use this as a metric. So if we're looking to buy a new TV and we load up a review website, it’s likely that we're not going to just read one review, so if we search for some Samsung TV model and we read the number one review and then we're like that was great, I'm gonna go back and read the second one and maybe the third one and say on average people read two and a half reviews, that would mean that this behaviour of them clicking on the first result, coming back to the SERP and then looking at the second or third result, is really common. Now could you conclude from this then that the reviews ranked number two and number three are better than the review at number one? You can't necessarily say that's the fact, because people might just want to read more than one review, so they're naturally going to read what's at the top first and finish later on. There's been a couple of Google talks, with Google engineers, when they've in context been talking about when they do SERP experiments and how hard it is for them to use click data and user behaviour to draw solid conclusions and build general rules. The other thing that's worth mentioning is, so Bill Slawski from SEO by the sea, who spends a lot of time, really helpful reading through Google patents and posting his thoughts on them. I saw him tweeting about pogo-sticking, basically saying that in all of the paints he's reviewed, granted and pending, he's never once seen any mention of pogo-sticking and therefore his thoughts were, if this is a important ranking thing for Google then surely it should be in one of these patents. So I think that one's a little less quite clear, I think we've seen various different results come through like click-through rates and stuff from SERPs, but if we loop back around to the myth which was about bounce rate, I'm fully behind this which is that I'm confident that Google isn't using bounce rate from Google Analytics in their rankings.

In at number three was, technical SEO is no longer important, and you might think that I've heard this the most from people who funnily enough specialise in content or link building and I mean I guess if you’re hammer, everything's a nail and you would have thought that technical SEO would get less important over the years because the technology from the search engines is improving and it certainly has you know, because now Google is rendering pages more, how chrome looks at them, so it's rendering CSS, JavaScript and it never used to do this before. But the actual web tech is still outpacing Google's resource to look at it in a similar way to how users are seeing it. So if anything we're doing, we're certainly as an agency doing way more tech SEO now than we were five years ago.

So you know we've talked about them before but the JavaScript frameworks, Vue Angular, Vue React are brilliant frameworks, but they are causing more issues around client-side JavaScript and this is because this rendering of JavaScript, as a specific issue for Google, is tricky because it's very resource intensive and they have finite resource and again, we've spoken about that before at length, about this delay between Google initially crawling pages, looking at the HTML, and when it gets queued up to be fully rendered and from a technical SEO point of view, we've had to put in whole solutions to deal with this. You know, server-side rendering is getting a lot more common now and these solutions really only exist for robots. Whether they are from social networks or from search engines, we're branching out and building these extra systems that not only we have to make but we have to continue managing them for robots.

So doing server-side rendering and at the same time serving a site that's client-side rendered for your normal visitors is a lot more work, because you need to make sure that both are working and both are in sync and caching on one isn't broken, so that's quite a in-depth task. And like we mentioned before, the universal search results are always changing, we’re getting more SERPs features, not less and things like schema are becoming more important. So again, I think it was the Backlinko study that said schemas not important and there's been a few things talking about how schema doesn't impact rankings, which directly again, I think it's fair to say it does help Google understand entity relationships for what is this thing, what is this thing, and how do they relate each other, but the very fact is, if you want to appear in a recipe card on Google, you have to have the schema or you're not going to be eligible in the first place for it. And again, these are all extra technical things that only exist now that have to be built and the cost of getting some of these technical things wrong is very high. You can kill the site's performance, its ability to rank, no matter what you do with your content and with your outreach, if you get this technical stuff wrong. So again I think even though we've got big, popular platforms like Shopify, like WordPress, like we spoke about in the last few episodes, it would be foolish to say that technical SEO is no longer important.

I thought as number four it would only be fair to get the other view which is that, as long as you write great content, people will find you and there's no need to build any links. So that's like the opposite end of the opinion spectrum to tech SEO isn't important, and for years and years and years, Google has been playing this line of content is king, just make the best possible content and good stuff will happen. It used to be that, in my opinion, that line was just laughable because SEO was primarily about links and enough links would fix just about anything and rank just about anything. Of course, with the updates we've had with Panda, Penguin etc. over the years, that's not the case anymore, you definitely need good content but links certainly still are important or actually just even some kind of outreach, even if the goal isn't this laser-focused ‘we need this many links, that are followed with this anchor text, from these domains’ - so whether it's social or more digital PR, which is certainly popular at the moment, is that these content signals on their own aren't strong enough.

If you go and register a new domain, you build a brand new website and you write the best article about car maintenance and don't tell anyone about it, while it may get indexed, it's highly unlikely that that content is just going to rank itself. There are a few sites, obviously big sites, that have the luxury of being powerful enough that when they publish new content it just starts to rank, but when you followed this trail of breadcrumbs, that's usually because those sites themselves have so many links that they're passing around this, whatever you want to call it, authority trust and link equity, whatever you like, they're passing that through to those articles through internal linking, that's what's causing them to rank.

It's very clear to test that - if you publish the same or very similar bits of content on an established site versus a new site, the new site isn't just going to start ranking. So if you're not in that position with a known brand, with a known following, you're going to need some outreach and ideally links to help you rank that content. So again, to say links aren’t important or links are dead, it's just ridiculous to be quite honest. And you know, when you see people saying, oh I made this content, I got it to rank without links - there are, of course, always links to that content because if they weren't links to it, people wouldn't be able to find it and normally those links are internal. So again, this new bit of content they've written that quote-unquote has no links, does have the benefit of links because the rest of the site has links.

Lastly, number 5, I wanted to finish on ‘spending money on Google Ads improves rankings’. It's kind of an interesting one actually. So off the bat, of course spending money on Google Ads does not directly affect your organic results. There's no, I don't think it's come up as a myth, so it's obviously still talked about but I think it's pretty clear for most people in SEO that's not the case. It is interesting though and I wanted to talk about it because there's some interesting side effects that have been quite well documented which is, if you're running Google ads and you're appearing as an ad for things you also rank for organically, is that you tend to then get more clicks than usual on your organic result, because it's double listing and that's likely just down to some kind of psychological effect of that double listing. But I think it's quite interesting. And you know even if you don't rank for something, if you are running Google ads and appearing for keywords, it's going have other halo benefits of, you know hopefully you're going to get more brand searches, you're going to get more of your brand searches related to the products or services, whatever it is that you're offering, and you're starting to build that digital footprint, as to your brand, your identity or entity within Google.

And of course, running ads, getting clicks, means more eyeballs on your content, more eyeballs means more exposure, more chances to be shared, more chances to get linked to, all of which are things that are going to contribute to your ranking. So again, it's a case of being careful with correlation, causation - Google have been very clear about this as well and again if we think, outside of just rankings and logically what would benefit them. I think, not my expert area in law, but there would probably be, to put it kindly, a legal gray area around not disclosing you're manipulating organic search results, based on people paying for ads as well and I don't even see what the benefit to Google would be to take such a massive risk, you know even if it's not legal, even with a PR perspective, why would they would want to risk that and how in any way helping people rank organically would positively impact their Google ad spend. If anything, I would see it reducing, so it really doesn't make any sense to me as a theory. Google have been very clear, lots of times, that is definitely not a thing but again there is this halo effect of, yeah if you run ads to your content, you're more likely to get links which is great - more eyeballs, more links, better rankings. The same is true when people produce new content, you might pay to boost it on social because just getting people to it is, again, going to help it get out there and help get links.

So I hope we've cleared up some of those SEO myths and you've enjoyed listening to them. We are going to be back on Monday the 18th of May. As usual you can get all of the show notes at search.withcandour.co.uk and I hope you all take care and we'll see you then.