Episode 77: Search term keyword data reduced, price hikes and Q&A

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, you will hear Mark Williams-Cook talking about the Search terms report data being slashed, Google Ads price hides and also a listener Q&A: Your SEO questions, on Javascript, indexing, and internal linking answered!

Show notes

Collin Slattery tweet https://twitter.com/CJSlattery/status/1301252078531235840

Links: Gov Digital ServicesTax: https://www.gov.uk/government/publications/introduction-of-the-digital-services-tax/digital-services-tax

Transcription

MC: Welcome to episode 77 of the Search with Candour podcast, recorded on Friday the 4th of September 2020. My name is Mark Williams-Cook and today we're going to be answering some of our listeners questions about SEO and we're gonna have a bit of a focus on PPC today. There’s been loads of different news stories popping up, but two things I want to talk about is, firstly, Google ads slashing the amount of keyword data available and, of course, raising the cost for advertisers in various countries, such as the UK, where everything is now two percent more expensive.

Before we get going, this episode is sponsored, very kindly, by Sitebulb which is a desktop-based SEO auditing tool for Windows and Mac. It focuses on making your SEO issues actionable and easy to prioritise. I've talked about it before, if you're learning SEO, it also does a great job of teaching you about those issues. Sitebulb has an exclusive offer for Search with Candour listeners, which is a 60-day trial, no credit card required, you can get it at sitebulb.com/swc

They've actually just released version 4.1.0, I was going through the release notes now - that may sound like a sad thing to do, but if you know about Sitebulb or release notes, you will know they are worth reading. They're pretty entertaining as far as release notes go. I'm really pleased to see and I've spoken before about how you can audit schema on your website with Sitebulb and that's been really warmly received by the SEO community. There's a great quote actually in the release notes that says, this is from a customer, ‘I've loved ScreamingFrog for years and never thought I'd need to look for another crawler until Sitebulb brought out their structured data update. Brilliant and really easy to identify the issues across the site. I feel like I'm cheating, but damn this is just so, so sexy.’ So, I mean it's going far if people are describing anything to do with auditing schema as sexy I guess, but it does make it really easy. The update includes even more improvements about how they do that. So they're merging node identifiers on schema, meaning they can actually detect, rather than say if there's some organisation schema, rather than saying there's a thousand bits of organisation schema on the site, they can identify maybe if that's one bit of schema that's actually just had multiple instances. So it made it much easier to get that overview with merging. There's updates as well with their content search, so you can now search your site for specific words or phrases in bulk. Absolutely brilliant update. As I said, a 60 day trial at sitebulb.com/swc - go and give it a try.

We usually end the show with this Q&A but I thought I'd bring it to the front because we've had some great questions. We've got a bit of a backlog of questions we're working through; if you have a particular SEO or PPC question you would like answered, you can tweet it to us @candouragency on Twitter, or if you can find me on LinkedIn, I think I'm the only Mark Williams-Cook on there, but if you can find me on LinkedIn and you direct message it to me, I'd be more than happy to answer it on the show. The only thing we try and say is, try and make it non-site specific, so it's a question that lots of people would benefit from. I've got a few of those questions that you kindly submitted that I'm going to try and answer. So the score is basically you send in the questions, I very quickly scan read them and just work out if they're not too difficult for me to answer, and then I'll try and do that.

So the first question we have is from Uppliappan Eswar who says, if we don't have access to a website backend, would you recommend using Tag Manager to overwrite canonicals and push href lang tags? Is there a downside to this?

So I guess firstly, thank you for the question and I would say the best practice if you can do it of course is to just code in your canonicals, any schema, hreflang tags, directly into the html if you can. Of course as you've pointed out, this isn't always the case where you have the ability to do that certainly we work with some “legacy systems” where we can't even update page titles, so if we break it down there's certainly not much for an issue in adding non-existing data via google Tag Manager. So for instance, if your site doesn't have href lang tags and you need them and there's no real feasible way you can actually just add them to the site, without using an external tool like tag manager, then absolutely great, definitely go ahead and do that. The downside is that yes of course, apart from the site being crawled in the html, it does need to go through caffeine - Google's Renderer - to execute that JavaScript before anything additional you add in will be seen but better late than never, so definitely in that case.

Now, for data that already exists, the answer I guess is really looking at what you're trying to change. So in my own experience, we have done things like overwritten title tags with tag manager, I've even changed things like footer links in tag manager where I was told it essentially wasn't possible to change those. All of those changes, as I said, took sometimes a few days to get picked up.

So Google's rendered the site again, but they did get picked up and they had the desired effect. For instance, the page title updates did start ranking for what we wanted them to. So if it's that kind of thing then again, I think it's better to do that than not have it at all. One thing I haven't changed on doing, and you're quite specific in your question here is you say overwrite canonicals. So canonical tags as we know are a hint to Google about which URL we're recommending is the canonical tag but Google doesn't treat that as a directive; so it won't put all of its decision into what we say. Overwriting canonicals could potentially be tricky because you're saying to Google at one time okay, this is the canonical version over here, and then when it picks up your tag manager stuff, you're changing that decision. So as long as that's the same over a long time, you should be fine. If they aren't there at all, again absolutely add them with tag manager, if they're already being created, I think you just need to be careful about what Google's going to do when you start changing those signals, especially on mass. So I'd maybe do that as a test first, especially if you've got a larger site, maybe pick a group of pages, change the canonical and you can, of course, go into Google Search Console and actually check which page Google has selected itself as the canonical tag and see if that matches up, so I hope that answers your question.

The second question we have is from Lewis Camp, who is a freelance content writer and Lewis asks, why do websites appear on Google for a specific search term and then disappear moments later? Is it down to the domain not being very old?

So off the bat, no it's not to do with the age of the domain, that's not why this is happening. There's a few possible reasons for this, depending on exactly what you're describing. So when you say that the website's disappearing on a specific search term, moments later, that could mean it's ranking in a different place to a previous search you've done. So you haven't mentioned in here, but there are factors such as the device you're using, or whether you're signed into Google or not, there's lots of different factors that can affect what search results you see. So Google may change the order of search results, for instance, if you search on desktop or if you search on a mobile, because that may reflect a different intent. So it may mean that you're on page one for one and page two for another, so it looks like it's disappeared. If that's not what you mean and you're saying it's completely gone, which I'm guessing that's what you mean, this could actually be related to the age of that page, so not the domain, how long that page has been in existence in context to when Google discovered it. So there isn't just one big index that everyone is dipping into when they're doing a Google search, there's a huge Google infrastructure, with lots of different databases that are all trying to sync up with each other. So especially when a page is brand new, it may get crawled, it may get indexed, but that page isn't necessarily present within all of the indexes across Google's infrastructure.

So I've certainly seen that before where when I first publish an article, it can get into Google and it's there and someone else does a search in the same room, and they get slightly different search results where it's not there. And I've actually demonstrated a few times with new pages on experiments where we can get a new page indexed, but that URL won't appear when we use a site operator, but it will appear in the cache, but it won't appear if you search for the URL, but it will appear for a search term, so we just have to bear in mind that Google is really, really big and it sometimes takes time for these things to sync up. If you've got a site that's got a decent amount of links, and it's you know fairly popular, and that stuff does get indexed, within maximum a day or so, I would expect it to be consistently there. If it's not consistently there, I would then start to actually look at your own site to see if there's an issue. I hope that helps.

Our final question is from Manish Bhickta who asks, how important is it to place important pages in menu navigation to make a better website architecture?

I think this is a really interesting question because this is one where SEO does run in parallel with the user experience. I would just answer that with a statement which is that, I would say, it's Important to prominently link to pages that are important. Just from your user's point of view, so imagine you have spent 15 hours creating one bit of content on your site, so you've really researched it, you've got experts involved, you've made assets and graphics for it, and it's the absolute best bit of content on that subject, and you want it to rank for whatever search term you've selected for that. It's really like a pillar bit of content for your site. It doesn't make sense even if we don't think about search engines to hide the link away for that page, so to link to it just from one category page on your blog. So within 12 months time, the only way someone will be able to find that article is if they go into your blog, and then look at the archives, and click back to September 2020 or whatever the date it was and then find it, that just doesn't make sense. If that's the case, it does make sense for search engines to come to the conclusion well, hey this page can't be that important because it's buried away here in the site, and they're not that worried if people find it or not.

The opposite is true for items or URLs that are in a main menu. It's a fair assumption that if a link is available on every page on the website, that is going to be an important page and we know that internal links and any links that exist as well are not created equally, or they are created equally, but they don't end up equal, I guess.

So normally a link, for instance, from a home page is worth more just because the home page has more links to it normally, than other parts of the site. So if you had that bit of content that was important, you could create a panel on the home page for it, or yes maybe, if it's a guide to using your products or some information you really want people to see, it might make sense to make it a permanent addition to the menu navigation.

So to circle back around to your question, it is important to place important pages in the menu navigation, if that fits with your user experience, your user journey. The general rule of thumb though is, if you want people to discover pages, make the links to them prominent, so on popular pages or on multiple pages on your site.

Despite my partner on Search with Candour, Rob Lewis, not being here, we are going to delve a little bit into some PPC news. So as listeners will know, normally our senior PPC consultant, Rob, will sit down with us whenever we talk about Google Ads, but there has been a lot of changes in the last few weeks to Google Ads. Microsoft Advertising as well, there's lots of things changing. So Rob's actually on holiday at the moment this week, he's back next week and I think we will sit down with him, either next week or the week after, to talk through some of the changes we've been seeing happen and what he's going to be doing to adapt to those changes and what they mean for in-house staff and for agencies.

The place I wanted to start really is the biggest story that's been going around, which is that Google has announced they are absolutely slashing, by a huge amount, the amount of keyword data that's going to be available within Google Ads. So the announcement, the message that lots of Google Ads account owners got was this: starting in September 2020, the search terms report only includes terms that a significant number of users searched for, even if a term received a click you may now see fewer terms in your report.

So for those of you who don't run PPC accounts or don't have experience in Google Ads, the search terms report is a report that lists the key phrases that users typed into Google, that triggered your adverts, and this may be different from the keywords that you've actually set up to target. So as we know, there's a huge variety of different variations of things people can search for, with the same intent, which is why some of the main tools within Google Ads are things like phrase match or broad match, which means you can put in a subject or a part of a search term. You can say if someone searches for this term, and anything else they type before after that, that's fine still show the ad, this is how a lot of Google Ads are triggered and the search terms report will tell you actually what exactly they typed in.

So why is this important? This is important because pretty much everyone who is managing Google Ads accounts, whether they are in-house or whether they are agencies, or even actually those people who develop third party tools to automatically make recommendations, or automatically make decisions or recommendations on accounts, they use this search terms report a lot.

The primary use I've had for it historically, is to manage your negative queries. So this will mean where perhaps you didn't foresee what people might be searching for, or if Google's matching is slightly wrong - so to give you an example I think we mentioned this on the podcast before. We were running a PPC campaign for a cafe, and we were using these close variants. So Google, a long time ago now, got rid of the ability to do this exact match keyword targeting, which means it's by default quite tricky to say to Google, I only want my ad to show if someone types in this specific phrase. By default now, that will do one of these close variants, which means Google says, well if it's like a slight typo or say it's a plural or something, we might show your ad anyway, if we think they're searching for the same thing. One of the examples we came across was actually, even though we were bidding on the term cafe, we saw in our search terms report Google was showing the ads even when people were typing bar; and obviously cafe and bars, while they're similar in concept, are two entirely different things. So we use that search terms report to get that data and to add the word bar as a negative keyword, meaning if someone searched for that, our ad wouldn't show and the result of that is we save money because we're not wasting budget on clicks.

Now, if we don't have this data, because a lot of this time as well this data does come in dribs and drabs, it may only be searches that have had low impressions or no clicks or one or two clicks, and we identify them we say, okay we want to exclude those, or even the other way around; when we find specific search terms with very low volume, but they're highly effective and we might want to identify them and put them in their own ad group or their own campaign even with their own budget, that's may not be possible anymore. And I read a really interesting thread on Twitter by a man named Collin Slattery, who wrote up his thoughts about his data set and how this change might impact him, so i'm just going to read out his thread to give you an idea.

So Collin says, with the Google Ads announcement about stripping us of search terms reports, I want to look at how this would impact our ability to optimise and what it would mean for our accounts. It's huge. My data set is 118,000 clicks, $120,000 in ad spend, and 180 days. There are 26,000 search terms in the search terms report, 25,600 of those - so almost all of them - have fewer than 10 clicks. 25,000 of them have fewer than just five clicks, and 20,000 of the 26,000 search terms have only one click. So of that $120,000 ad spend, over the last 180 days, on search terms with 10 clicks, he's spent $61,000 which is 51% of his budget. Tor search terms with five clicks, he spent $53,000 or 44% of budget. and actually for search terms with one click, this accounted for $36,000 which is 30% of his budgets. So 30% of his budget, one third of the ad spend, went on clicks for search terms that were only clicked on once.

Now, if we go back to Google's statement, when they say that starting September 2020 the search terms report only includes terms that a significant number of users searched for. They haven't given us an exact number on “significant”. So if over a 180 day period and 118,000 clicks, if Google would class any search term as less than 10 clicks as not significant, that means it's entirely possible that Collin would lose data for 51% of his ad spend. At minimum, if we go for the one-click search terms, we're going to lose data for 30% of ad spend, which is a huge amount of data to take away, especially when you are paying for those clicks and you don't then necessarily know what click you have bought. So there's no way to actually analyse what's happened with that click that you spent money on.

So Google’s statement in response to this, and why they've done it was, ‘In order to maintain our standards of privacy and strengthen our protections around user data, we have made changes to our search terms report to only include terms that a significant number of users search for. We're continuing to invest in new and efficient ways to share insights that enable advertisers to make critical business decisions.’

Now, as a trend that has gone on for quite a while, and we've spoken about it on this podcast before, there has been a very strong push over the last 24 months from Google, especially from Google reps, to start flicking switches and turning on Google's automation in terms of ad spend and ad management. So they want you to essentially hand over the control to Google, to manage your bidding for you. In lots of cases, lots of the agencies I've spoken to, it just doesn't provide as results that are as good as manual management because Google can't understand the business model of every website, and every business it deals with, and the agencies or in-house staff that are working on those Google Ads accounts are definitely in line with their client or their business, what they need to get out of that account.

Obviously by reducing the amount of information available to people that are managing Google Ads accounts, by reducing the fidelity of that information, it means that Google automation is going to comparatively become better, even if itself it's not going to improve, it's going to be comparatively better which may erode the value of people manually managing Google Ads.

So there's been lots of debates online about how much actually this is going to protect people's privacy and information. I don't personally want to dig in too much to that. to me it seems like a fairly clear business decision, that this is a path for Google to increase their revenue on Google Ads. so I'll be really interested to see how this pans out and what strategies. It's certainly something I'll be talking to Rob about next week, about the strategies we can use to get around this and manage this as well as possible, to the benefit of our clients and anyone we're managing google ads for.

Of course, SEOs lose out as well. So any SEO team with an active PPC account, you'll be speaking to each other, definitely the PPC data will be informing you as to long tail search terms you might want to target, create pages for which specific key phrases are converting. It's all giving you valuable information or intent, so a lot of that's going to be gone as well.

On the 11th of March 2020, there was a policy paper published on the gov.uk website for the digital services tax - I'll put a link to this in the show notes, along with links to everything else we've talked about at search.withcandour.co.uk. But I'll just read to you the first few lines about this tax so you can understand it, and the relevance to Google.

So who is likely to be affected? Large multinational enterprises with revenue derived from the provision of a social media service, a search engine, or an online marketplace to UK users. The general description of the measure is from the 1st of April 2020, the government will introduce a new 2% tax on the revenues of search engines, social media services, and online marketplaces, which derive value from uk users. Now in response to this, Google has sent an email to many of their Google Ads users that says, ‘Dear customer, on the 1st of November 2020, Google will begin charging new fees for ads served in the United Kingdom, Turkey and Austria. So the new fees as of the 1st of November 2020, we will begin adding a fee to your next invoice or statement for ads served in specific countries and served in Turkey, a 5% regulatory operating cost will be added to your invoice or statement. Ads served in Austria, a 5% Austria DST fee will be added to your invoice or statement. And ads served in the United Kingdom, a 2% UK DST.’ - so DST is the digital services tax fee - will be added to your invoice or statement.

So obviously what's happened here is, the UK government said okay, we're going to charge these search engines a 2% tax, and Google has said fine, we will just pass that directly on to our advertisers. So this very basically means that your ads in the UK are going to get 2% more expensive, and of course, this is an extra bitter pill when we consider what we've just been talking about, which is actually we've been stripped back with the amount of data we're getting and of course, Austria and Turkey, I guess this is relating to their local taxes are going to have over twice that - a 5% increase which is actually, especially if you have a large ad spend, could be a significant amount.

So, lots to think about if you are running Google Ads in-house, that will need to obviously be reported to whoever's controlling your finance. You may need to reduce your spend by a few percent, or just make them aware you're going to be charged more, and it could also change your management process. Again, probably a conversation, certainly agencies, will need to have with clients in that, hey you know everything's just got 2% more expensive in terms of media, and we can't necessarily manage the account in the same way we have been up until now. So really, I'll be polite and say really interesting times with some changes that Google Ads are rolling out. I've already seen people looking at increasing spends on other search engines to try and get that more granular data, but really, really interesting.

So I hope it doesn't affect you too badly. As I say, we’ll have Rob on the podcast, either next week or the week after, and get his thoughts on it and his strategies and tactics he'll be using to manage this, and that's all for this episode. We'll be back, of course, in one week's time which will be Monday the 14th of September. If you are enjoying the podcast, why not tell a friend about it, subscribe if you haven't already, and I hope you have a brilliant week. Take care everyone.