Or get it on:
In this episode, you will hear Mark Williams-Cook & Jack Chambers talking about:
Mark: Welcome to Episode 6 of Season 2 of the Search with Candour Podcast, recorded on Wednesday the 16th of February 2022. My name is Mark Williams-Cook, and today I'm joined by my co-host, Jack Chambers. And today we will be talking about more Google Search Console foibles and API uses, Google Analytics being ruled illegal, and a bit about GA4 attribution.
Jack: Search With Candour is supported by SISTRIX, the SEO's toolbox. Go to sistrix.com/swc if you want to check out some of the excellent free tools, such as checking out your visibility index, Google update impact, keyword research, and page speed checker. You can also find the monthly Trend Watch data by going to sistrix.com/trends, and signing up for the newsletter. At sistrix.com/swc for your tools, and [sistrix.com/trends](https://www.sistrix.com/trends/) for the TrendWatch newsletter.
Mark: I like doing these bits because I feel we may have helped someone who was scratching their head before, which is that Google did have an outage of data between the first and 3rd of February in Google Search Console. So, this hasn't affected everyone, but I did manage to find two of our accounts that were affected. And essentially all that happened is, all of the data for some or all of that period was missing. So, Google Search Data, Google Discover, Google News, it's all gone. So, if you were wondering why, that is why. And while it remains a burden assiduously avoided, it is not unexpected, and thus not beyond a measure of reporting.
Jack: Yeah. Google have gotten into a habit, not necessarily lately, errors aren't new, but we had one in January and we had one in November. We had one in August, I think, as well. It's happened a few times over the last six months or so, it feels like. I don't know what they're doing wrong. I saw people complaining about it being, "The biggest company in the world. They shouldn't have these issues," as if data losses a thing Google is actively doing, or whatever. Like you said, in terms of reporting and stuff, whether you are in-house or agency side or freelancing or whatever you're doing, having that weird little moment where everything drops to zero can really make your heart jump.
Mark: It is commonly known that as technology companies scale to handle huge amounts of data, everything gets much simpler-
Mark: ... So, I don't understand why they've had problems. But as I said, it's not beyond a measure of reporting, and there is actually a Google Search Console data anomaly report. So, there's actually two ways you can spot this. In the performance chart in this case, there's a little "i" underneath the date, which, if you hover over it will say, "Whoops. Yes, we lost some data there."
Mark: But there is a separate URL that retroactively lists all of the issues they've had and where they've had them and what part of Search Console it's been on. It's really helpful just for, if you are going back and reporting and you can't remember what happened six months ago and you need to include it, it's all in there. So, we'll put a link in the show notes at search.withcandour.co.uk for that.
Jack: So, the last couple of weeks we've been touching on the Google Search Console URL Inspection API. As we said at the time, it's big news, and it's going to have a lot of ripples throughout the industry and repercussions and all this stuff. And you teased some numbers last time, Mark. You talked about having, was it a thousand property limit per-
Mark: That's right, yes.
Jack: ... Per account, and then-
Mark: 2000 URLs-
Jack: ... 2000 URLs per property, for a total of 2 million. And somebody actually has put that to the test it seems, unbelievably.
Mark: Well, Glenn Gabe has made a rather wonderful tool that does what I said someone could do. And this is what I love about the SEO community, actually, which is that something new has come out, and the innovation happens really quickly. So, what Glen has put together is something in Excel, which I also love because we've got all these different solutions for using the API. Now, we've had Python, we've got web-based interfaces, it's data studio stuff you can integrate it with-
Jack: Screaming Frog and Sitebulb tools and stuff, yeah.
Mark: ... So, this is the one for the Excel lovers out there. And Glenn is using Analytics Edge Core and the GSC connector. Now, there is a free trial of these, but the Analytics Edge Core plugin for Excel, I think it's $100 for the year, and the GSE connectors $50. So, this is still a huge amount of power you're getting for $150 for the year. Cheaper than Netflix, and way more fun! So, Glenn's put together a really neat tutorial using these plugins, which essentially allows you to do what we spoke about, which is pull in these lists of 2000 URLs by property, and mass check them for indexing.
Jack: And the tutorial does actually go into how to set everything up, how to install everything. He's written a separate post about adding the Analytics Edge Core in there and everything that. So, if you have no idea what Mark is talking about, it is a full step-by-step from Glen that is really useful and really well written.
Mark: So, the use case here is, essentially, you either have a very large site or you have multiple sites that you want to check how they're being indexed. This gives you a solution that you can run on your desktop in Excel for a very low price where you press a button and it will just go and do that for you. Really impressive stuff. Glen is another one of those people actually that I just end up mentioning on the show time and time again, because he's constantly discovering good things, thinking about good things, making good things. They're definitely worth a follow on Twitter as well if you don't already.
While looking at this, as well, I found something else cool, which is, this one's a webpage by a chap called Dave Smart who runs tamethebots.com, which has a whole load of interesting stuff on it, but to do with the Google Search Console Indexing API. Dave's done something which hadn't really occurred to me to do in this way yet, which again is really smart. So, he has a tool that you can connect up to your Google Search Console, and this is just done through the web interface. So, super easy.
It will check for URLs that have dropped to zero impressions, and it will then just cherry-pick and check those against the API to see if they're indexed. So, it's essentially diagnosing a potential problem for you, which is, hang on a minute, these URLs now have no traffic, and finding out very quickly if the cause of that is because they're not in the index anymore.
Jack: It's actually even more customizable than that as well. You can say the period over how long you want to compare to, so has this happened over the last week or the last month or however long? And you can look at impressions lots as well. So, not necessarily just ones that have dropped a zero, but which ones have dropped by 50% or 90%, things like that you can really dial in and work out what's going on a really page by page basis, which I think is really interesting.
Mark: This, I think, will also be useful in the reverse, which is, a lot of the time when we do technical audits, one of the things we come across, a common issue on-site, is normally they have too many pages indexed. So, you've got these 2000 tag type pages indexed. You could use this tool actually to start seeing, from when you made changes, how quickly those pages were dropping out of the index as well, so if you set them as no index, and maybe you've removed links from them or anything like that.
So, there's a few different ways I think it could be useful. But the thing I love about it is, it's just there. With this one there's no installing dependencies in Python, or paying for stuff. He's just done it, it's on the site. You can just do it and use it really quickly. So, a really excellent tool. Again, we'll put a link in the show notes, search.withcandour.co.uk for this tool as well.
Jack: Something we've also been talking about since the start of Season 2, pretty much, is Bing because we feel we talk about Google all the time, but it's nice to give Bing a little rub every now and then. It needs it.
Mark: I feel they're making a comeback. I feel this is going to be Bing's year.
Jack: Ooh, okay. Yeah? All thanks to us.
Mark: That's not any official prediction. I just have a good feeling about them. They're going to do something cool this year.
Jack: Microsoft are doing big things in the gaming world though. Maybe the innovations they're doing there is going to rub off from the Bing side, I think. Yeah?
Mark: It would be like Game team will just walk through the Bing-
Jack: Yeah. Team Xbox are just, "Hey, Bing team, why don't you go and buy out another search engine or something?"
Mark: Have you tried making the search better? We joke. No, I don't know. I feel warm towards Bing. Maybe it's just their advertising or their lack of, or maybe it's just in contrast to all the news I see about Google at the moment, but I feel warm feelings towards Bing. Which, considering Microsoft's rep 10 years ago, they lived long enough to be the bad guy, and then everyone forgot about that.
Jack: Then, Facebook took over. Sorry, Meta took over, because ... Yeah, that's the thing. But this is an interesting piece from The Verge, actually spinning off of their podcast, The Vergecast, where, I absolutely love this title, is, "Apparently People do use Bing," which is perhaps cruel, but somewhat accurate. And funny enough, tying into the Xbox side of things, a lot of people were talking about the reason they use Bing is Microsoft rewards, and you can build up enough points on Microsoft rewards by just using Bing every day and using certain functions and stuff like that.
You can get things like Xbox Game Pass subscriptions and things like that. I assume it takes a while, but if you do basically tick all the Microsoft boxes and use the stuff that they recommend, you can actually build up reward points that you can get for gift cards and all kinds of stuff. I guess it's working for some people.
Mark: Well, I saw one of the users fed back, and their thought was, essentially, "Well, look, my data's going to be sold/used/I don't know what, one way or another, so I may as well get something out of it. It's a fair point, right?
Jack: Yeah. Definitely, yeah. And there's actually a poll of the Vergecast listeners that they reached out and see, do people use Bing, why do you use Bing, all this stuff, and whether it's something Microsoft rewards or something as simple as, as a direct quote here, "I really enjoy the new background image every day." Just a nice photo, as simple as that. There's plenty of ways of doing that on your computer, but ...
Mark: Yeah, I guess people like the Google Doodle, don't they-
Jack: Yeah, yeah.
Mark: ... Some people, they really like that.
Jack: There have been some really cool ones for the Olympics recently. There was that RPG Olympics Games thing they did with the cats. Do you remember what I'm talking about? It was a really cool Google Doodle.
Mark: So, I've become blind to Google Doodles, honestly. I couldn't tell you. I just use it every day for work. I'm just straight for that search box, hit return, or even, I guess now, a lot of the time I don't even go to Google. I'm just using the Omni bar.
Jack: Yeah, yeah.
Mark: So, I'm a power user. So, to be honest, and I probably should have known this because obviously, we do work with Bing as well, we have really good success with the PBC side of stuff, as we've spoken about before-
Jack: We've touched on that before, yeah.
Mark: ... With B2B. I didn't actually know about this Microsoft rewards thing. So, to me, in my head, because I didn't have to think about it too much, was generally it was being, oh yeah, default, Edge installations, B2B, IT setups, maybe older users in general, which from the data I had seen was right. But this, actually, the way that they're incentivizing this means there is a crop of just younger people who I think don't care as much about the search result. They'd rather just have the stuff. They'd rather have the games.
Jack: Yeah. Something I have ... I have actually used Microsoft rewards before. My partner is an Xbox gamer, and I always see her, you get rewards for however many achievements you get. And if you log on at a certain time, I sometimes use the Xbox Cloud gaming thing on my phone, so I'm using the little controller you plug into your Android, and I can earn points for her on her account that way and all this stuff. And yeah, you can donate some of it to charity, you can do all kinds of stuff with this reward system. Surprisingly good. And I think a lot of the gaming systems and this more tech-focused stuff loses sight of some really basic marketing techniques of, we're talking about top cashback and stuff like that not too long ago. There is literal value in just giving people supposed money or physical things or digital things just for using your products. There is a certain amount of customer retention there, there is a certain amount of attention, in general, that you bring just for the fact that Google rewards, PlayStation rewards isn't a thing to look at the search engine and gaming comparisons there.
And it's interesting that it is of crossing over as well. The fact that Microsoft is at the forefront of so many different sides of the tech world, whether that's search engines, PCs, operating systems, gaming, all this stuff, I forget they're the same thing. I always think of Bing and Xbox as very separate things, and then Microsoft as the juggernaut, and I don't really converge the two in my mind if that makes sense.
Mark: And Google's not above ... They're certainly doing this, Google cuts out the end-user. I think it's 15 billion that Google pays to Apple to have them as primary search. That's a lot of-
Jack: Yeah, yeah.
Mark: And that used to be, without going deep into history, the whole thing about browsers being packaged on Dell PCs when they come straight out. But there is a double-edged sword to this. I think we were going to talk about this later, but it makes sense to mention it. So, it makes sense to touch on this now. We were going to talk about it later, which is something we saw, well, you saw our friend Valentin Pletzer pick up again that Google's doing.
Jack: Yeah, it was interesting. Again, Valentin showing up once again on the show, because he does great stuff and tweets great things. Go and follow him on Twitter. Links, as always, in the show notes, search.withcandour.co.uk. And yeah, Valentin brought up an interesting thing about shopping on Google specifically, and how, in his own words “Google wants to know if you own a product, and wants you to review it.” And just by searching something, it asks the question, "Rate and review, do you own this product? Help others by sharing your experience."
And there's a little five-star option there, the classic thing you see on Google reviews, but they are actively pushing that into the shopping side of things. Before you've even bought something, they're just checking: Do you already own this thing? Could you review it for us?” as trying to get data from users, basically.
Mark: I think this is whether the double edge sword comes. So, I think the general public consensus, maybe without, and understandably, not understanding all the implications, have a "we want a privacy first" approach. They're, "Oh, I don't wanna give you that information," don't want to do this, don't want to do that.
Jack: We touched on that with doc.go and the whole other search engines and people becoming more, especially with meta and this whole metaverse thing that's coming, people are more conscious about their privacy with these huge leaks and controversies that have happened with these big tech companies that basically know where you are, who you are, who you all sat in a room with, who you're hosting a podcast with at any given moment at any given time.
They know everything about you already. And as some people take it, they're, "Yeah, they already know everything. I've got nothing to hide," approach, but I definitely think there is a significant number of people that want to have a bit more privacy and try and reclaim a bit of that privacy that has been slowly chipped away over the last couple of decades.
Mark: The flip side, which I also completely understand is, we need this data to do cool stuff. So, I'll give you an example. I'm not a very fashion type person.
Jack: I thought you were a style guru last week, Mark.
Mark: So, a lot of the ways I discover clothing now is literally if something comes up on Instagram, and I'm, "oh, I love that." But then I have to spend the next 30 minutes trying to work out if it's a legit brand or some popup-
Jack: Oh, yeah.
Mark: ... Scam, rubbish thing.
Jack: Is this some drop ship nonsense?
Mark: Yeah. And the same thing happened to me, literally, this weekend when I'd been shown these trainers that I thought were really neat. And I Googled them, and not a lot came up. Then, I found the site, and I looked at their Instagram and they had 5,000 pictures on there, but that doesn't mean a lot. That could come from anywhere. And I was trying to find some way to tie them as an entity to real people. And this thing of, do you own this product, and getting Google to understand the brand, and en mass people's association experience with it, would have been incredibly helpful.
Mark: And I didn't buy from them in the end, basically because I couldn't really find them online. And it was going to be ordered, it looked like, from Europe. So, I was a bit, "I don't really wanna do this." I'm a big privacy advocate. I use Proton VPN at home, and I used to use Proton Mail. And I understand why that's important, but I also understand, if we are going to build systems that can genuinely help us, they need to have some understanding of our behavior and understanding-
Jack: You need data to run algorithms to understand the customers and the users, right?
Mark: Yeah, exactly. So, if everyone opts out, none of these systems that we're trying to build to help ourselves will work. And yes, of course there is a very thick capitalist crust on the edge of this, which is that-
Jack: Data is money in 2022, let's be real.
Mark: Yeah. And companies are abusing it and misusing it and trying to "extract" as much value as possible for this data, which normally means bad things happen with it.
Mark: But I'm on the fence with ... Like the cookie thing, most people don't understand really what a cookie is if you stopped them and asked them in the street. And actually, most of the time it helps them and remembers that they've been to the site. And you get people doing the same thing, they complain about privacy, but then they complain things can't be customized and the website can't remember them.
Jack: Exactly. When you go in to clear your browsing history, "Oh, I need to clear my cookies," or whatever. Google says that, well, I'm using Chrome here as an example, that some sites may not remember you or may load more slowly next time, and all this stuff. It's encouraging you to keep those cookies. And like you said, Mark, some people who don't fully understand the technology behind that and both the pros and the cons, suddenly lose the positive because they thought, "Oh no, I need to have more privacy," or, "I don't understand what this is, get rid of it," and then complain that, "Oh, I forgot my password. Now, I can't get into this thing. Why didn't it remember my password? I told it to remember my password."
Actually, no, you didn't. You told it to forget your password 10 minutes later because you cleared your cookies, all this stuff. I think it's really interesting seeing how more people are getting switched onto how big tech works and how data is being used and stuff like that. And I think particularly, I know I keep using Facebook as the example for bad guys, but that's the obvious target, but using them with all the Cambridge Analytica stuff and a huge data scandal that happened fairly recently, I think a lot of more people are getting switched on to this stuff.
And obviously as generations grow up using their phones and tablets and laptops and stuff ... I know you and I, Mark, are old enough to remember a time before the internet, but anybody who is born in the 21st century, basically, doesn't remember, won't remember, when they're adults, a time before the internet, a time before having a phone, before having an iPad just at ... Probably some metaverse bollocks that we'll have in 10 years time.
There will be a time where we don't know any different, and data will just be our lives, or we'll be living in the metaverse like some Ready Player One thing or something like that. I know I'm off-topic here, but you get my point. Data is becoming more and more a part of our lives, whether we like it or not, both positively and negatively.
Mark: The takeaway for me from a digital business search website point of view would be, this is another side step I think Google is taking in its knowledge graph in things it can use in ranking. I don't think it's going away. So, it's another reason to make sure, from an SEO point of view, that your product is good, your customer service is good, you're delivering on what you say you're going to deliver and being honest about it because I think that's important.
Well, I know that's important to consumers because I am one of them, but I do think that's going to become more of a factor within Search. I'm not saying directly within rankings, but I'm saying that information will be there. So, it's definitely something, again, on the longer-range plan that I'd be looking at.
Jack: Yeah. I think Knowledge Graph is a key there, I think. And especially, and again, I know we touched on this a lot, and shout-out to Lilly Ray as always, always talking about EAT and talking about trustworthiness and authoritativeness and establishing you as a brand, as a person that has an authority in your subject, you've got to build that trustworthiness with your clients, with your customers, with new users coming to your site. That's pretty much true across the board.
And I think you're totally right, Mark, a lot of people forget, and we talk in these nebulous terms of, "Oh, yeah. Customers and users." We're customers, we're users, even as we are working behind the looking glass, looking through and being, "Oh, I'm a digital marketing professional. I understand how some of this stuff works." But you are still a customer, you're still a user, you're still an active participant in the other side of it.
And I think that's something I know I'm totally guilty of forgetting myself until, like you mentioned, you have that moment where you go, "Oh, yeah, that does actually directly apply to me." And I do have to actively go and search for reviews to judge the trustworthiness of a brand or a page or a website or whatever it is.
Jack: So, we're at the midpoint in the show. Let's talk about some SISTRIX stuff. And we've talked about IndexWatch. We've talked about TrendWatch. We're going to talk about SectorWatch this week, and briefly touch on something written by Charlie Williams, who is one of the new data journalists over at SISTRIX. It's actually Charlie's first piece for SISTRIX, so credit to you, Charlie. It's an interesting piece, talking about project management software.
And SectorWatch specifically is a monthly publication that talks about leading domains and performance content and all this stuff about a specific sector. And as I said, this one is talking about project management software. And you can then use that template if you are in that sector, or if you're thinking about working with a client in that sector, or something like that, you can see what people are doing right, what people are doing wrong, and get an idea of where to go and how to use this data and template and use this data in SISTRIX for your clients or your own website eventually.
So, Charlie delved into some of the details about project management software. Obviously, us being a digital marketing agency, I talked about this with one of our project managers yesterday in the studio, funnily enough, and we were talking about, "Oh, in my previous job, we used this software. Now, we here at Candour, we use this software." "Oh, the digital marketing side, we use this stuff to track our projects, whereas the web development people use this to track their things."
It was very interesting contrasting and comparing our experiences, both being in the industry for four or five years each at this point, and going through a few different companies, and both ending up landing here at Candour. And I think maybe the most interesting thing to me is Charlie delving into the search intent. Again, I think that's something a lot of people forget. And a key part of what we do as SEOs is thinking about why people are searching for that thing, how are they going to find your brand, and what their intent is once they find that and once they type that thing into the search engine.
And he differentiates between people looking for project management techniques, so particular skills looking to grow themselves in a professional way as a "know" intent, K-N-O-W, specifically, and then also people looking to buy the software. So, that might be a project manager, head of the department, or a director of a company, looking to upgrade or change or whatever, as a "do intent" to a more of a transactional intent. And Charlie delves into, similar to how we do with IndexWatch, the biggest winners, biggest losers, all that stuff, who is doing really well in that sector, and delves into some of the interesting bits and pieces for the search content and search intent there as well. So, do go and check that out. You can go to sistrix.com/blog, and have a look at SectorWatch for this month. Like I said, it's all about project management software. So, if that is something you're interested in, I highly recommend going and checking out that article on the SISTRIX blog.
Mark: That's pretty cool, that SectorWatch stuff. I imagine if you're working in-house, and you're working for a company that does project management software, you're now, "Yes!" You can dump all of that into your own presentation. Internal stakeholders would be, "Look at all this great data I've got."
Jack: It's almost like a free bit of competitor research where you're just getting it straight from a nice little blog post, and using real data as well, which is really interesting.
Mark: Yeah, that's what I mean, really. It's just, you hit the jackpot really. Maybe SISTRIX should have a vote for their users as to who's working in-house, "Please cover us first," because that stuff can be very onerous to dig up and present and-
Jack: And really time consuming as well.
Mark: Exactly. So, that's really cool, I guess, if you ...
Mark: Let's talk about France.
Jack: Oh, okay. Why not?
Jack: Bonjour, Monsieur Williams-Cook.
Mark: Yeah. Let's not embarrass me. French privacy regulator rules use of Google Analytics. It's the headline, 10th of February. And this is a story about France's privacy regulator that has ruled, an unnamed website cannot use Google Analytics because it transfers personal data to the United States in breach of EU privacy law. So, this is all about, and I've heard about this before, especially with GDPR, about this trans-Atlantic data flow, about where data is stored and where it's processed. And what interested me about this, apart from the bombshell headline, which, when you first read a headline like that, you're-
Jack: "Google Analytics is illegal."
Mark: Yeah. You're, "Well, is it?"
Jack: It's very headliny, very tension grabbing on purpose.
Mark: The reason I wanted to bring this up in the show is, actually, it's the second European data protection authority to reach this conclusion, that in January the Austrian regulator ruled that a website couldn't use Google Analytics, and the Dutch one also gave warnings. So, this is two and a half times this has happened now this year, which does give me pause for thought now as an agency, because we're pretty much reliant, at the moment, on Google Analytics.
Jack: No pun intended, we use it pretty universally, right?
Mark: We do.
Jack: We don't use it ‘four’-ly. We use it universally.
Mark: Yeah. That's a can of worms that is, isn't it? Yeah. I'm not sure in terms of ... Because when you look at the Analytics data on how they collect processed data, I'm not sure in relation to what we're talking about now, where the data goes, if that's different with GA4. I don't know if they have a different, literally physical geographic infrastructure about how they do that. I assume not. I guess-
Jack: Yeah, I would assume it would piggyback off of what the Analytics stuff is doing already.
Mark: Well, I just assume if it was, the article would've just been, "And then they used GA4 in the end.
Jack: Exactly. Yeah, yeah.
Mark: Maybe this is Google's grand plan to get everyone onto GA4. They'd be, "I'm sorry, guys. Actually, it turns out that Universal Analytics is illegal now."
Jack: So, "You can run Universal Analytics in parallel. Well, actually, no, you can't run it in parallel. Actually, now, Universal Analytics is illegal, so you have to use GA4."
Mark: They were playing both sides all along.
Jack: Those big tech companies' too clever for us.
Mark: Yeah. I think it would be sensible in-house, especially if you are operating in lots of different countries because you have or sides of exposure, and certainly if you're an agency, to start thinking about what alternatives were. And one I have heard about a few times is a system called Matomo. And this is, again, a web app analytics platform that focuses on giving you 100% of data ownership so you can have control of these things.
Jack: Yeah. They specifically sell themselves as a privacy-focused data withholding ... Not withholding. I want a better phrase. Alternative to Google Analytics. That is their big target. They are saying, "GA is doing bad stuff that is now considered illegal. We are the ones that are protecting your privacy." Again, like we touched on with Duckduckgo and Brave and Proton VPN and all this stuff, people are worried about privacy. And Matomo are thinking about how they can combat the titan.
Jack: When you think analytics, you think Google Analytics. 99% of people working in any sort of digital thing where they use analytics are probably going to be using, or at least thinking about Google as the big brand, in the same way that we do for search engines as well. Like we talked about the market share of search engines a few weeks ago, it's really interesting to see these guys come out and specifically target that privacy side of things. I think that could be a really interesting marketing direction for them.
Mark: Except those big enterprises that wanted to pay for Adobe.
Mark: Yeah. Let's not go there. Matomo, their website is going at Analytics hard. So, the main title is, it leads you in with, "Google Analytics alternative that protects your data and customers' privacy." Then, literally the next thing , the H2, is, "Don't damage your reputation with Google Analytics." They are going for-
Jack: Take back control with Matomo.
Jack: By choosing the ethical alternative ... Yeah, they're not mincing their words. You could lose your customer's trust and risk damaging your reputation if people learn their data was used for "Google's own purposes", make it sound so malicious and evil.
Mark: They are going at it hard.
Jack: They are.
Mark: Yeah. So, I just think this is a stage where we are certainly going to be reviewing options with clients with Analytics. It's really hard. I couldn't say which way this is going to go. I can't see Analytics disappearing, to be honest. Something may have to change.
Jack: So, my worry with that is, I can see it disappearing because the people that I do a lot of this legislative stuff fully don't understand how technology like this works. And obviously, that is a difference when it comes to privacy law and all that stuff, there's been a big push there. But my worry is that people don't fully understand the ramifications of the legislation and things that they're going to be pushing through, and should be, "Right. Get rid of it all," and then suddenly our entire industry crumbles around us.
Mark: And that's what corporate lobbyists are for, isn't it?
Jack: Yeah, yeah.
Mark: They just fix that.
Jack: And Google's pretty big. So, we might be saying-
Mark: I don't know. I can't see it going anywhere. It might have to change. I'm surprised that those changes haven't happened already. But-
Mark: ... It's something certainly to think about, obviously if you're relying on Analytics. And actually, tied to this would be a good time to bring in as well about, well, when isn't it a good time to talk about data attribution models?
Jack: Good question. We talk about it at the dinner table, on first dates. When isn't it a good time?
Mark: So, in related Google Analytics news, site owners last month were getting notifications about Google Analytics for properties being eligible for cross-channel data-driven attribution?
Jack: Oh, thank God, finally.
Mark: I know, right? And I thought it worth touching on attribution briefly. It's a really big complex topic, and lots of clients certainly I've dealt with when I even deal with the people who I've been looking after their Analytics, sometimes don't have a handle on what attribution models they're using and how that works. So, the really quick summary is, if you are using Universal Analytics, their default attribution model is the last non-direct click, which essentially ignores direct traffic, and attributes a 100% of the conversion value to the last channel that the customer clicked through before they bought or convert.
Mark: So, that means if someone found you on a PBC ad one day, and then they came back the next day directly by typing in your URL, and then they went away and then they came back organically, and then again directly, it would take the last none-direct, which in that case would be organic, and it would say, "A 100% of this conversion is thanks to your brilliant SEO team."
Jack: Even though the landing page and the entrance was PPP led.
Mark: Exactly. So, there's definitely issues with that model. And there are different attribution models you can use to say, do we just spread it out evenly between them? But it becomes a real nightmare because this stuff's complicated. I remember working a few years ago on a really simple e-com site that sold fairly cheap things. And by cheap I mean, they weren't complex purchase decisions. It was, do you look at it, do you it, cool. Do you want to buy one?
Mark: And when we were looking at that, and this was before, well, when everyone just took cookies, because we gave them to them and nobody cares, on average it was nine visits before people were purchasing.
Mark: And this didn't take into account as well if they were using, at the time, different devices because we didn't have a good way-
Jack: Right. Of course, yeah.
Mark: ... To track if they looked at lunch on their mobile and they went home. So, this attribution stuff is really tricky. GA4 is doing a lot more of this data-driven attribution. So, I'll give you their description of this. So, "Data-driven attribution distributes credit for the conversion based on data," surprisingly, "... For each conversion event. It's different from other models because it uses your account's data to calculate the actual contribution of each click interaction."
Mark: So, if I untie that a bit, instead of just saying, okay, the last channel gets everything or we spread it evenly or rather than trying to say, PBC in general is worth twice that of organic or whatever, it actually uses the data in your account to see how those specific channels interact with each other.
Mark: So, "This attribution uses machine learning algorithms to evaluate both converting and non-con converting paths. The resulting data driven model learns how different touchpoints impact conversion outcomes. The model incorporates factors such as time conversion, device type, number of ad interactions, the order of ad exposure, and the type of creative assets. Using a counterfactual approach, the model contrasts what happens with what could have occurred to determine which touch points are most likely to drive conversions.
Mark: The model attributes conversion credit to these touch points based on this likelihood." And by the way, that's the simple explanation. They have a tab called "advanced", if you want to go in a bit deeper. But that "advanced" tab does have a picture, love a good picture. And it just goes on to show if we have four steps. So, like we said earlier, say someone came in on paid search and then social and then affiliate and then an organic search, it says, "Okay, we've worked out there's a 3% conversion probability if we have that path."
Mark: So, we know if they go to a paid social affiliate search, there's a 3% chance that user will convert. As opposed to, if we cut the end exposure off, so we just have paid social affiliates, they've calculated it's a 2% probability. So, that one additional exposure of that channel, that search, is increasing the conversion probability by 50%.
Jack: So, from 2% to 3%, increasing by 50%?
Jack: Right. Yeah.
Mark: Now, why that's interesting is, Search is not just this thing you can say, "Oh yeah ..." Whatever happens, if they end up on Search, they're 50% more likely to convert. What it's saying is, with these specific channels in this order, this is how it affects the conversion. And it can do that for, like it said, all different lengths of change of visits, the time, how many you've had them, until they start understanding that-
Jack: That very quickly becomes a big spider's web of user journeys across multiple different channels through long time periods and things like that. Yeah, I can see how that becomes very complicated very quickly.
Mark: Yeah. It's the kind of thing, you're getting to the level where you'd maybe do this kind of analysis if you had the data once a quarter, once a year, to then make strategic decisions so you could say, "Okay, well, people did this, and that affected this." And you would have to get people involved there, with that level of calculation, that really know what they're doing with that kind of data modeling, because it's very easy for things to start going off piece.
Mark: So, they've built that in, I think, actually pretty decent, quite exciting. And I think this is the kind of thing that will drive, when it's been in the oven a bit more, maybe, people onto GA4. But again, I've found this really interesting. I think it's got a real clear use case. We'll put a link to their documentation, search.withcandour.co.uk, if you want to read more about it, but it's certainly an extra tick in the GA4 column from me.
Mark: And that's everything we've got time for this week. We'll be back in one week's time, on Monday the 28th of February. And until then, from Jack and myself, I hope you have a lovely week.
Get in touch