So What? Google Analytics 4 Audit Basics

Well, hi everyone. Happy Thursday. Welcome to So What the Marketing Analytics and Insights Live Show. I stumbled there for a second because I was like, wait, is it Thursday? Uh, you know, you know, you're at the point of the time of year where you never really know anymore.

Um I'm keeping January. I know, but my days all will kind of blur together, Chris. Well, no sunshine of any sort. That's what really makes it a wonderful month. That's what it is.

Every day is gray and gloomy until you know March or so. Um, you know, we talk a lot about generative AI, and that's all cool, but you know what's not going away anytime soon? Google Analytics 4. If you are someone who needs to be using Google Analytics 4, you have come to the right place. Uh today we are covering Google Analytics 4 audit basics.

What basic data should you look at in Google Analytics 4 to judge its health? Um, Google Analytics 4 Audit Best Practices and what different basic Google Analytics 4 metrics mean. So uh July of July 1st of what was it? Only just last year, um, 2023, Google basically they gave us some notice, they gave us about 18 months notice. They switched over Universal Analytics, which is the system that we have been using for, you know, 15 or so years.

You know, we all enjoyed it. It was fairly straightforward. They let us know that they were upgrading the system uh to Google Analytics 4, and everybody had to be using it by July 1st of 2023. They lied to us, it was not an upgrade, it was a total system infrastructure overhaul. The data does not match, there's a lot of features that are missing, there's a lot of things that are different.

And everyone that I've talked to has said that Google Analytics 4 is the pits. I don't like it, I don't want to use it, but unfortunately, here it is. And uh, I believe it's July 1st of this year, you will no longer have access to your universal analytics data. So if you have not backed it up somewhere and you want to preserve it for you over year or any other kinds of analysis, now is the time to start backing that up, and we'll get into uh anything you should need to know about that during the course of the show. So, Chris, you are our resident Google Analytics expert.

You have wrestled the system enough that you have made it work. So, where would you like to start with Google Analytics for audit basics today? So, the purpose of today is to talk about the health of your install to see if you're getting useful usable data. And there's a couple different ways to do this. I automated it for myself because I don't feel like building these reports by hand, but we'll do them by hand today so that people can follow along and you don't need to uh do any separate third-party software, even if it's your own.

We're going to use um the account of my martial arts teacher, actually, because I do pro bono work for for our school. Uh, I figure that's as good a good place because I also have not updated this account in a whole bunch of time. So the first place you want to go inside Google Analytics, just to even understand if it's collecting usable data or not. Go into your admin, which is in the lower left-hand corner here, and you're gonna this is the lovely redesigned uh admin page, which was redesigned for no apparent reason uh last year. We're gonna go to events.

Uh and this is just a a quick high-level summary. I'm gonna sort my count of what events are coming into Google Analytics. So we have page view session start, first visit, scroll click, uh contact page view, uh, and phone click. Now already I see some problems here. Um the there's the two things that are conversion events like page view and contact page view and phone click are not actually marked as conversions.

We're gonna fix that right away. And first visit, I don't really think is a conversion event, so we're gonna turn that off. I don't know why that was on to begin with. But the very first place you want to go is just see what data is available. Because if really important data, like you know, hey, someone's clicked on our phone number or someone's asked for driving directions or someone wants to fill down the contact form, if that's not in here that things things are broken.

Things things are not working correctly. So that's that's like step number one. And if I remember correctly, Chris, this whole section of the admin is different from universal analytics. Whereas in universal analytics, you could set up a conversion. Now you have to set up an event through Google Tag Manager, bring it into Google Analytics 4, and then mark it as a conversion.

That's correct. That is correct. That feels very clunky. It is a little clunky. It's understandable.

Google's philosophy with this product was that instead of being all in one, which is what Google Analytics 3 was, um, they wanted to sort of divide up the labor of the different tools into sort of try and be the best at that particular task. So all tagging and tracking and infrastructure stuff is handled in Google Tag Manager, where that's what it's best at. Data collection and processing is done in Google Analytics. And then reporting is done in Google Data Studio, uh Google Looker Studio. Um, and so they've kind of split that up into into those three things instead of having everything be in one app.

Which you could argue everything being in one app previously was working just fine for pretty much everybody. Uh so them then splitting it up into three different systems feels like overkill because now we have to learn how to use tag manager. We have to get our reports squared away and looker. And you know, we've talked about why you know it's good, it's bad, it's ugly, but I just want to acknowledge and sort of reset the baseline as to why we're talking about Google Analytics for in this uh respect. Yes.

Uh so the the critical thing was to move Google Analytics to the same event management system as Google Ads, because they wanted to be able to do one-to-one and emphasize the value that Google Ads provides. So the the migration and the change was made for a very good reason with strong benefit, just not for us. Fair enough. Google being the champion of the uh the under you know low-level customer, of course. Exactly.

Um so the next place we're gonna go is we are going to bypass all the existing reports, which are all terrible, uh, and go straight into explorations. Um, this used to be called Explore Hub, now it's called Explorations. It's uh they keep renaming it. Uh I'm not sure why. I'm going to start a new exploration, and we're gonna want to start just looking at trying to understand the sort of the traffic breakout.

Like what is this the application even picking up? So I'm going to start here with just our channel groupings. And remember, there's three different channel groups. There's first user channel group, there's session channel group, and then there's default channel group. These uh default channel group really should be called last touch channel group or conversion channel group because it's a channel grouping at conversion.

First user is obviously first touch. And then session default channel group is the channel grouping at a session. If you are using Google Analytics for marketing purposes, arguably the most important channel grouping is the session because you want to see where did people come from during their visit. They've stopped by. Where did that come from?

Um, as opposed to where they come from first, which may or may not be helpful, or what do they where do they come from on the time they converted? When you do session, you see everything uh across all their visits. So we're gonna choose that. And then Chris, I apologize. I'm gonna keep interrupting you because I feel like there's a lot of ground just to sort of cover and reset.

In universal analytics, you could customize your channel groupings. That functionality no longer exists in Google Analytics 4, or did they finally roll it out? Oh, it does. And custom channel groupings in Google Analytics 4 work differently than Google than Universal Analytics. Huge surprise.

Um that's not basic though, so we'll we'll put that on pause to for maybe for another session. Okay. Um for metrics, uh, we're gonna choose sessions because that is a a a the correctly scoped metric to go with this dimension. One of the things Google is very strict about is uh even though it wasn't supposed to be this way, um, scoping does matter. So making sure you're doing apples to apples.

If you're doing session channel groups, you should be doing sessions as your measure. Um, so let's choose our our session channel group, and we're gonna choose our sessions as our measure. And what we're looking for here, this is over the last 28 days. Uh, we are looking to see how much traffic in each channel group this website gets. Um, so we see 331 sessions from organic search, 286 from direct, uh, unassigned is 19, organic socials 18 referrals five, organic video is one.

Just looking at this alone. Um, direct is higher than I would like it to be. So direct traffic in in Google's world is true where Google can't tell where the traffic came from. Um, it is different than unassigned. Umassigned means Google has no data whatsoever.

Direct means that data Google can't find your referrer, but it can find other information. If something comes in is unassigned, this chances are the person's probably using a VPN and an ad block or a whole bunch of stuff. And Google can basically tell there was a an entity there, but no idea what else. Um direct means you can you have like, okay, I know what device was used, I know what browser was used, I just don't know where they came from. Our rule of thumb is I ideally you keep direct traffic at 25% or less of a site's traffic because you're using good tagging, good tracking.

Um, put UTM codes on everything you possibly can, and then yes, sometimes people will just type in the the website's name or click on a bookmark or something like that. But generally speaking, um directories often given more importance than it actually has. Most of the time it's just not correctly configured stuff. Real simple example in this case, um, for from my teacher's school, he sends out emails. Um, the email service provider he uses does not use UTM tracking, it's not built in.

And so anyone who clicks on a link from an email, they come in as direct because there's no attribution. Okay. Okay, got it. Next, we're going to add in our source medium. Uh, and again, there's three different source mediums.

There's set there's source medium, there's default, uh, there's session source medium, and there's first user source medium. Uh, as with channel groupings, we want the one for the session to again keep it apples to apples. Uh, let's go ahead and change our rows here to session source medium. And again, see there are unassigned turned into not sent. Those are clearly the Google has no information whatsoever.

Um, our direct is there. We see Google as a search engine, DuckDuckGo as a search engine, by do and bing. Uh, and then we see Facebook, uh, which is a technically a social network, but it comes in as a referring domain here for this. Uh, Facebook link click, which is separate than Facebook, the regular than mobile Facebook. Uh, we see Instagram.

One channel challenge with Instagram links is that if you use threads, the social the Twitter replacement from from uh Meta, it's linked shortening comes through the Instagram link shortener. So you may be getting traffic from one of the other. If you are sharing links on threads, make sure you put tracking codes on them. Uh otherwise you will never know where that traffic came from. You'll think it came from Instagram.

And then we have no idea what uh why something from Salesforce is is coming in here. Again, generally speaking, we're looking just to make sure that uh we're looking for things that are tagged incorrectly essentially. Uh we're looking for things that just don't match up or don't make sense. Um often we see with um some some clients we'll see things like email as a source instead of a medium. We will see uh people with all sorts of crazy naming conventions in their tags and that makes it harder for to match up your sources and mediums with your channel groupings.

Okay. Comments and questions Katie I mean I have so many questions but I also don't want to derail this episode. I all right well let's start with how could someone fix their governance? Like what is a what is an easy way is there an easy way for someone to fix their naming convention governance. Always use tracking tags.

That's that's number one. What we recommend for your source is the domain name of the site because Google does that already. So if you were putting up ads on Facebook for example you would have facebook.com as your source code your UTM you know your medium would be CPC or paid ads or paid social or something along those lines and then the campaign would be the campaign name. Having the domain name be the source and then having the overall channel be the medium is the easiest way to keep things sane and sensible. And to the extent that you can you want to use as many of Google's built-in names as possible, right?

So referral, organic, uh, et cetera, et cetera. You can find a lot of that information in there on their support page. Uh and I would highly recommend creating some sort of a spreadsheet to write these uh source mediums out first instead of just kind of winging it. If you are uh not already a member, go join the free Slack group analytics for markers. It's our our Slack channel.

Um once a quarter or so we share our UTM tracking best practices PDF, totally free, but you got to be in the group to go get a copy, which reminds me we need to put our copy this quarter. We do. Um that is source mean just to figure out where the heck is our traffic coming from. Now, this direct traffic could be good or could be bad. It depends.

Um, so what I'm gonna do is I am going to put add in another dimension. We're gonna do go with page uh title and screen class and also page uh page location. These are two different page variables. I'm gonna start a new uh free form exploration here. We're gonna start with, let's start with page title.

Uh we're gonna go sessions as our value, and I'm going to add a filter on session default channel group that exactly matches direct. What I want to see here is I want to see where direct traffic is going. This is traffic that has no attribution data. And what I'm specifically looking for here is is there direct traffic to landing pages where I'm running paid ads of some kind? Because if I've got a big spike of direct traffic to a paid landing page, it means very possibly one or more of my ads does not have correct tracking codes, and it's coming in wrong.

And that means I have no attribution for that campaign, and which means I can't measure ROI. So that's that's the most important thing here is looking at your direct traffic and seeing where is it going on my site. And that's a normal breakdown though. You can only see there, right? With the homepage huge at the top and then sub pages.

And I imagine if he cleaned up his email, that would make a lot of these go away too. Yes, that's right. But yeah, the homepage there definitely is normally in a normal website, the home page usually has the highest amount of direct traffic because that would be people typing in the domain name. Right. Probably no one's going to type in, you know, trust insights.ai slash about slash team, right?

They they won't remember that. We are the only Looney Tunes who would remember that. Um whereas your average consumer will not. So yeah, your homepage should generally have the highest amount of direct traffic. Um the next thing you want us to keep an eye on is that not set traffic.

Where is that traffic going? So I'm gonna go from direct here and we're gonna change our instead of our um session default channel, we're going to go with source medium this time. Exactly matches. And again, same thing. We want to see is there uh completely unknown traffic, and where is it going?

Um here, um, you will okay if you've got problems with bots, bots very often come in as not set. Um, and so you will see weird page spikes. We had one client uh last year who one of their product pages just kept getting hammered by bots, and uh it turns out there's actually a bot attempting to do fraudulent transactions on their site and it was targeting a very specific page. So it was relatively easy to uncover with this this analysis. Hmm.

Yeah, it's delightful, isn't it? Okay. The next thing uh you'd want to look at just from a uh let's clear that um is real simple. So you have your page title, and you can also use page URL if you want uh page location, which depending on your site structure uh can be easier or harder. Just clear that filter.

And now you're looking at sessions by page. Uh maybe add you know a few more rows here. This is not, I wouldn't call it content attribution, but this is just what content is top performing on your website, like what pages are getting any traffic at all. And if really important pages are not in like your top 10, then at the very least, your site may have an internal navigation problem, or you may not be you know uh nurturing people the right way. But we here we see here the contact us page is number eight.

So this the site's is going well, but you you want to just take a real quick look. What is getting attention? Why are you using page title and screen class and not just the metric or the not just the dimension page title? Like what does that tell you different? Uh well you page title itself doesn't exist.

I think it does. Well, he's not using those tags, is that what you mean? Oh, yeah, page title. Yeah, we could so page on screen class. If you have a site that has a mobile app, um, your screen class will be part of that.

So if you have a blended site where you are using um GA4 from for the web and for like iOS and Android. Uh the page page title and screen class will bring in the naming pages of both your mobile app and your website together in one unified view. Gotcha. But page title does exist. It does.

It didn't used to be there. Um next. Uh notice next. Oh, um for sort for looking at your pages and things. The other thing that is sometimes helpful to do is you can do filtering.

Let's go ahead and put our channel groupings back in and look at the channels that typically matter to marketers. Um so for example, I might want to look at organic search. What pages on my website get the most search traffic? So uh are there pages on here that are doing really well? I might want to look at what are there pages that are that do really well from our email marketing efforts, right?

What pages might those be? Email's not here this week, okay. That's fine. Um look at organic uh social instead. Um what pages are getting social media traffic.

And this is again, this is just basic inventory. What pages are getting traffic from where and does it match up with your efforts? Right. So if you have been logging uh a page, maybe like you have a restaurant client and they flog in this one specific drink on your on your menu, and you're you know, you've put on Instagram every two hours, and there's just no traffic to it, like then you know that tactic is just not working. Okay.

Next. Um, so yeah, it's a good idea to go through and just evaluate uh the pages of getting traffic by those different dimensions. You can construct you know bigger reports, but just in terms of a basic audit, just to dive in quick. This is a simple way to do it. Um let's go ahead and go back to our dimensions.

And we want to look at page referrals next. How traffic is being referred to us. Let's add that in as so. And this is telling us where is our referring traffic coming from? So we see uh a big blank spot.

We see Google in here. We have some self-referral traffic because I think we need to change the configuration of the site for its own domain. Uh we have Facebook, we have DuckDuckGo, Baidu, and so on and so forth. And we could see we can go all the way through these uh this chart and see all the different places that referring traffic is coming from. This is a really good thing to do to see where your traffic is coming from, particularly for unexpected uh traffic sources.

So let's see, let's add a few more rows here. What if you see uh so you're looking at Boston Martial Arts Center, but you're also seeing in refersh in referral traffic boston marshalarts.com. Like so, should you be seeing a lot of the website itself showing up in the referral traffic? You should not. That's that's one of the things they're saying.

That shows this uh configuration issue in Google Analytics. For the property details, we actually want to make sure that the property is excluded from itself, um, which Google Analytics does very well. Uh we've actually had issues with this with a different analytics system recently. Um, and it's not it's not doesn't handle it as well as Google does. Okay.

Um let's uh we'll tackle that later. All right, let's go back to our explorer. So for some of those other places, if there's a if there's a known referral that you want to exclude, you can obviously filter that out. So in fact, let's do that. Let's do our page refer.

Does not contain. Just clear that up. So these are the essentially the external sites. Um if this was my site, I'd be saying, okay, well, are there any is there anything in here that could be a potential partner of some kind? Let's expand our timeline here to go back to maybe September, just so we can get a bit more data to work with.

Okay. So there's like a this appears to be some other uh martial arts website. There's an ESPN for some reason sending traffic and things. Uh oh, yeah, there's an ESPM story there. So there's there's opportunities in here if you have a good PR team or a PR agency or an uh an SEO team to look at where you're getting your traffic from and see if there's an opportunity to get more links or more prominent links uh as well.

And you can you can tell that just from looking at where you're you're referring traffic is coming from. Um next up is conversions, and this goes back to uh we'll go back to admin, and where we looked at our events, we're now gonna look at our conversions and see uh what's in there. Now we just made some changes, so this is all gonna show zeros for right now because we of those changes we just made. Remember that with Google Analytics, unfortunately, nothing is retroactive ever. Um, so if you've made changes, um, for example, to events and conversions, uh, you will not see that until new data comes in.

And it will just be kind of messy. So those are the those are the basic places that we look. Those are the basic places that we look uh in any Google Analytics as a starting point just to judge the health. Because ultimately, you can mess around with a lot of the configuration, and there are absolutely our best practices for every single section of this site in different settings you want to check. But you want to look at the data first to see what's in there.

Because if you don't do that, then you don't know how bad it is. Uh, and whether for the if you're when you're talking to the site owner, like can you even use the existing data or not? Um in this case, this site is in reasonably healthy condition for understanding where its traffic came from. It is not healthy for understanding conversions. Um, you could not do use any of the attribution modeling right now.

And I think, you know, and I think that that's a really good uh call out because these configuration is so much more, it feels so much more complex than universal analytics, that I think it really does make sense to just sort of see what data is coming in, especially if you've inherited the configuration from a different consultant, a different team member who's no longer there. You want to make sure that you're getting the right data out of it first. And I think that it's a really good opportunity to bring up the 5P framework, which can sort of help you put together those requirements. So the five P's are purpose, people, process, platform, and performance. Purpose, what is the question we're trying to answer?

People, who are the people involved, who's affected by this process, how are we approaching the task? Platform, what are the tools we need to use and performance? Did we answer the question? In this instance, you know, we could start with the purpose is we want to do an audit of Google Analytics for to understand can we get conversion data? Or is it uh reporting on the basic metrics correctly?

Who are the people involved? Well, the people are the person who is the admin and probably the person who is the owner, the business owner, so that they could be aware. Um what is the process? Chris just went through the process in terms of what reports to start with. Platform, Google Analytics, makes sense, and then performance.

Did we answer the question? In this instance, we did. We answered the question and we were able to see the things that specifically need to be tuned up in that um profile in Google Analytics for. And so it's a really good, I wouldn't call it a shortcut, but it's a more user-friendly way to approach auditing the system if you're not comfortable with the configuration itself. Exactly.

Now, one of the things that we do just for my own sanity, um, is we we have built all this into automated reporting stuff. So the first time whenever we onboard a new client uh with their Google Antics account, we just plug straight into their the API and say extract all this information out. Because as you can see, just going through that last set of um very quick reports took almost 30 minutes, um, which to gather the data, not even to make judgments about it, just to gather the data, you know, clicking and and building explorations and stuff. That that's not a good use of time. Um, it's better to expull the data out yourself with you know programmatically, and then you can make those judgments very quickly um just to judge the health of the the instance without having to uh to manually do that, or even just open up a pre-existing notebooks and you know, change change which accounts is tied to, et cetera.

I have two questions on that. One is so you were just in the explore hub, or I guess now it's called Explore that I didn't realize they changed your name. Um why if I just built those reports in Explore Hub, why can't I just keep going back to those reports? Okay, one. But I in my understanding is that once you build the report in Explore Hub, you can't change the date on them.

Like once it's built, it's a built report unless you keep modifying it. Um, this was a issue that we ran into when we were building this for one of our clients. Is that if you build a report and save it, you can't then go back into that report and change the date unless you rebuild the report again. It's a little clunky. So is that something that they've resolved?

Yeah, you could change the date on the report and change the reporting windows, like that. It doesn't migrate by itself, like you have there are some basic rolling windows that you can set last 28, last 30, etc. Do the comparisons, but they have made explorations a little easier to use when it first came out two and a half years ago. Yeah, it kind of sucked. Gotcha.

Um, so my second question is why would we build this in Explore and not Look or Studio and then have the reports just there to look at to monitor the health of our Google Analytics for instance at any given time? I studiously avoid Looker Studio for anything GA4 anymore because there are query limits. Um what happens is if you have an uh a workbook that has you know a half dozen visualizations and stuff pulling from different tables and stuff, you burn through those query limits really quick. Anyone who's you ever used a Google Looker Studio dashboard with GA4 for any amount of time, you inevitably get the uh you have reached the quota limit for this visualization and all your charts just error out. Uh when you're using explorations, that doesn't count against the query limit.

So you can use this as much as you want. Uh, this is especially important if you are um if you're using this with a a client and they are they have production Looker Studio reports and you're doing diagnostics, you do not want to blow away their query limits and then their production reports stop working. Um I feel like there's also sampling issues with the data that comes into, I may be saying this incorrectly, but my understanding is that to bring the data from Google Analytics 4 into Looker Studio, you have to you know connect the data systems. And this may have been true a while ago. There was something about the data connector API that didn't bring all of the data.

And so the question we got a lot was why is the data in Google Analytics different from the data in Looker Studio? So Looker Studio uses the GA4 data API. Um there it's not just sampling, there's there's limitations and restrictions on what you're allowed to export um from certain fields, uh, particularly in anything that involves um their their ad data, because they don't want to provide personal identifiable information and things, but there are sampling, there are some sampling limits uh in the API. The easiest way to get raw data um is not through Google Analytics, uh uh amusingly. If you integrate BigQuery, the BigQuery integration in GA4, you will get individual hit level data within BigQuery that you can then do raw data analysis on with the caveat with the caveat that the data in BigQuery will not match one for one with what's in Google Analytics because the the BigQuery schema is Firebase's schema, and so the definitions of some metrics like sessions are different.

So that makes me wonder if I'm and I am this person, if I'm the person who's looking at the data, and if it's different in three different places, if it's something in Google Analytics Explorer, if it's something different in BigQuery and something different in Looker, which one's right? Don't know. Um and it sounds like an absolutely you know, ratchet crazy thing to say, but it the thing that we always tell clients is you need to have a a further down the funnel source of truth, and then essentially do a regression analysis to see which of your different data sources are are closest to that number. Um, for example, if you use a system like Cloudflare, Cloudflare will give you web analytics as well. Oh, that's ironic.

Um and Cloudflare's numbers will not match be even close to GA4. Um so you you run a regression against an important thing like leads generated or something, and then you can say, okay, which numbers are you know are are closest to that? Where's the you know what's getting filtered, where's it getting filtered, etc. And the thing that we usually counsel our clients is it's not you don't need the exact numbers to make a decision. This kind of data is meant to be directional and to tell you, you know, is the campaign working?

Um to your point, Chris, make sure there are other data sources further down the funnel that are your source of truth. But systems like Google Analytics should really be used as directional in terms of making your quote unquote data-driven decisions. Um every single system has its own quirks and flaws. And so, unless you are really diligent and hand counting and triple checking everything, getting to that exact number with all of the infrastructures we have out there feels nearly impossible. And that's not meant to discourage you, it's just meant for you to rethink your perspective on how you're using the data to make decisions.

It's a hundred percent impossible. Um but that's my point is that you're trying to you're not using the data to get down to exact numbers, you're using the data to directionally point you in, you know, where you need to go with your decision. Exactly. I mean, these devices uh with the ever since iOS what iOS 14 or I yeah, I think it was 14. Um, there's additional uh blocking of certain types of trackers and certain types of information.

Uh the more recent versions of iOS blocked some additional stuff. Uh there's proxy blockers, there's VPNs, there's ad blockers. If you use Ghostry, for example, Ghostry blocks Tag Manager. Um, so it would it will restrict uh that data collection. Um, there's also dark traffic, you will get a fair amount of dark traffic, which is traffic where you don't know where it came from.

One of the more advanced um visualizations that we've done in the past for people is to do again a regression analysis of their direct traffic, which doesn't have that attribution data, uh, versus known sources of traffic to see which traffic source your direct traffic is likely most composed of. Uh, and there is such a thing as dark search. Um, for example, the people going through search engines like DuckDuckGo or going through uh even more restrictive search engines. That traffic may or may not show up with any kind of attribution. So yeah, you were you were using this as a guidance tool, and you're using this tool as as a very broad top of funnel tool.

There's going to be a whole bunch of stuff that Google Analytics just can't measure either, um, that are important but just totally absent. Uh things like branded searches. Uh, you need to use Google Search Console for that because that is data that Google sees on their servers. It can't account for traffic from things like text messages, um, text messaging clients. Uh, if depending on the level of security you have in certain apps, it will just come in as as unknown traffic.

There's dark social traffic, which is also fun. Uh so yeah, there's there's any number of places where you have information that you haven't no idea where it came from, and there's no way for Google to get that information. Additionally, there are some there's some weirdness with identity resolution. Um this is a this is a more advanced topic, but it's one that's come up a lot in the analytics for marketers Slack group, where choosing which reporting identity can change your numbers because if Google doesn't have data uh for certain types of uh identities, it it sort of drops that from the data, and so you'll get less accurate reporting. You'll get more thorough data on the people you have information on, but you'll get less overall data, which which kind of sucks.

And I think that that's been one of the real frustrations with systems like Google Analytics is the industry as a whole has been really working towards being data-driven and using data to make decisions, and now here we come saying you can't do that. Surprise. Yeah, I was gonna say one other thing I wanted to dig in a little bit more on was so when this originally came out, we were kind of sold on the idea that looker would be the front end for this, and so you've seen now after some field time that it's not worth it for a lot of our clients because it can't handle the data load. How about what we had also talked about where data is getting dumped to BigQuery and Looker is pulling that data? Is that worth doing, or do you find that Explorer has got enough that most clients are happy with that?

Or where does that line get drawn? And is it the does that work? Exploration is is usable now, right? So it's it's certainly usable, and it's it's good enough for a lot of basic client tasks. For Looker Studio is really good for things that where you want to have almost like a canned report.

Um, we have this setup with one of our clients where it it just emails a PDF, right? So here's here's your monthly report. Um it's it's good in limited doses uh because again, because the API limits if you upgrade to um the Google Analytics 360, you get 10x the query limit for you know tens of thousands of dollars a year um for your install. But I would say, yeah, if you if you have like simple like one-page dashboards, Looker Studio is is perfectly fine. If you were talking, you know, workbooks with 10 or 15 or 20 tabs where you're gonna have a bunch of people banging on that workbook and um then yeah, you you want people in exploration instead, just because they'll hit query limits real fast.

Um I kept looking off the side because I was following along as you were doing it, because I admittedly I'm not the Google Analytics for expert. I felt fairly comfortable with universal analytics, and I think a lot of marketers did, but Google Analytics for I find to be very unapproachable and intimidating in terms of the UI. So I was following along as you were building the report in the reports in the Explore Hub, and I feel a little bit more confident because I could never get them to work, and I still struggle with Looker Studio uh sometimes. And so, you know, I was able to, while you were doing that, diagnose one of our clients that we need to sort of go back into and fix up their property now that their website is a little bit more robust. And I, you know, that to me, like in the course of those, you know, 20 or 30 minutes you're doing that.

I'm like, oh, and now I have a whole to-do list of things that are gonna make a huge difference right away, including seeing how some of the meta descriptions on the web pages themselves, the naming conventions are a little bit off. And so it's it gave me a lot of information really quickly without having to get into all of the settings. Yeah. One thing to be careful of with page titles, if you have pages with duplicate page title names, they get lumped together. So that's one of the reasons why you may also may want to use page location for some of the reports, just so you can see the individual URLs.

Makes sense. So I guess the next best step is for people to embrace Google Analytics 4. You can, you know, join our free Slack group and you can complain about it to your heart's content. But the reality is that it's here. If you're using Google Analytics, you have to use it.

You have to figure out a way to get along with it. Um, if you want some additional training, we do have a Google Analytics for course that you can find at TrustInsights.ai slash GA4 course. If you want us to just do it for you, which a lot of people do, you can find us at TrustInsights.ai slash contact. You know, Chris, I give you a lot of credit for really diving in and embracing Google Analytics for when they started announcing it and learning it. Um, you're one of the few people that I feel confident really has a mastery of it.

And even then, you know, we found out there's things that they're rolling out all the time that you weren't even aware of. You know, and so I want to encourage the people who are watching this to not be discouraged if you don't have a mastery of Google Analytics for it, it's hard for all of us. Yeah, they keep moving things. Like literally, like, oh, that's in a different place this week. That's that's great.

Um, the other thing, and I think it's it's worth reiterating this, it is a useful system when it's properly configured, but it is it should never be your system of record, right? It should never be your source of truth because it's not close enough to business outcomes. Your marketing automation systems closer, your CRM is closer. And to the extent that you can bring you Google Analytics data into those systems, you can build things like better attribution models uh using data in those systems as opposed to in GA4. GA4 is good for figuring out what happens up you know from the time someone gets to a digital property you own to the time somebody does some kind of conversion event on that digital property.

If your sales process goes outside of that, then it's only part of the picture, right? So, for example, when you fill out a form on the Trust Insights website, that is the end of Google Analytics role, right? It says, okay, form was filled out, great, my job is done. That is not the sale, right? That is not when you you you book us as a consulting firm.

A lot happens after that. And so for us, our system of uh record is the CRM to say, like, okay, this deal actually closed. Um in the course, and obviously within a bunch of other places, um, we'll talk about how you amortize the value of your actual sales down to conversions that Google Analytics can see, but it is not the source of truth. And if it doesn't cooperate uh or you know, things just don't work for uh for you, there isn't a lot of harm anymore in relying on what's in your marketing automation system or what's in your CRM, because you know, the closer you get to the bottom of the funnel, the closer it should be to revenue anyway. And the the ugly dirty secret of attribution modeling, as much as we love it and we love how fancy it can get and how sophisticated it can get, is that there's there isn't and there shouldn't be huge percentage differences between you know the most sophisticated machine learning model and last touch, right?

If there's a wild difference between the two, then something's broken in your tracking somewhere. Um your last touch attribution in a web analytics system is going to often going to be okay-ish enough. Right? And it's okay enough to at least know, okay, something worked, or hey, look, no conversions happened at all from this thing. I guess that that didn't really work, did it?

Right? Mm-hmm. John, final thoughts. That's it. Come jump on board.

We'd be happy to give you the full tour of this stuff. So uh we have a lot of clients that have help and get stuck and just need eyes on to get them there. And I think the big one, what's always missed is you can look at this stuff snapshot at any time, but the key is to have your UTM process locked down, follow the uh procedures that we talked about and check on that weekly or monthly, you know, so you're you're actually tracking the data. Don't presume that you can just throw that link doc out there once and it's gonna get done because we have bad news for you. Yeah, humans, humans are imprecise.

So all right, folks, that's gonna do it for this week's show. We will talk to you all next time. Thanks for watching today. Be sure to subscribe to our show wherever you're watching it. For more resources and to learn more, check out the Trust Insights Podcast at Trustinsights.ai slash TI podcast.

At our weekly email newsletter at trustinsights.ai slash newsletter. Got questions about what you saw in today's episode? Join our free analytics for marketers Slack group at trustinsights.ai/slash analytics for marketers. See you next time.