The Unofficial Shopify Podcast: Entrepreneur Tales

Your Data is Probably Bad

Episode Summary

"No data is better than bad data."

Episode Notes

Everyone likes to say they make data-driven decisions, fewer actually do, and even fewer are making those decisions with good data.

As more and more Shopify businesses drill into making data driven decisions having the proper foundation is imperative.

I have Rich Hannah from RHUX analytics here. One of our mutual clients has been working with him and was really impressed.

Hot takes:

Prior to starting his own agency, Rich was the former lead analytics engineer for lululemon and consulted for brands like Nordstrom, Major League Soccer, Zendesk, Starbucks, Tesla and others.

Episode Transcription

The Unofficial Shopify Podcast
8/24/2021

Kurt Elster: Everyone likes to say they make data-driven decisions, don’t they? But few actually do, I suspect, and even fewer of those are making decisions with good data. But as more and more Shopify businesses drill into making data-driven decisions, having the proper foundation is imperative. You can’t make data-driven decisions with bad data. You need it. You need good data. And the advantage of data-driven decisions is that you can run a leaner, more profitable business when you have clear insight into what does and doesn’t work, and that’s the point of all of this.

Sound Board: Dollar dollar bills, y’all.

Kurt Elster: So, to help us think through that, I got Rich Hanna from RHUX Analytics here. And one of our mutual clients has been working with him and was really impressed, sang his praises from the rooftops, and so we’re joined by Rich Hanna. And prior to starting his agency, listen to this. He was the former lead analytics engineer for Lululemon and has consulted for brands, stop me if you’ve heard these. Nordstrom, Major League Soccer, Zendesk, Starbucks, Tesla… Ooh, my favorite. And others. Wow.

Mr. Hanna, oh my gosh, you have quite the resume here.

Rich Hanna: Thanks, Kurt. Yeah, I’m happy to be here.

Kurt Elster: Well, yeah, thanks for joining us. So, I touched on your background there, but in your own words, what’s your background, or rather why should we listen to you?

Rich Hanna: Actually, backing up, at a former life I worked for a company that consulted for those brands that you had mentioned. You know, Tesla, GoPro, a bunch of others.

Kurt Elster: Oh, I didn’t even mention GoPro, but I like them too.

Rich Hanna: Right. You know, so working with that company, I got to learn just best practices across the industry, just what these giant billion-dollar enterprise brands are doing and how they implement analytics, and how even they mess up, right? Even they are confused on things and even they have data collection issues, and reporting issues, and just everything from like, “We don’t know why this is tracking and why this is broken,” or how to even go about doing this, or like what’s the strategy.

Kurt Elster: God, that makes me feel so much better. Like every… I have these exact same issues with eight-figure brands, and it feels like the bigger the brand gets, the harder it gets to get good data, clean data, actionable data, or maybe it’s just that it becomes more important. But I’m always like I second guess all of it, I don’t know what is and isn’t real. It’s hard to tell and it drives… I’m like, “What? Am I doing any of this right?” It’s just like imposter syndrome and second guess… It's horrible.

Rich Hanna: Right. Yeah. Well, and the people that do it every day, like these giant industries or these giant enterprise brands have entire departments that are dedicated to this kind of stuff, you know? They have a team of five to 50 analysts and a team of five to 10 implementation engineers that are doing this stuff, and even they have difficulties with it. Everybody’s got a different opinion as far as how it should be done. And I think the unique perspective that I was able to bring from that was just the difference in how everybody ran their company, you know, and just seeing kind of like their best practices, and I was able to kind of pull from each one of those and just say, “Okay, these are the ones that I’ve noticed that work across the board.”

I think I was pretty fortunate to be in that position to be able to see how that works. And then from there, I was able to go to Lululemon. I had a great opportunity with them to be a lead A-B testing engineer for the company for a few years, and then I transitioned into the lead analytics engineer for them to help implement everything from Google Analytics to Adobe Analytics to app analytics and things like that. And then here I am, working for my own company now.

Kurt Elster: Well, A, congrats. And B, I didn’t know that you were also a split testing engineer at one point, because that I could do an entire nother show on with you, as well. But you mentioned Google Analytics. Okay, let’s go back to basics with Shopify. I’m setting up a Shopify store. It’s not a joke. I literally am setting up a Shopify store for someone as we speak. And I’ve got to add the Google Analytics snippet, and in Shopify you go online store preferences, paste in the snippet, check enhanced eCommerce data, click save. Have I already gone wrong here?

Rich Hanna: Yes.

Kurt Elster: No! Why?

Rich Hanna: So, the reason being is that just how it sets up the data model just doesn’t make sense. I mean, it may make sense to a Shopify dev or whoever actually was the product manager for that particular product in Shopify. I think that the data model itself does not follow the Google enhanced eCommerce just schema or best practices. There’s kind of an industry standard as far as just you set up your enhanced eCommerce with like the different tagging setups and things like that, and we’ll get into that for those that don’t know what tagging is, like we’ll chat about that, but I would say that if you’re wanting more than just your basic data collection, and you’re wanting clean, accurate data that’s actually actionable, that you can pull reports from and actually know what you’re looking at, then no, you don’t want to do that.

In the case of a store owner that is just kind of getting into it and they’re just starting out, I think it’s totally fine to do that. So, I guess to answer your question, kind of backtracking on myself a bit, I think that it just… It depends on the situation of are you a company that’s doing a million dollars a week in online revenue? Are you somebody that’s just starting out brand new to build your own store and your context of you’re just filling in a store, starting from the ground up? It's fine. You’re gonna have to redo it later if you do end up scaling the business. So, just kind of keep that in mind.

We always recommend kind of going with GTM out of the box and setting up basic tracking with best practices as far as that goes, but yeah, that’s kind of our thoughts on it.

Kurt Elster: All right, so when we… With that native Google Analytics integration, is my data bad? What’s wrong with it?

Rich Hanna: So, it’s collecting data and it’s… The way it puts it into the model for analysis for like your top event reports doesn’t make sense in the aspect of like if you go through and you try to do any sort of product level reporting, or like detail to cart ratio, or anything like that, it gets challenging with how that current setup is configured. If you go to checkout, and we have tools that we’re able to look at the events that occur on somebody’s site, if you go to checkout there’s like three or four different events that are firing in a single checkout step when all of that can be consolidated into one when it’s actually falling in the Google-recommended schema, right?

So, it will send one for the product, one for the checkout level, and it just… It’s creating duplicate events, which if you become a bigger store, GA will actually have sampling rules that will set in place and they’ll cap you at two million hits a month, which we have seen from stores out there. And so, the more events that you’re sending to GA, then you’ll actually start losing data the bigger you get, so there is some drawbacks to it if you’re a bigger company.

Like I said, as a smaller one, your odds are you’re kind of just starting to dip your toe in things like that, so you’re not gonna be doing the level of analysis that will be… you’ll start to notice those little inefficiencies that are occurring.

Kurt Elster: Okay. So, what should I do instead?

Rich Hanna: So, we recommend that you start off with using GTM. If you have a Shopify store, you’ll have it installed directly in the theme. If you have a Shopify Plus store, it’ll be in the theme and the checkout file. But typically, what you’ll do is you’ll install GTM. You’ll have your analytics are running all your core tracking, like your core page views, your core settings, which are capturing all the data that you need on an every hit level, so different things like the client ID, or the Shopify user ID. That way you can stitch together users in GA and Shopify, right? Or different things like that, right?

So, I think that we always recommend GTM, just because it’s more scalable. It is more challenging, and it is a bit of a learning curve, and there’s… You don’t want to just dive in and start throwing a bunch of stuff in there, but there’s plenty of great vendors out there that are willing to help, and I think that it’s a good start to just start off on the right foot. And if you need basic level tracking anyways, it’s actually not that difficult to implement. And you can get good, basic, enhanced eCommerce tracking setup within just a few days, really.

Kurt Elster: So, GTM, you’re referring to Google Tag Manager.

Rich Hanna: Correct.

Kurt Elster: Okay.

Rich Hanna: Yeah.

Kurt Elster: You said, “Oh, you can set it up in a few days.” A few days? I can copy and paste one snippet for regular Google Analytics. What am I getting out of GTM that makes it worth the hassle here?

Rich Hanna: Yeah, so I say a few days to set the expectation longer. It really doesn’t take that long if you are able to find a good way to do it. If you have a vendor do it, it should take like a day, right? If a good vendor is doing it, they’ll QA it, so it’ll take two days. But if you’re implementing it yourself, there’s all sorts of good resources out there, Google documentation that will enable you to set it up. You will probably need a dev or to be able to implement what are called GTM templates, which are built into Google Tag Manager themselves that you should be able to set up enhanced eCommerce tracking.

There’s also a few apps out there that will help you integrate this, as well, which is an option.

Kurt Elster: What’s your preference? App or vendor? Or DIY?

Rich Hanna: So, vendor is definitely my preference, just because you can get it going, but if you want to get it any… and I’m speaking to the kind of the newbies here, you can get it going. You can get it set up with the core tracking and things like that. If you want to go anything beyond that, then you’re gonna need a vendor. And then you’ll have to start worrying about like personally identifiable information and things like that, right? And GDPR, which is like a global data protection regulation over in Europe, right?

So, there is just… It’s a big world and it’s a black box for a lot of people, so going back to your initial question of like yes, you can check the box and it will collect data, and that’s kind of the point of if you want to collect data just right out of the box, that’s totally fine, but if you want to do good analysis, you’re gonna have to set it up through GTM. So, it’s kind of a pick which path you want to take at the beginning.

If you’re a bigger vendor that’s more established, you have maybe a small… Or sorry, not a bigger vendor, a bigger business that’s more established and you have maybe a small team of devs, or a marketing team, or something like that, I would definitely recommend going with the vendor route and working with someone that’s gonna be able to set up clean, accurate data collection. We have this saying that no data is better than bad data. And the reason why is that there’s… You know, you’re just more likely to put all your eggs in the data basket, right? And if you come back later and you find out that that data is bad, which more often than not it actually is, then you kind of question the business decisions that you’ve been making, putting all your decisions in that basket.

Kurt Elster: Oh, no. I totally agree with that. That was kind of the crux of the intro, is like hold on, if we’re making data-driven decisions, is anyone questioning the veracity of the data? That’s the issue we get into.

Rich Hanna: Right.

Kurt Elster: You’d mentioned for installing GTM, you mentioned apps are an option. Any you recommend?

Rich Hanna: Yeah, so they’re actually a competitor of ours, but Elevar has a great app. They are able to install some tagging directly into the site. They’ve got a great model over there. There are a few that you can use that are… and I don’t recommend them, per se. I guess like in the aspect of each one of them that we’ve seen, outside of Elevar, which is a good… They’re a vendor and they have an app, right? You’re kind of on your own with a lot of them. And so, you’re kind of running into your same issues.

So, if you start having data collection issues, you’re kind of on your own with that and then you have to kind of find your way in the dark with those, as well. So, I guess my recommendation would probably start and end there.

Kurt Elster: Okay. And truthfully, of the GTM data layer apps, the only one I’ve used is Elevar. I’ve seen it used several times and it’s been consistently recommended. So, if you’re going the app approach, and really this is just one piece of the puzzle, this is adding the code, the data layer into the Shopify theme. But then you still have to set it up in GTM, et cetera. So yeah, no, I’d agree with that.

You mentioned GDPR. If I’m using GTM with Shopify, and so I’ve got this non-native setup, do I have to worry about people’s data getting transferred in some elicit manner that’s gonna get me in trouble?

Rich Hanna: Yeah, so I’ll say this with the caveat that I’m not a lawyer, but the-

Kurt Elster: Good news. Neither am I.

Rich Hanna: So, the idea of GDPR, and CCPA, and it’s going to be more prevalent here in the United States over the next few years, we’ll start seeing laws passed throughout these different states, is that there are companies, or excuse me, regulations for companies that are put in place to ensure that the user has privacy, right? That there’s privacy rules around data collection. So, if you are currently doing business in Europe, then you will need to essentially put in place what are called consent triggering rules that are essentially there to make sure that when a user… If a user doesn’t accept the cookies, like everybody’s been seeing these, “Do you accept cookie,” popups over the last year or so, right? On every single website they visit.

So, if there’s a cookie consent popup and the user says, “No, don’t track my information or change my cookie settings,” then there needs to be something in place that is actually blocking the ability to fire those tags and triggers to Google Analytics, and Facebook, and things like that. So, yes, that is something you will need to set up directly in GTM to ensure that there are consent rules being followed.

Kurt Elster: Okay. How terrified should I be?

Rich Hanna: I don’t think terrified right now. I mean, and this is me kind of going back to not being a lawyer, but you know, I think that as of right now, everyone in the United States is kind of getting on it. I think the more prevalent it is here in the United States, having cookie consent rules and things like that will be more and more common, and we’ll need to have things in place like that. I know that at my previous company, when CCPA dropped, there was a pretty big scramble to make sure that there was different elements that were on the app itself that were needed to put in place to be able to ensure that this California Consumer Protection Act was actually being followed.

There are fines that can be imposed. I don’t know off the top of my head what those fines are, but those can be put directly on the business for collecting data. If it’s being sent to Google and Google sees that there is a violation of GDPR or CCPA, they will delete the data and could delete your account.

Kurt Elster: Geez. All right, so a little scary. As long as we’re on the topic of things that scare or should scare merchants, another one that always seems to terrify people is, “My PageSpeed scores, oh no!” With the level of data you have and the analysis you’ve done, have you looked at the impact of PageSpeed? Like real world, what do you think?

Rich Hanna: Yeah. So, I think PageSpeed matters, and I think that if there is a significant detriment to the page load and the user experience, then absolutely it matters. The user will leave if it takes more than a few seconds for a page to load. You know, we’re talking six, seven, eight, nine, 10 seconds, right?

Kurt Elster: Well, at that point it’s like it’s not like… “Oh, I got a PageSpeed score issue. My site’s slow.” No, it’s like your site’s straight up broken if it takes eight seconds to load.

Rich Hanna: Yep. Exactly. And so, I think that there-

Kurt Elster: Eight seconds, it’s like let me just turn my phone off.

Rich Hanna: Right. So, in that instance, yes. Absolutely. The PageSpeed matters. If we’re talking like the difference between 4.5 and 5 seconds, there’s data that suggests that there actually is a lift in revenue and conversion rate, but there’s also data that suggests that there’s no such benefit to it, right? And so, I think what that shows us is that there’s like that it can help, but it’s… I think there’s conflicting data right now and my personal experience, at Lulu we were able to increase our PageSpeed score by I think it was about three quarters of a second, and there was a small increment, because just over the mass volume of people, like it naturally… Your page is faster, so it’s going to-

Kurt Elster: This is in reference to Lululemon.

Rich Hanna: To Lululemon. Correct. Yes.

Kurt Elster: Yeah, so no one listening has anywhere near the traffic or revenue of Lululemon. Just straight up.

Rich Hanna: Right. And so, in that case, there has been… There was a lift, right? But I believe that if it’s a second or less, you’re kind of just splitting hairs at that point. I don’t know that it’s gonna make a huge lift in revenue. Potentially 4% to 5% as far as conversion rate lift goes, that’s as high as I’ve seen in the real world outside of Lululemon. You know, but that’s just our particular dataset that we’re looking at, as well.

Kurt Elster: Interesting. It sounds like inconclusive, or I think charitably a fast site doesn’t hurt, a slow site doesn’t help is where we’re gonna land on. As long as we’re operating within like you can actually load it.

Rich Hanna: Right. Exactly. Now, there are some benefits to… Because I think what we’re talking about here specifically is probably like a headless site, right? And I think-

Kurt Elster: You brought up the word.

Rich Hanna: There it is. So, I believe that with a… There are benefits to a headless site in that you don’t have to… Your initial PageSpeed score is only going to matter for the page load, where the PageSpeed score doesn’t actually indicate the savings is on each subsequent page load after that, right? Because a headless site is built on what’s called a progressive web app, or a single page application, and those are typically built in React, or Angular, or something like that. React and Angular really actually only loads the page once. And so, if you look at the top of your browser and the little refresh icon that’s happening on your tab as the page loads, that only happens once on landing and then every subsequent page load after that will actually… your page never refreshes, right?

So, in theory, you’re not loading all those additional libraries. You’re not loading all of those… You’re loading GTM once. You’re not loading GTM every single time you go to a new page like a static site, right? And so, there is cost savings, or I guess speed savings to be had on subsequent page loads. So, the PageSpeed score isn’t a true indicator of what actually happens directly on a user journey. So, we’re talking four to five pages deep. Whereas on the initial load, there’s relatively no difference in that aspect.

Kurt Elster: It is so refreshing to hear someone just give us the straight facts on web performance. It really is.

Rich Hanna: Does that not happen a lot? Or is it kind of like a dance around just like headless in general?

Kurt Elster: I think a lot of people just don’t understand the deep technical aspects of it and I think it often comes down to like, “Hey, Google says there’s a score that if my score gets better, my site does better and I get more free traffic,” so it really doesn’t matter what anybody else says, because Google says so.

Rich Hanna: Got it.

Kurt Elster: And it’s so frustrating.

Rich Hanna: Got it.

Kurt Elster: It’s just like you’re just wasting resources and tearing your hair out over something that is way less impactful than you think.

Rich Hanna: Sure.

Kurt Elster: You mentioned the H word. If you go headless, what data considerations do I have here?

Rich Hanna: Yeah, so great question, so this is something that we’ve seen. We’ve had the opportunity to work directly with some headless companies out there and actually Shogun is one of our partners, and we’ve had the opportunity to be able to work with them, and some of the challenges that we’ve seen not just with them, but with a few other headless sites that we have worked with is that with the improved performance between pages and site speed and user experience, I guess upgrade if you will, the challenge in migrating over to headless is that you need an entirely new analytics implementation.

That also, your Shopify out of the box does not actually translate, right? So, you will need to set up GTM. You will need to set up different just analytics data collection tools and things like that, so you’ll have to set up Facebook, you’ll have to set up Google Analytics, you’ll have to set up pretty much any third party, Klaviyo, all that will have to be set up. And it has to be set up differently. The data layer needs to be restructured. The data layer is what sits there to collect all the data, to send it off to these different third-party vendors. The data layer will need to be restructured because how a React or single page application or progressive web app handles just data collection.

So, your conversion rate may have some issues, like if you… Artificial issues, I should say, because what’ll happen is as you’re navigating across the site, or when you land, if you come in from Google, for instance, like a Google ad, CPC, and you land on the site, and then you navigate to a different page, there’s a chance that the session behind the scenes will break. And so-

Kurt Elster: Oh, geez.

Rich Hanna: What that will look like according to Google Analytics is that there’s two sessions, right? And we’ve seen upwards of 13 sessions for one user when in actuality it should have been one session, so now Google Analytics takes that with the conversion rate and says, “Okay, this is a… It took 13 sessions for this one user to convert.” And so, it’s like but that’s not true, so your conversion rate is actually better than what it would be showing in Google Analytics, which is a good thing, right?

But it just has to be accounted for on the setup. There is something that you have to set up to be able to do a rogue referrer, and that is essentially-

Kurt Elster: Rogue? R-O-G-U-E?

Rich Hanna: R-O-G-U-E. Yeah. And so, that’s like-

Kurt Elster: Sounds badass.

Rich Hanna: Maintaining that original referrer. So, the rogue referrer issue, once that’s fixed your conversion rate returns to normal, right? So, that said, working with Shogun to develop an out of the box solution that enables their clients to be able to just move seamlessly over to Shogun, and we’re working with a couple other companies with this, as well, to be able to have a one-to-one integration based on what they had in Shopify.

So, if you’re gonna go with a different company other than Shogun, or something like that, then that’s something you’ll have to think about. So, there’s benefits to switching to headless as far as the user experience goes. There’s also things you want to think about, and I guess in summary would be Shogun is developing this solution to be able to not have to worry about your analytics as you are rolling over.

Kurt Elster: Potentially, ideally in the future with Shogun, my analytics don’t break, and is that a GTM-based solution? Or straight Google Analytics 4?

Rich Hanna: Yeah, so that’s a can of worms right there.

Kurt Elster: I realize. I’m like, “Oh, boy. I opened something here.”

Rich Hanna: Yeah, so yes, so currently, there’s something that we have with them that’s a… It’s a GTM solution. It has GA4 capabilities in it. It also has server-side capabilities in it because there’s server-side data collection that’s occurring now, and this is in response to ITP and all of the Apple changes. And so, you can actually collect up to 99% of your data now, as opposed to missing out on a lot of the Safari tracking issues and privacy browsers and things like that, which his a whole nother story.

But that’s Google’s response to a lot of what’s going on with getting rid of cookies and all that kind of stuff, right? So, there is a solution we have in place right now to be able to roll over. There’s a more advanced solution that’s coming down the pipe that’s called Headless Analytics that we’ve developed. It’s a proprietary software. It’s actually one of the first in the world that is piggybacking off of a technology called Module Federation.

Kurt Elster: You’re just making up words at this point. You’re like, “All right, the rogue module installs into the headless server side.” Anyway, everything is better after that.

Rich Hanna: Right. Exactly. Yeah. So, anyways, so that’s coming potentially. There’s still some kinks we have to work out. But there are… Yeah, so there’s something that will be coming down the road for that. As far as GTM, GA4, server side, all of that kind of stuff, that out of the box is pretty much ready to go with Shogun.

Kurt Elster: Cool. And we’re gonna do… We’re gonna get some Shogun customers on here to talk us through headless, and I wasn’t seeking out Shogun, I just like… I said, you know, I gotta find some people who are headless, and they were Shogun frontend customers. That’s just how it works out. It seems to be the market leader in this instance.

So, you mentioned server-side analytics, and iOS 14, and this privacy thing. So, iOS 14 is very… Apple in iOS 14 has this big privacy push and as part of that, on my iPhone, I can opt out of a lot of tracking. It limits a lot of tracking. And most tracking is based on session IDs and cookies, meaning my device helps advertisers keep track of me. And when you say server side, it means the server providing the info, so on the other side of the web, not my device, is gonna do the tracking.

I’m shocked that that’s not how we always did it. Because it would run faster, it would be more efficient. The server side has total control of it. It’s not reliant on goofball JavaScript. Talk me through… Well, just give me some thoughts on server-side analytics.

Rich Hanna: Yeah, so server-side analytics, it really is… I’m surprised it wasn’t kind of the how everything was set up in the past, as well. I think just a natural progression of Google bought this company called Urchin, and Urchin had UTM parameters, and like that’s where it’s like literally UTM source medium comes from, is Urchin tracking module is where that comes from, right?

So, I think they had this model, and it was able to be collected on the web, and somebody decided that the best way to do this was cookies, and so everything else just latched onto that. From there, obviously Apple was like smacked that idea and said, “No, we’re gonna-“

Kurt Elster: Many years later.

Rich Hanna: Many years later. Yep. And that we’re not gonna allow that, right? Which some businesses are like, “Well, is that your decision?” There’s that whole big argument that’s kind of going on right now, right? And so, the idea of server-side analytics, you’ve already called out a few of the benefits, but it just… It puts the control back in the business’s hands if they want it, right? And takes it out of the browser’s hands.

And so, what can be done is that you’ll be sending data from your website to this server that’s set up, and it’s actually relatively easy to provision. There’s just an out of the box solution that Google has that you go, “I want to set up a server-side container.” And then you click a button that says provision server-side instance, and then it sets it all up for you, right?

And there’s still some small configuration that needs to be done to kind of scale it if you’re a bigger business, but the simplicity of it is really, really good. So, what it does is that you can now reduce the amount of tags and reduce the amount of network hits that are being sent from your web container to your server-side container, and then your server-side container can then take that, like let’s say in this instance like a single event that’s sent, so like an add to cart event or something. It takes that add to cart event and now you can broadcast it out to multiple different locations. You can fire multiple different tags, right?

All of your logic and all of your heavy lifting can live on your server, and you can have a much lighter container that lives on your site. And as a lot of people know, you were talking about PageSpeed scores, one of the things you’ll see that’s kind of the problem child is Analytics. And the Analytics load time to load in a big tag management system that’s taking up kind of the main thread, dev talk there, is… It takes that and makes it smaller, so it doesn’t have to take up as much space. It’s not like forcing everybody else out of the way to essentially get there and load first.

And as everybody always tells you, and Google even tells you, put this all the way up in the head of your site so it loads first, right? And so, now you’re loading GTM but you’re also loading Facebook, you’re also loading Klaviyo, you’re also loading all these other tags, and third-party libraries up at the top of your site, which is why it’s the problem child. And so-

Kurt Elster: Yeah. Really, I think unknowingly, when people talk about, “Oh, my PageSpeed score is bad. Oh, my site’s slow.” The culprit is almost always third-party JavaScript. You’re loading all this JavaScript from external sources on the web and that slows things down dramatically. And if you get rid of all that JavaScript, the site loads way faster and your PageSpeed score gets better. But ‘lo and behold, no one wants to give up their analytics. No one wants to give up their remarketing snippets. And all these goofball apps.

Yeah, and truly, part of the reason headless is often faster is because you no longer have access to a million crazy apps at your fingertips to install, all of which is like, “Hey, I’m gonna install three different versions of jQuery and then leave these here for you to deal with later.”

Rich Hanna: Right. Yeah. Exactly. Yeah. And I think that the benefit of having server side is that you’re not necessarily firing a bunch of network hits. Like if you look at the network tab, if you have a server-side instance set up, you are sending far fewer, right? Depending upon each setup is different, so like I can’t throw a number out there, right? But it’s… You could theoretically take 10 different hits for a single event down to actually just one network hit going out.

And to kind of clarify what a network hit is, that’s when you’re actually reaching out and trying to send data across the internet, right? And so, it can take it down from 10 to one, and so when you do that, that’s what actually helps speed your site up and things like that as well, too.

But on top of just the performance impact, you’re also able to, and this is kind of some points of contention of like if and just like should it be done, but you’re able to take it out of like the privacy browser’s hands of blocking Google Analytics and blocking Facebook Analytics and things like that. Because you can actually send it in your own URL, which those ones are actually looking for, right? They’re looking for a specific structure in that URL that’s sending to Google Analytics and sending to Facebook, so you’re not… Since you’re not sending it to those anymore, you’re sending it to your own, you can actually bypass the ad blockers and the privacy browsers and things like that if you choose, which gives you all of your data.

Since there’s all these things out there now that block these particular third-party vendors, there are… You’re gonna be missing data, like we see it all the time that you don’t collect all of your data because of those things in place.

Kurt Elster: All right. Actually, I like that. I think that’s a good place to wrap it up. But all right, where do we want to leave it? What piece of misinformation do you want to correct? What do you want, what action do you want people to take? Where do you want this to go? What’s your legacy on The Unofficial Shopify Podcast?

Rich Hanna: Don’t feel like… I guess this huge pit in your stomach if your data is bad or if you find that your data is bad, or you think your data’s bad. It’s fixable. Almost everybody’s data is bad just starting off, you know, and so there’s no one out of the box solution that fits all websites, right? And so, it’s gonna be Google Analytics doesn’t know how to handle that. If you go to GA and you look at your pages, odds are if you’re a relatively large company you have over 600,000 URLs, unique URLs in your pages report.

And so, what that does is that causes fragmentation, which is like it spreads all of your data across all of those different rows, and so you can’t analyze that because it’s like if you want to know what your top performing page is or something like that, you can’t necessarily tell because it’s actually spread out across 500,000 rows. So, Google doesn’t know how to handle that out of the box.

Kurt Elster: I don’t know how to handle it either.

Rich Hanna: So, you have to do things to make sure that your data is clean, and that your data is fixed, like we see bots all the time, bot traffic in there, so putting filters in place and things like that. So, don’t beat yourself up if you feel like you have bad data. It is fixable.

Kurt Elster: It sounds to me like you gotta make do with it right up until you’re like, “All right, it’s time to hire a professional to just tackle this for me.”

Rich Hanna: Yeah, and that’s up to you to decide when that is, right? If you want to start making business decisions based on that data, then hire the professional. If you’re just kind of like wanting to know like, “Oh, okay, how are we doing?” Then you can make do with the out of the box implementation and just watching how many users you have come to your site, how many sessions, and things like that. But you know, I think that through all of that, with hiring somebody like a vendor is they come along with best practices, you know? Setting up these best practices. And then from there, it’s strategy, and a lot of people think best practices and strategy are the same thing, and they’re not. Best practices are just like what happens across the industry and then strategy is unique to your business, you know?

And so, a good vendor will provide both of those to you and be able to help you be able to grow your business based off those data-driven benchmarks and metrics that you have.

Kurt Elster: And let’s say I wanted to hire Rich Hanna as my vendor to put these best practices into place. Where do I find you?

Rich Hanna: Yeah, so you’ll go to our website, RHUXAnalytics.com. You can see us on there. You can find us on LinkedIn. Pretty much all your favorite social media outlets.

Kurt Elster: All right, and I will link to that in the show notes. Rich, this has been quite educational and a little bit freeing. A lot of good info in here. I appreciate it. Thank you so much for doing this and if people have questions, they should check out RHUXAnalytics.com. You even have a pretty good blog, I noticed.

Rich Hanna: Thanks, man. That’s a shoutout to you, Brenda.