Measured Direction Podcast

Download the Episode

In Episode 1, we discuss, perhaps unsurprisingly, getting started in Digital Analytics as well as developing a Testing program and the joys of Tag Management.

Show Links:
Web Analytics 2.0 by Avinash Kaushik
Successful Analytics by Brian Clifton
Digital Analytics Association
#measure Slack
Learn Python!
Learn R!

Transcript

Jason Rose:
Hey, guys, welcome to the first episode of Measured Direction, as you can tell from Tom laughing the background, this is taking me a couple takes, but I feel like this is really the one that sticks. So what this is, is Measure Direction is a digital analytics and marketing strategy podcast produced on myself. Jason Rose, a content strategist here, and our leader of our analytics practice, Tom Miller. I don't know if you saw our kind of kick off pilot episode video, but this podcast is really about just addressing analytics questions and any kind of marketing strategy that questions that we have. So this first episode, we're going to address internal questions that we have here at digital surgeons. But in the future, we really want you guys to write it and let us know what you want to hear about in the future. I guess we just start with the first question telling you anything you want to add or introduce yourself or whatever is out there.

Tom Miller:
I'm our Elix Practice lead here, digital surgeons.

Tom Miller:
We are setting this podcast up to answer practitioner level questions related to digital analytics and marketing effectiveness. Really marketing strategy focused on digital channels.

Jason Rose:
And we let's get started with the questions. It sounds great. So I guess the first question we had come in is, you know, someone who is just getting started on digital analytics. What kind of recommendations can you give them for learning discipline? And just a programming language or they need to execute it.

Tom Miller:
Ok. Wow. That's a great question. So we've solicit these questions from internal keep folks here at DHS, digital surgeons. Yeah. So this is a I mean, this is like a perfect kickoff question for this episode. So, you know, I'm going to give sort of an insight into how I got started with Dejoy Lytic. And we'll take it from. Now, I think hands on experience is the best I learned being a hands on practitioner.

Tom Miller:
You know, one of the easiest things that you can do is to find analytics software. And fortunately for us, that one of the most popular and most widely used pieces of clickstream analytics software is Google Analytics, and that is free. So what you can do is if you've a blog or if you have a Web site that you might have access to that is getting some traffic. Go ahead and install new analytics. There's there's really no risk to the current function of a Web site to install Google Analytics. Now, getting to the point where you're installing it might require a little bit of. Technical learning, a little bit of access to things that you might not have previously had access to. But that's OK. You know, learning the what we call the implementation side of an analytics tool is important. And it does. You know, you can't really understand the data that's being collected until you understand the mechanisms by which it's being collected. So makes sense. Yeah, kind of. So, you know, I think you can get pretty far with either getting access to a Google Analytics profile and doing some analysis on that profile, be it on your own blog, on a site that you might have some access to, et cetera.

Jason Rose:
So really, before getting super deep in the theory of it, your number one recommendation is just create a Google analytics profile, go on there, do the certification, play around with it, set it up for whatever kind of personal site you may have.

Tom Miller:
Yeah. You start getting your hands dirty your way in front with the certification.

Tom Miller:
I would recommend getting a, you know, properly implemented Google Analytics profile set up and going into the interface and just getting a feel for the types of data that's being collected. You know, Google does a really great job of producing a whole mess of content related to not just its tool in its tool set, but this idea of modern digital marketing strategy. In fact, I think Google is probably driven the ball forward on that discipline, perhaps more than any other company. So they are you know, they're very transparent. They win when people are executing on digital marketing because a lot of their digital marketing budget flows into Google. That's that's sort of Google's angle with with the whole marketing suite that they've produced. So, you know, that's sort of step one. You know, I think is there there's opportunity to then extend this technical knowledge ento starting to get to understand, you know, the business context for why Google Analytics is a widely used tool. And one great way to do that is that there's an organization called the Analysis Exchange. And what that organization does is it pairs students of Dejoy Analytics. So people that are sort of new to the field are just starting out in the field with mentors and nonprofits that are seeking some help with their digital analytics.

Tom Miller:
So I've been a mentor with this organization for a march or maybe I mean several years, maybe 10 years. I don't know if that's possible for years. And I've probably worked on seven or eight of these projects. And what you do is you basically work on a very discrete question with a student. The student creates in some cases identifies a problem, creates some analysis around that problem, and then creates a presentation that is presented to the lead stakeholder of whoever the nonprofit is. It's a it's a really great little mentoring program. You know, it's it's limited in scope. So if you're if you're working a day job, you know, you're talking about a project scope that is in the low tens of hours. I mean, like 10 hours, you know, 20 hours at the most over the course of several weeks. So anybody should have the time to be able to do it if they're willing to invest a little bit of time in their own DEJOY analytics education. And, you know, you're also helping out a worthy organization. Right? An organization that is helping other people out by their mission.

Jason Rose:
This sounds great. So where would where would someone go again to find that and get involved?

Tom Miller:
Sure. Sure. So it's run by a an analytics consultancy called Web Analytics Demystified. And if you just Google Analytics Exchange, it'll come up and it looks strange. Yep. And I believe the URL is something like web analytics demystified. Slash a call. Yeah.

Jason Rose:
Already, you know, books for failure, maybe plughole. They should put pickup apps. Absolutely.

Tom Miller:
You know, I think the the great the best darter book is still Web Analytics 2.0 by Evan Oshkosh. Yeah, I think it's the must read introductory book for Digital Analytics. He really has a relatively. Vendor neutral breakdown of a lot of various aspects of Web analytics. So, I mean, he's he's an excellent digital marketer. He's a blogger, prolific blogger. But this book that he wrote, you know, at least five or six years ago, it might be a little bit dated with the tools, but the underlying theory is sound and will be sound for the foreseeable future. And he goes over things like clickstream voice, a customer, competitive analysis, really the whole spectrum of digital analytics, you know, sort of intro practitioner level digital analytics, you know, really, really excellent book. If you are focused on Google Analytics, I'd also recommend Brian Clifton's book Successful Analytics. It covers a lot of those sort of driving business value questions, a lot of those organizational and process oriented questions, which sort of as you get into the discipline of Dejoy analytics, you realize that the discipline is really about business process, aligning teams and not really so much about the tools. That's really the hardest part about digital analytics.

Jason Rose:
Do you think that's a problem, that a lot of practitioners become too obsessed and they kind of lose sight of that, that the general underlying themes and theories are why they're doing what they're doing here? I think a lot of organizations become tool obsessed and issues themselves are kind of fighting against that.

Tom Miller:
Yeah, I mean, I think it's it's sort of, you know, since since the discipline has been created, it's sort of been an ongoing discussion. Right. It's it's tools versus people in process. And, you know, the consensus among the practitioners set is that it's it's really 10 percent tools and 90 percent people in process. And I tend to agree with that. I mean, I think that these things that marketing analytics need to be successful and to flourish within the organization, organizational alignment stakeholder by an executive sponsorship, you know, team ownership. These things are much, much more difficult to pull off than a simple, you know, even if you you whatever. Right. Even if even a super complicated implementation of a clickstream tool, you know, that that's a problem that you can throw dollars at. Right at the time at the it the organizational alignment actually takes you know, it takes it it takes a different type of currency. It takes a social currency to to be able to understand and manipulate and align to effectiveness. And, you know, in some organizations that it's purely a bigger organizations that gets to be really, really challenging. So I'm sure that this will be an important topic of this podcast.

Jason Rose:
Before we get back to me asking you about programming languages and finishing up the rest, the question I want to apologize to any listeners that are hearing some background noise and the construction going on. We're not actually drilling while we're doing this, but we're fortunate enough that we're expanding our offices next door. So that's the occasional hammering or drilling that you may be hearing. So back in answering this question. Sure. Sure. So let's let's keep got so.

Tom Miller:
So what else about learning discipline? There is a trade organization related to Dejoy Analytics called the Dejoy Analytics Association. Conveniently enough, you know, I certainly think that they are a pretty valuable gateway to a community, a gateway to learning resources, you know, with a.D.A membership. I remember you get a thought leader conversations, you get discounts to conferences, you get discounts to graduate degree programs which are which have been set up to support support the discipline so that, you know, you can get a graduate degree from several universities in the USA and in Canada in digital analytics or digital intelligence or what have you. The community is is also how you are going to advance and accelerate your career. The digital analytics community is an open community. It is a nurturing community right now. I would say that the best way to gain access to the community is through the measure. Slack Measure Slack is at Bitly b t that ally slash ad measure slack. If you don't know what slack is. I don't know. I'd recommend you Google it. We use it at D. S. You know the. I've got three Slack's that I'm. Love and its I guess some people would speculate that it's going to replace e-mail someday. I don't think that that will happen, but it is and I guess I'd call it an enterprise grade instant messenger.

Tom Miller:
It's really like you can almost think of it as like a Facebook for work or it becomes a central communication hub that everyone goes into and they're building other integrations in there. So everyone kind of more or less goes to slack first and then ends up being able to get done whatever they need to communicate how they want to. How do you like it? I really like it so far. I mean, I'm not going to drop the messenger service here using before we switch to slack. But Selek has been a nice change and I feel like it's increased productivity. I don't I feel like the channels are relatively good at keeping communication out in the open and keeping kind of backroom or private stuff at a minimum. How about you enjoy it so far? Oh, yes, sure. It's great.

Jason Rose:
Is this your favorite? It's. Where are you using slack then before we implemented it here with you. I love. Yes.

Tom Miller:
Yes. It's. It's great. I don't know. It's it's it. I'm amused and bemused by it. So. So the other you know, another great resource for somebody starting out is hashtag measure on Twitter. That's sort of another place where the community puts a lot of content. You'll see a lot of the same articles getting. Tweeted multiple times, you know, a lot of classic articles get tweeted from time to time. Eh, it's great.

Tom Miller:
I mean, you can learn you can learn a lot, a whole lot about this discipline from the blogging community. So I know that's sort of how I would get started is read a lot, you know, read some books, read a whole lot of blog posts and get to a point with a tool, you know, a freely accessible tool where you're you're able to actually perform some basic understanding of what's going on with the Web site and then get to a point where you're doing some basic analysis with a Web site.

Tom Miller:
Go ahead and do whatever hands on things that you can do and look within your own career for opportunities to extend what you're doing with some intelligence that's based on Web usage or the people that are using your Web site programming languages. Oh, man, I don't know if I'd love these two questions together, but that's how we got them. So, you know, the tags for a click stream analytics tool and a lot of, well, every marketing tag and platform that's being used online is is driven by JavaScript, by JavaScript tags. When you hear about tags, you're actually hearing about snippets of JavaScript that are executing within a browser environment. You can learn JavaScript. I mean, JavaScript is a programming language if you're interested in. Getting more to using a scripted language for data analysis. You know, I'm I'm a developer on a python developer. I mean, if I had to choose one to say, learn it, learn python, because it has nearly universal applicability to any other types of programming tasks that you might want to endeavour to do. You know, both are in Python are easy to set up in any environment. And they're generally free to learn. So the the documentation and the community support is certainly there for both are in Python. You can find a question and answer to any question. I've not been able to not find a question related to ah or to Python on stack overflow.

Tom Miller:
So, you know, there's there are literally millions of people using both languages and they're running into the exact same problems that you might be if you're working at any type of enterprise. Were you using a database and you're doing data analysis? Well, you're probably you don't want to learn, Eskew. All right. What Eskew Al does is allows you to pool queries. Well, it allows you to create queries to pull data out of a database. You know, it's extremely helpful, extremely useful or a requisite language to know for certain roles within organizations. And, you know, a lot of database driven marketing and really, you know, operations or, you know, there's a there's a multitude of uses for Eskew out now. You know, in some cases, you might be a layer removed from that. Right. You might have a developer writing applications that is pulling data out of a database and presenting it. And the mechanism by which it's pooled is as well. But you don't really get visibility into that because you're using an application that's pulling the data for you. Right. But it is extremely useful. It's it takes a while to to really fully understand what you're doing with escudo.

Tom Miller:
And, you know, my recommendation for anybody learning Escudo is to also do do some reading on set theory and do some exercises that are absent, you know, programming language exercises that really just directly relate to set theory. And what's that theory does? Is it really helps you fully understand certain operations. And Eskew like Selex and joins and. Pretty much everything else. So I would I would certainly say that learning ask you out does open up a lot of different types of roles to somebody that's doing digital marketing. You know, it's just a question of where do you want to work and what role do you want to work? Right. B.I tools are sort of the same thing. If you want to work in certain types, organizations, typically larger enterprises that have a B.I tool as their basis for their business intelligency. Well, you could learn how to develop on that tool and develop reports that come out of a data warehouse or some other, you know, large data source within that organization.

Tom Miller:
So I, I feel like I've talked a lot about this, but yeah, I mean, I kind of feel there's a natural kind of segway into our next question.

Jason Rose:
So let's let's question let's assume I've I've gotten my hands dirty and Google Analytics, separate web analytics, successful analytics are on the slack. Learn the programming language is awesome. So what are the biggest mistakes that I can now make when I'm really starting to develop my first testing program and really getting into it?

Tom Miller:
Oh, boy. Okay. So testing program. What are the biggest mistakes that you make when you first when you're developing a testing program? Well, you know, the number one mistake is not really fully understanding the business context for the interfaces that you're testing. So there most certainly is a. A ranking or a a different valuation for different interfaces across your Web presence, across your Web site where you want to be testing is in your maximally impactful. Key interactions that are going to have the most the most impact on your ultimate conversion business, bottom line, that is what you're referring to really bad.

Jason Rose:
Of course.

Jason Rose:
So if you're an e-commerce site, you know, it's obviously revenue and it's e-commerce revenue. So, you know, how do you get to. How do you get to that? Because I think one of the other biggest problems is you install a shiny fancy testing tool and first thing you do is you pick a page and, you know, you start changing all the button colors or something like that.

Tom Miller:
Right. And. That's sort of the opposite approach. Tell me what what you should be doing. I've done this here and I've done this. Other organizations is really developing a structure and a framework for prioritizing your tests. And so, you know, the way that I think about it is I think about the business impact. Right. Ranked on some scale, multiplied by the ease of implementation for any given task. So easy tests that have a low impact might have a higher priority off of really hard to implement tests that have a moderate impact. Right. So you know that those are the two basic dimensions. I blow it up to five or six different dimensions. But you know that that's a basic idea. And then what you can do is you can say, OK, well, here's here's my time. Here's I going to allocate my time. Here is the amount of traffic coming to this interaction. So then you can sort of back into how long you want your tests to run. And what does the testing cadence look like for any given test subject? And you know what? What do I expect to get at it? What's the business impact? Whereby raising the floor. Right. I see. Testing as an exercise of raising the performance floor. Not necessarily raising the performance ceiling in most cases, but it's it's really, you know, improving the baseline performance of your Web site.

Jason Rose:
So what you say about raising the floor versus raising the ceiling, doing a little more detail, what exactly you mean by that? Sure.

Tom Miller:
So, you know, my theory on testing is that what you're doing is that you're you're eliminating a lot of friction. So you're limiting friction to your key outcome. Right. To your positive outcome for both you and your site user. And when you're eliminating that friction, you're not necessarily creating a higher high for somebody that is motivated to do something on your site. Right. There's there's this concept and it's sort of a it's a it's sort of a really interesting way of thinking of your site users. But like some of your site user, you think if you're an e-commerce site. Right. Some of your site users are. There's a spectrum of their interest in actually purchasing from you. And a big chunk of the people that are coming to your site are hell bent on buying a specific product from you in that session.

Jason Rose:
And sight testing might not necessarily. Change that in 10.

Tom Miller:
But if they're running into problems, problems with the user experience, problems with the actual function of the site, that is preventing them from clearly being able to say, click by now. Finish the cart. Submit the cart. That will help. Right. But you're not really creating more of those people with testing. Another segment of your site might might be shopping around and they might be interested. Right. So making sure that they have a you know, a a clear path to conversion is hugely important. But other things might be important, like how you're presenting your value proposition or how your product images is even laid out or things like that. I mean, you know, again, you don't really know until you start testing what what is more effective and what isn't for any given interface. And then, you know, another side of your users might not be interested at all. They might be previous purchasers. They might be people that are coming to your site just for product description, information or reviews. And they have absolutely no intent to buy. Right. And so.

Jason Rose:
Ap tests in that case is not going to mean anything because they weren't going to. You're wasting your time trying to appeal to them by.

Tom Miller:
Well, maybe you are maybe, you know, maybe something about your value proposition actually gets them to convert. Right. So so that that additional level of segmentation. And by the way, that segmentation, you can't it's it would be extremely difficult to get to that on clickstream alone. Right. So, you know, you're talking about an A.B. testing program that is informed by a voice of customer program. Right. And, you know, you're sort of you know, that X percent of the people that come to your site fall into these different purchase intent categories. And then you can sort of extrapolate your baby testing program based on that. And you can attempt to assign them to a category based on some of their behavior. But in a lot of cases, you're talking about.

Jason Rose:
Right. So the best educated guess you can.

Tom Miller:
So correct. So. Yeah. And, you know, all of that is, you know, again, achievable with current technology. You know, not specific vendor technology. Right. I mean, there are many vendors that will help you achieve that type of insight. But that's sort of where where you're thinking should be with testing is, you know, what is this overall business context and how do we how do we really raise the bar performance for our Web site?

Jason Rose:
All right. So that's it. That's all we got for this week, right? Well, you stuck around this long. Thank you very much. You know, I'm going to promote our form for submitting questions. You know, as of this episode, we are open to questions from the general public. So the the forum where you can submit your question to us is Bitly b.i.

Tom Miller:
T. L. Y slash measured direction. That's the name of this podcast. I'm Tom Miller. Practice, Lead and Digital Search. And you can follow me on Twitter at T and L. L. R. S. T Miller. Without any vowels. T. L. L. R..

Jason Rose:
And I'm Jason Rose. Content strategies. You can find me. J t rose. That's j a y t. Rose my Twitter handle. If you wanted to be those questions, use the hashtag measure direction. I'll be sure to find it on there.

Tom Miller:
Oh, awesome. Yeah, that's a great idea. As we do, our podcast was produced by Jason Rowe sitting here and mixed by Adam Chambers. That's at C, H, A and B, A Z Chabon's.

Jason Rose:
He went about his his half of a duo called Kamon sound really cool. Local New Haven rap duo. Definitely. Check them out.

Tom Miller:
All right. Well, thank you very much. And good luck measuring.