Power Up Campaign ROI with Data Analytics

Joining host Greg Oberst on this episode of The Whole Idea Podcast by DCG ONE is Kristen Crandall, director of analytics at The Agency at DCG ONE. Learn how to use data analytics to bring unprecedented visibility to your campaign and drive ROI throughout the customer journey. Get insight into how to integrate data into your next campaign, hear best practices and tools to leverage along the way, and learn how to avoid mistakes that could compromise your data and campaign.  

Other links you may like to check out:

About us - https://www.dcgone.com/about
Strategy - https://www.dcgone.com/strategy
Technology - https://www.dcgone.com/technology
The Agency - https://www.dcgone.com/agency
Let's connect! https://www.dcgone.com/contact

Email us: podcast@dcgone.com

Check us out on social media:
LinkedIN, Instagram, Facebook

Transcript

Copy Transcript Copied

Air date: April 17, 2024 

Transcript: The Whole Idea Podcast 

Episode:  Power up campaign ROI with data analytics 

Greg: 

Hello everyone and welcome to the Whole Idea Podcast. I'm your host, Greg: Oberst, senior writer at the agency at DCG ONE. For this episode, I'm joined by Kristin Crandall, director of analytics at DCG ONE. Welcome Kristen. 

Kristen:  

Hello. Thank you. 

Greg:  

We're going to tap into your expertise on data analytics. Of course, data analytics isn't a new practice at its core, it's been around, but as a capability boy, it sure has evolved. Let's talk about that evolution in the last 20 years or so. 

Kristen:  

Sure. So really, I mean really we started evolving data analytics with the invention of computers probably, I mean is longer than 20 years ago. Sure. But that's kind of where we started to be able to look at bigger sets of data and make predictions now starting really like in the early aughts when we got into big data is where kind of what I do started to happen. So there's actually this great book that came out 2010, 2012, uh, by Nate Silver called The Signal in the Noise. That's all about using this huge amount of data to distill it down and find trends and make predictions and do tests. Um, highly recommend that book if anybody wants to nerd out about it. Uh, but that's kind of where we started getting on track with what we're really chasing after today, which is just trying to make predictions about what's gonna happen. 

Greg: 

But now it's easier and faster. 

Kristen:  

Yes, definitely. And it's changing really, really fast even in just the last five years or the last year with AI that's coming up for us a lot more machine learning capabilities, it's a lot easier to query a database. You used to have to have pretty robust SQL knowledge, which isn't terribly difficult, but now there's tons of no SQL options for getting data. So it definitely is getting easier and easier and faster and faster for those of us that are in it every day. 

Greg: 

You're spending less time on the logistics, call it manual part of, of data analytics and thinking more about the analytics.  

Kristen:  

Yeah, the takeaway, right. I mean that's definitely the goal. We still spend a lot of time, even if it's faster to get the data, we have to spend actually a little bit more time now in some, some instances making sure that it's clean data and that there's not a bunch of noise in it. And so there's still quite a bit of time that happens at the front end. We're not quite in a place yet where we just get to think <laugh>, uh, about the analytics part. But certainly the most fun part is actually spending time once that data's clean.  

Greg:  

Yeah. Thinking about the analytics still have to do the work, it's just that the work is different now and approached a little bit differently. 

Kristen: 

One thing that has been happening lately is more accessibility to stitching together a true customer journey. So clients have been asking us to do this, been asking me to do this for as long as I've been an analyst, which is longer than I'm going to admit. And the data has been so historically siloed where you wouldn't be able to stitch what was happening on your website to your Google ads. Your email was a completely different team, uh, that was attributed to nothing. Your direct mail, you just hoped for brand awareness and really didn't know. And your CRM is just a completely inaccessible black box. Now we have API connections and other ways to connect that data together where we can actually deliver on that ask. That's been out there for over a decade of putting together this full customer journey. So you can see how all of these pieces interact together to either ultimately convert or not.  

Greg:  

That's interesting. So not only is, uh, technology assisting us from a speed standpoint, some of that speed is coming from this idea that technology is doing a better job talking to each other. 

Kristen:  

Exactly. And so I don't have to be a really, uh, advanced engineer to stitch this data together. I can be an analyst and stitch this data together. 

Greg:  

Let's move on to the, the parts of data analytics. How would you unpack the different segments that fall into your work? 

Kristen:  

So really I think the best way to think about it is when we approach any data analytics project. I'm a little bit old school and so I just go back to the scientific method. So the first thing that we do is say, what is the goal here? Or what question are we trying to answer? That seems simple, but a lot of times that's the hardest part is having a very clearly defined what's our purpose here? And then after that you wanna go into the research phase, which is really critical. This is where we're looking for an external truth that can either validate we're, you know, working on something that's meaningful or that can sort of say like, whatever prediction you've made here is biased or too optimistic and there's nothing that supports like that's gonna turn out well. Sometimes we have to say, we're on the wrong path here, we have to go back to step one. 

Kristen:  

So it's really, really critical to do that research step, which I think it's easy sometimes for us to skip past because it takes time, but it's really important. And then after that, we're gonna go ahead and implement whatever we need to collect the proper data. So in our world, a lot of times that's putting a GA four script on a website or making sure that we're pulling in some like paid media data, whatever it might be. It's making sure that we have all the tools in place to collect the data. Then we give it some time to collect the data. That's another really important step. A lot of times like we'll implement something and maybe the next day the customers are like, Kristen:, how's it going? Like, it's gonna take a little bit more time for us to be able to say anything for sure. 

Kristen: 

So you just kind of kind of let it do its thing for a little bit. And then we get to do the fun part, which is actually analyze. So that's taking a statistically significant set of data and just seeing what it tells us. For me, I think visualization is really important at this step. Um, it makes things more digestible for me and then for my clients, internal or external, having something where you can like obviously see a trend or an insight is really important. So that's kind of what happens at this step. And then you're gonna report on it, learn from it, and adjust. So a lot of times in this phase we'll see that maybe a prediction that we had early on proved to be false or we'll see that something we had is is not performing and we wanna scrap it and try something new. So this is where we just make adjustments and then we go right back into that cycle and start over and over and over again forever.  

Greg:  

Where Woes that data come from? 

Kristen: 

It can be anywhere. It can be from going out and doing interviews with, you know, whoever your kind of target customer is and finding out what they wanna see. It can be from looking at sort of demographic type information. Maybe what you're trying to do as is a retention effort. Then it's about talking to your current customers and finding out what uh, would convince them to stay. And sometimes it's things like looking at like research papers or just kind of whatever's out there for you. Almost like you're gonna, you know, write a paper, you need to get all your sources in place. It totally depends on where you're starting. Um, but a lot of times it's just conversations with the people that you're trying to reach. 

Greg:  

That would be, uh, qualitative research. 

Kristen:  

That's correct. 

Greg:  

As opposed to quantitative research. Correct. Why don't you talk about the difference between those two <laugh>.  

Kristen:  

Yeah. So, it's really just kind of about the amount of information that you have. I think that's a good way to distill it down. So obviously if you're talking about quantitative, you are measuring a certain number of items. You're measuring a number of clicks on your website, you're measuring the number of people that are looking at your Instagram ad. There's some like quantity. It's right there in the word of what you're looking at. Qualitative is still kind of statistical thing to do, but it's a little bit more subjective in that you're really going for more of a back and forth and trying to find a little bit softer trends that will later inform how you're setting yourself up to quantify that data. I mean, the questions that an interviewer might be asking is structured in a qualitative world, but the answers that are coming back are not just gonna be ones and zeros. humanity behind it.  

Greg:  

They've got a little bit more – it might be opinions. Alright. Let's talk about marketing specifically, and the types of data analytics that are most relevant in our world in marketing. 

Kristen:  

Yeah. For us it's all about predictive analytics and more and more using that in like full customer journey analytics. This has been really, really fun in the last year where we have a lot of customers coming to us and we're able to look at their paid media data offline and online. So we're looking at, you know, paid search in addition to direct mail or email. Then tying that directly into what's happening on the website and then seeing even further down funnel. So are they making a phone call from the website? Are they filling out a form fill all the way in some cases to like truly converting and becoming a client or a customer and then staying a client and customer. And it's so you can see like one tiny change that you make. You can see the ripple effect all the way down through conversion and it's so cool <laugh>. So we're doing that more and more and I think there's more and more clients that are, were able to do that in a way that we weren't even two years ago. 

Greg:  

Yeah. Predictive data analysis brings business and marketing insight that makes for more informed decision making Of course. Sure. But also, boy, it really creates effective targeting opportunities as well, right? 

Kristen:  

Yeah. And it's, it's, I think that's exciting too. And I mean the business insights portion still is really, really important. That still needs to be there as a foundation. Sure. But yeah, you can really target customers exactly where they are and speak to them in their language. And it's good for the customers too. I mean, I know it's getting better, but in the last couple years there's been this kind of like, oh my gosh, you know, my ads are so targeted. It's like Big brother. Um, but it's also just efficient if you're seeing things that you're looking for anyway, uh, or things that really speak to you. And so it allows our customers to speak to their customers very, very directly about what they're already showing signals of wanting to do. Like in the travel example, if I see an Instagram ad or an email with like a beautiful cabin in the wilderness, there's a 99% chance that I'm gonna click on that. And they know that about me and I appreciate that they know that about me. 'cause I'm probably not gonna click on a beach ad, which I almost never see. Let's 

Greg:  

Talk about the data on data analytics. <laugh>. I'm thinking about the cost of, say, executing a marketing plan without using data. Does that turn into a larger expense because now you're kind of shooting in the dark a little bit and maybe wasting money? 

Kristen:  

Absolutely. A really important part of any data analytics program is making sure that you're calculating ROI. Uh, and that includes on us as analysts <laugh>, we are as accountable as a media team, making sure that your media return on ad spend just showing up positively. Uh, without that you really have no idea and it's a pretty, uh, slippery slope in terms of just spending and thinking. Maybe the traffic you're getting is from a certain campaign or, uh, it's just risky to not have that in place. Sometimes customers or just anybody out in the market can become fearful that analytics is too difficult and too complicated and cost too much and takes too much time and just backs away from it. And kind of just what we were talking about in the ROI conversation, that's a really risky path to go down to not look at what's happening. Like it might seem scary to say, but what if I look at it and I have a negative ROI that might happen, but it's much better to look at it and adjust than not look at it at all. 

Greg:  

So that seems like a pretty big first mistake, not using data analytics at all. What are some other mistakes you've seen? 

Kristen:  

The second mistake that I see a lot is people wanting to make a decision without enough data. So just because, you know, maybe especially in an AB test, so say you're running an AB test and uh, option A has 30 clicks and option B has 28 clicks, and you might go, okay, I'm definitely going with option A, but there's just not nearly enough information to make that decision. Um, you really have to let analytics take its time to get enough information and be patient and also realize that at the end there might not be a clear answer and you can't force one if it's not there. And then the third thing that I see customers or anybody again do is just be too blinded by their own bias to trust the data. So even if the data back to the AB test is saying, for sure option A is better and you do have a significantly a statistically significant data set, you still might just be so convinced that customers wanna see the color blue instead of the color purple, that you don't believe the data. 

Kristen:  

So trust the data if you're going to use it. For example, we have a customer that is looking to test video placement on their website. Their prediction is that the video being higher on the website will both increase engagement with the video and increase engagement on the page in general. So when the client first came to us, they were like ready to test five different things at once. My job is to make sure we're following the scientific method we were talking about a minute ago. I said, okay, let's start with just one simple test, which is first we need to put the video on the page and test how much interaction it's getting at all. And if just having a video on the page impacts engagement with the page. And so we're doing an AB test where half of the website visitors will see the video and half won't. And so we're establishing a baseline of does having a video at all make a difference? Once we have that, we can move it around and see if that impacts people engaging with the video in the page in general. Kristen:, 

Greg: 

So what advice would you have for brands who are looking to ramp up a better data analytics practice? 

Kristen: 

The most important thing to do is to just start doing it. Start at the first place where you have easy access to your data. A lot of times when I'm guiding a customer through this, we'll start with just a website. Um, Google Analytics as frustrated as we get with it sometimes actually makes it fairly easy to start tracking web engagement. So just start there or just start with your Google ads and just look at the data. Once you start looking at that data, you can then think of the next step and add that on very incrementally. Um, but if you never start, you'll never get there. And once you do start, it's not. Um, I mean it definitely takes like some analytic thought for sure, but it's not wildly beautiful mind on a chalkboard. Scary to do. It can be done and you can always call me if you have questions. 

Greg:  

What frustrates you about Google Analytics? 

Kristen:  

They change things fairly frequently and even if you're subscribed to like every news alert in the world, they'll still just change things, and you won't know. And suddenly your data will look weird and you have to ask all your friends why and figure out together what they've done. <laugh>, they're just powerful and they, I mean, it's great that it moves quickly. Um, it would be better if we could follow those movements before we see weird stuff in the data sometimes. 

Greg:  

Maybe they could be a little more transparent. 

Kristen:  

Yeah, they could. They could. And they, uh, when they switched from Google Analytics to Google Analytics four, that was a huge adjustment. Um, and they did it pretty quickly. So it was a lot of work for those of us that were making those transitions happen smoothly, uh, to keep the data apples to apples because they changed almost every data point. Um, but now that we're all getting settled into it, there's some pretty good stuff in there. So I'm starting to come around now that the hours and hours and hours and hours of transition time are over. 

Greg:  

Let's follow that thought a little bit more. What has you excited from a tools data analytics tools standpoint? 

Kristen:  

Some of the cool, like cooler things that we can track are becoming easier, easier to track? So I shared the video example earlier. It used to be that implementing any kind of video tracking was like this very massive custom HTML task. And now because Google and YouTube are such good friends, if it's a YouTube video, Google does it basically automatically. And you can see that progress just like you'd see page scrolls. And I have to do minimal work to be able to see that, which is awesome. Same with things like form fills and just some of the lower funnel, higher value actions that happen on a website. Uh, Google's starting to do a lot more of that automatically, which means less setup work for me and more looking at data time. 

Greg:  

That's Kristen: Crandall. She's the director of analytics here at DCG ONE Kristen:, very informative. Thanks for joining me today. 

Kristen:  

Thanks for having me. 

Greg:  

My key takeaways for this episode on data analytics include: 

• Technology continues to open up exciting opportunities for data analytics, especially with regard to managing big volumes of data speed and understanding the full customer journey.  
• As you dive into data analytics, start where you have easy access to the data, then add incremental steps with more data as you get access to it.  
• Understand your purpose. Don't skip the research. Calculate your ROI otherwise you are guessing about success or failure, and give your data time to develop. 
• Make sure you have enough data to make good decisions and be careful with your own bias. Trust the data.  

If you have questions for Kristen Crandall or would like to talk to her about using data analytics to maximize engagement for your campaign or brand, write us today at podcast@dcgone.com.  

By the way, you can check out a story by Kristen: on our blog@dcgone.com. Of course, it's about data analytics.  

Thank you very much for listening. Kelsey Brewer is our whole idea podcast producer for this episode. I'm Greg: Oberst. Watch this channel for our next podcast and more expertise inside and inspiration for whole Idea marketing.  

Take care.