Season 2, Episode 2: All About AI & Data Analytics

In this episode, our Vice President of Data & Insights Cody Harker explores all the nuances of combing through client data and finding ways to make it more efficient and innovative for our clients.

Transcript

Tom Marguccio 0:00

Welcome to Coffee and Tea with SCG, a podcast from the industry experts at SCG Advertising and PR. We are a full-service woman-owned agency that offers advertising, public relations, recruitment marketing, and association and event management. In this season, we'll be chatting with some members of our team about innovation in their respective industries. So grab your coffee and your tea, and let's brew up a good conversation.

Lupe Dragon 0:28

Welcome back to Coffee and Tea with SCG. I'm your host, Lupe, and today we have Cody Harker, Vice President of Data and Insights on a special episode recorded online. Thank you for being here, Cody.

Cody Harker 0:42

Thank you, Lupe. I'm excited to uh chat today.

Lupe Dragon 0:46

Of course, me too. So, first question as we ask everyone, coffee or tea?

Cody Harker 0:52

So I'm a big coffee drinker. And I um and I'm I'm one of those weirdos that drinks iced coffee year-round, even if it's 10 degrees outside. But I'll tell you my my little secret on weekends, a little splash of Bailey's in the coffee is just sublime.

Lupe Dragon 1:12

I mean, that sounds great. I mean, why not? It's a Saturday. You're fun, you're having time at home, maybe you're reading a book, have a little bit of Bailey's inside your coffee, go for it. Um, I'm definitely more of a tea drinker. Um, I do like coffee occasionally. The last time I had coffee, I had a brown sugar shaken espresso. Uh, never had one before. That was actually pretty good. I I have to say I was surprised. It also kept me up all night. So, you know that watching a lot of Netflix. So let's put that out there. Today we have you here on the podcast. This whole season has to do with innovation, and you're the data guy. You're all on that. I'm gonna ask you some stuff about that. So, to start off, uh, what are the biggest challenges you see now in traditional data and analytics today?

Cody Harker 2:01

Yeah, so that's an interesting question. And I'm gonna start off by answering it in probably not the coolest way. I see the biggest issue with data and analytics being more of a business process thing than a technology issue. A lot of companies want clean data, they want dashboards, they want visualizations of data and really good understanding of what's going on. But at the end of the day, data in equals data out. So if you're not doing the right things to produce quality data, the outputs aren't going to be what you want them to be. And that can look like a number of things. In a marketing agency, on behalf of our customers, it can mean using the right URL source parameters. It can mean making sure that source codes are being collected and that we have the proper tracking in place and all of the little sort of bells and whistles that come along with managing campaigns. So that way, when you look at the end of it, you're able to see something that's good, reliable, and hopefully actionable as well. Aside from that, I think scale isn't as a real big issue. Scaling reporting and analytics is just tough because when you think about scale as a concept, it doesn't really lend itself to meaningful insights because meaningful insights tend to be really pointed towards a specific organization or a need. And when you build, when you scale, you're hopefully building a product that serves 80% of your customers' use cases 80% of the time. And so there's a lot of value left on the table. And really, there hasn't been a lot to sort of reconcile that thus far. And so we we just run into issues where you try to create templated things, and those templated things, whether they're ongoing reporting deliverables, dashboards, et cetera, just don't quite serve the fits and needs of the client that you're working with.

Lupe Dragon 3:54

Yeah, I mean, that's important, trying to address what each client wants and needs. They might not be the same. So if you're trying to measure something that doesn't really apply to what they're going for, then it doesn't really serve a purpose in uh trying to drive what they're looking for. So that that makes a lot of sense. What are some of the challenges being addressed by innovative approaches?

Cody Harker 4:16

I've been seeing the rise of data center of excellences a lot and more data leadership roles out there. I also see data consultancies and data service providers rising in prominence as well. And you even have more sort of data in a box solutions where people or companies rather will come to the table with an end-to-end data management solution that allows you to connect data pipelines, manipulate the data, transform it, store it, and then visualize it. And I think that's been very helpful, especially for smaller and medium-sized companies that don't have complex data needs and or let's say might just utilize a few different types of channels and SaaS platforms where you don't have too many moving parts. I think we'll also see that generative AI will help more in the future. But for now, I haven't seen a ton of value in generative AI, specifically when it comes to data and insights. I think as that technology progresses, we're going to see AI be able to do very valuable things like transform data more easily, pull out trends, run different types of analyses, and have a higher degree of confidence that the outputs from it are usable and reliable.

Lupe Dragon 5:38

Yeah, honestly, that's fair. I mean, AI is such a new thing that has come into the scene. I mean, it's been developing for a while, but it's starting to learn and grow still. So finding how that finds its place in data, I think might take another year or two, maybe more to kind of figure out how that's something that we can use in what we're doing now. Uh, with that being said, are there any trends that you're following closely that involve data analytics and maybe approaches or strategies that are now uh new in 2025 or going further?

Cody Harker 6:10

I think no code solutions are really interesting to me. That might be somewhat of a personal preference because I am not by trade a data scientist or data engineer. So I don't have a heavy background in Python, SQL, or R or some of the other languages that are common in the data space. But, you know, graph databases for as an example of being sort of a quasi-no-code solution are somewhat interesting to look at. For me, I think the most exciting trend that I'm starting to see is a larger amount of more narrow applications of artificial intelligence and how that can help with decisions that we make as marketing professionals, PR professionals, folks that, you know, do all sorts of different things. When we think about Gen AI being, you know, and we equate it to Chat GPT and some of these very large companies that are boiling the ocean, it seems hard to imagine a place where those applications happen very quickly for specific niche needs. But I think training models that have more narrow applications, like let's say ones that maybe just focus on labor market data or on marketing platform outcomes, things like that are really interesting. And of course, I do have to say, you know, explainable AI, quantum computing, those types of things are really fascinating to me without having an in-depth knowledge about them. But I just don't know what to expect yet. There are so many unknown unknowns when it comes to applications of AI and implications of using AI and technology that enables it, both the software and the hardware, that it's it's hard to expect what the outcome is going to be for some of those really new emergent trends. I do think we have a fairly good idea that Gen AI, when trained for narrow use cases, um, is going to be highly beneficial. And I think for marketers, whether you're a talent marketer or a consumer marketer, having that as a sort of co-pilot or assistant in doing your day-to-day job, I'm I'm most stoked for that.

Lupe Dragon 8:25

I think there's so much that could come from that as far as you know being able to figure out the prediction of the AI or like what it might be able to spit back at you as far as what we can do differently, what's working, what's not working, be able to have that instantaneous result from the data. So I think that might be something cool to look forward to. How do you see your role of data analyst evolving in the face of these innovations?

Cody Harker 8:51

This is really interesting and hits home to me. I've managed a lot of data analysts in the past. I've moonlit as a data analyst and now lead our data and insights practice group. So I've seen a progression in what we understand someone who works in data doing. I feel like 10 years ago, if someone told you they were a data scientist, you would look at them like they were a wizard or something. And we didn't really have this sort of understanding of what data people do. It's just this very esoteric field that sits out there. I think, you know, number one, we can look at some tangible numbers. I'm going to quote a good friend of mine, and I believe he quoted someone else on this, but there's been a 456% increase in prompt engineering as a course of study. While we've seen core skills like SQL and Java actually decrease in interest. So if I take that and paint with broad strokes, what we're looking at is a theoretical upcoming class of data analysts that are not necessarily as technical in background, but are better able to leverage tools like Gen AI to help them out. Certainly, I know folks in data and software engineering roles that use ChatGPT and other tools like that to help them write better code or troubleshoot what's going on. So I think we'll definitely see less technical minds and data. And I think philosophically, too, we're going to see more non-traditional data folks get into the data space. I'm kind of reminded of a trend that happened about 20 years ago when I was going around and applying to colleges. A dean of admissions told me that more and more med schools were taking people from outside of STEM backgrounds because they really appreciated having the diversity of thought coming into the world of medicine. Basically, up until a certain point, you were on a pre-med track, and that led to medical school, and then you led to your specialization from there. And more med schools were looking at how they could bring in people from the humanities, bring in people from other types of degrees, and teach them what they needed to learn to become a doctor. But having that diversity of thought when it came to medicine led to a lot of innovation. And I think when I think about where we're at with data right now, we could use that same sort of innovation that comes from having people with a breadth of different backgrounds looking at data, piecing things together and helping shape what the future of analysis looks like, especially if it's not code intensive.

Lupe Dragon 11:36

I think that's a really interesting point that you brought up about the med students and how we're picking people who don't only just have a medicine background, but someone who has all these different skill sets. Because I feel like I went through that in college too, where I went on a PR track, but I'm also doing like like this, I'm doing podcasting, I'm doing video, which I think it's important, you know, especially if you're doing data for like the job market, that, you know, if these employers want employees with different skill sets to be able to drive the diversity to not just one group of people, but to cast a wide net. Um so I think that helps a lot. Going back to AI, since we've talked a little bit about it, can you explain how AI is revolutionizing data and analytics and maybe what are some of the pitfalls of it at this point of where AI is in 2025?

Cody Harker 12:21

I'll um I'll answer it in a couple of different ways. Number one, for me personally, I see AI revolutionizing data in terms of the ability to add context and iterate quickly. So um, when I'm looking at a client's campaigns, when I'm thinking about what's going on with um some of our firm's larger clients, why certain things happen the way that they did, numbers are great. We want to empower our clients to have the numbers, be able to do whatever with those numbers. But to tell a story of why, you need a lot more context. That context requires research. It requires thinking through lots of different things and scouring. And so tools like generative AI are good at providing summaries and going out and fetching things that help me contextualize what's going on within the work that we're doing for a specific client and taking care of repeatable work that I could probably do myself, but don't have time to do in the day. You know, when we try to understand why the research component ultimately takes a lot of time, effort, and there are unforeseen variables. Um, I think that AI, especially as it grows and progresses, is going to be able to help provide that context and summarize why things have happened, or at least provide analysts with a much larger toolkit to be able to answer that story. In terms of the pitfalls, biggest thing that comes to mind is really around legal issues. I attended a really fascinating conference about a year and a half ago, and um a managing director from Accenture or Deloitte was there. And I remember his conversation, he was talking about how you know generative AI is really going to reshape the world of work and certain jobs like copywriting will be upended, and by upended, not necessarily meaning they'd go away, but they would change dramatically because of it. And and then he caveatted all of this to say that there are so many unknown unknowns regarding generative AI and its uses that there are going to be lots of small, medium, and large-size companies that don't adopt it because there are no legal precedents in how to handle certain things. So this is a bit of a morbid analogy, but this this person said if you're chatting with a generative AI tool and it tells you to jump off a bridge and you jump off a bridge, who's responsible? Is it you for committing the action? Is it the generative AI tool for telling you to do it? Or is it the underlying data set that powers the gen AI tool? And, you know, taken to an extreme, that's obviously not something that likely would ever happen. But you can imagine where you could apply that in a business sense where generative AI points you a certain direction and that winds up being the wrong direction, or something that infringes on a legal issue arises as a result of it. And we just don't have tons of legal precedent for what happens in that scenario, which creates a lot of risk. And I think ultimately it's limiting how companies use AI and whether companies will dive all in in the near future, I'd guess.

Lupe Dragon 15:46

I mean, that's fair. Definitely the legal implications are obviously important. Nobody wants to do the wrong thing. We want to do the right thing by our clients, who whoever your client is. So I I think it's important to definitely see ahead of that and make sure that we're, you know, doing the right things to have things in place once we start to get AI more um adopted in the practice. With that being said, how can organizations best prepare for and adopt these innovative approaches, including AI, including these dashboards, and anything else that comes to your mind when it comes to data and analytics?

Cody Harker 16:21

So I think number one, there needs to be a sort of top-down approach to whether and when and how you're going to implement tools that use AI and ultimately what your data strategy is as a whole. There are a lot of unknowns, unknown unknowns, as we've talked about when it comes to the use of AI, and company leaders will need to assess that risk when they look at adoption. Additionally, I think it's important to understand what analytics drive your business forward, how they manifest themselves into actionable outcomes and change, and be conscientious of the fact that just because you make a leap forward into, you know, something really cool and innovative, like using AI, like creating a really sophisticated data strategy, it doesn't necessarily mean that there will be huge dividends that come as a result of it. A good friend of mine uses this analogy, and I really like it a lot. I say it all the time. He likens a company's data journey to whale watching. If you go to the Pacific Northwest or wherever you go to watch whales, you can pay someone to take you out on a boat and go watch whales. There is no guarantee that you're going to see a whale when you pay that person money that he, she, or they can definitively tell you is we'll put you in an area where we believe whales should be. So if we take that and apply to the business world, just because you've invested time and effort doesn't mean that Moby Dick is waiting on the other side when you do this sophisticated data strategy.

Lupe Dragon 18:04

Honestly, that's so fair and whale watching and data. I didn't know how you were going to tie that in. I was like, you know, that that makes a lot of sense. Um last two questions. Uh, what are the ethical considerations that need to be addressed as data analytics become more sophisticated?

Cody Harker 18:23

You know, I'm this my biggest personal ethical consideration is environmentally focused. I I don't know that that would necessarily be the most popular answer if you pulled a large crowd, but uh the amount of power that it takes to use generative AI and and to power a lot of these sophisticated data tools is just it's it's enormous. And we're already seeing um nuclear power plants being recommissioned, and it could significantly alter the way that we power our grid. Um, I really think that we need to be environmentally conscientious of how and when we use tools that use such computing power and such electricity, because, you know, being candid with you, I live in a longleaf pine forest and it's absolutely adorable, and I wouldn't ever want to see it go away. So that's my biggest personal ethical consideration. I think, you know, a better answer to that might be data security and privacy. Those are obviously pressing concerns. You know, what types of data are being used for training models? Do you consciously know that the data that you put out into the world, even if it's a photo of you and your family or an Instagram post about you being at a concert, do you know whether that's making its way into an AI model or not? Are you able to have agency in that decision? And so, you know, I I would be worried about that as a society, that we have clear boundaries, restrictions, and understanding of data concerns and data privacy when it comes to widespread usage of Gen AI and the um large language models that power them.

Lupe Dragon 20:07

That is so important. I mean, that's something that we're talking about a lot right now in the media is who is in charge of our data and how are we able to take control of that and how are we able to use it to our advantage for and or not? You know, um, we all deserve to have agency over that and to figure out where we go with that. So that's definitely a topic that a lot of us are talking about right now, and maybe AI can assist in that. I guess we'll find out um in the next few years as we start to use it more. Last question What are your predictions for the future of data analytics in the next five to 10 years?

Cody Harker 20:45

I'll say I'm a little bit of a Luddite. I I tend to not be as forward thinking on the technology front as others. I am bullish that we're going to see a lot of innovation happen. I think that there's going to be tons that happen, but I think this era of big data and democratized data is going to calm down a little bit and we're going to see that analytics become more targeted and the outputs are refined to specific use cases. I remember, you know, 10 years ago, first time we had a dashboard that we put in front of a client, we threw every single number we could possibly fit into that dashboard because we just thought, oh, it's cool. We we don't have access to these numbers before. And I think that caused things to that caused a trend of basically these very data heavy insights poor dashboards that have kind of become commonplace across the digital marketing world and really just in general. So I think we're going to say bye to those days of data heavy dashboards and get more pinpointed in what analytics we need, what use cases they serve, and how we're driving change from. From the analytics that we're providing to customers and internally as well. I also think that context is going to become more centralized to data. One of the things I believe in very strongly is that data on its own is not necessarily all that useful outside of, let's say, sounding an alarm, right? If we're spending $1,000 a day on Google and then out of nowhere we see that number skyrocket to 10,000 in a vacuum, you know, we should be able to say, shut this off, something's wrong, let's go and investigate. But context into understanding why things happen is going to become more and more important as that data becomes more accessible to marketers, PR professionals, and really anyone out there in the um, in the sort of sit-down, more professional uh type roles. And then lastly, two more things that are somewhat related. I think we're going to see a rise in senior leadership focused around the creation, maintenance, and dissemination of data. Many companies and clients I've worked with have data strategies that are in place, but data exists in these little pockets here and there, and knowledge is not necessarily centralized. Sometimes you have duplication of SaaS tools to do the same thing, and you don't necessarily know when and where data is being used. And I think it there needs to be C-level visibility and understanding into what the data strategy is at an org. I don't know necessarily what that looks like, but I think you're going to see more of an understanding of data AI systems and procedures within different companies out there. Alongside that, I think that you're also going to see even more and more data consultancy and service providers out there to help navigate the very complex world of data. There are so many different types of tools, whether you're looking at databases, connectors, um transformation tools, visualization platforms. And it's a very murky, obscure world if you don't know it inside and out. And it can become a very, very expensive one very quickly. So I think you're going to have more professional services based around data and analytics to help get your organization in a good place, especially when it comes to those that have been invested in or are investing in other companies. For example, if I'm a private equity firm, I would want to make sure that all of my portfolio companies had a very good data strategy in place. So when I'm looking at their financials and I'm understanding their business plan, I have a high level of confidence that everything I'm receiving is in good working order and reliable. So, you know, in short, I think we're going to see a lot of a lot of changes in the data and analytics world. Some simplified, some things are going to become more complicated. And I think we're going to see a better, broader understanding of the data world, hopefully for everyone out there.

Lupe Dragon 25:10

That's all we hope for, is just to make a better world, make it easier and faster, and more efficient. So thank you for being on the podcast, Cody. I'm soaking in all the wisdom that you're giving us right now. So I really appreciate having you here. And uh yeah, thanks again for being on the podcast.

Cody Harker 25:26

Thanks for having me. It's so much fun today.

Tom Marguccio 25:32

Don't forget to subscribe whenever you listen to your podcast so you never miss an episode. And leave us a review. Until next time, keep those mugs filled and those ideas flowing.