[Listen to the discussion or watch it.]
The next episode of the Data Cloud Podcast features an interview with Kevin King, Senior Director of Business Intelligence and Analytics at Coca-Cola Consolidated, hosted by Dana Gardner, Principal Analyst at Interarbor Solutions. They explore the evolution of data strategy at Coca-Cola Consolidated, the shift toward greater data availability, and empowering business analysts with dynamic data access.
The discussion also covers the critical need for continuous refinement and innovation in data practices that drive business results.
Dana Gardner: Welcome to the Data Cloud podcast, Kevin, we're delighted to have you with us.
Kevin King: I'm glad to be here.
Dana Gardner: Tell us about your role at Coca-Cola Consolidated, a bit about your IT background, and what excites you about data strategy and its return to the business over the next few years.
Kevin King: I've been with Coca-Cola Consolidated for 20 years. I lead our business intelligence and analytics department.
As of recently, I also picked up our data quality and architecture organization. What I really enjoy about data strategy is understanding the foundation of what we do as a data organization and making sure we have good data, the right data, good quality data in order for our end users to get the best experience out of what they're trying to do to drive better results for the business.
Dana Gardner: And what have you sensed has changed about what businesses are expecting from organizations like yours? How are you enabling organizations to be better, faster, and more accurate in their decision-making?
Kevin King: The thing that's changed the most is the availability of the data. Our organization is looking to have data in the hands of our analysts throughout the organization, empowering them to be self-sufficient.
The bottling industry changes day by day and what we look at today may be very different the next day. And so, giving them the ability to be more dynamic with data and ability to drill down without necessarily relying on a traditional business intelligence (BI) team is probably one of the biggest opportunities and one of the biggest pushes that we have as a data organization.
Dana Gardner: Tell us about Coca-Cola Consolidated. What do you do? And that way we can drill down more about how analytics is helping you and how you're developing it better.
Kevin King: We're the largest Coca-Cola bottler in the United States. We manufacture, sell and deliver Coca-Cola products, not only from our digital footprint with online, but also from the brick and mortar. We deliver the products to the stores and actually take orders from those stores. That's what we do best.
Dana Gardner: And how long have you been in this business?
Kevin King: I've been doing it for 20 years, so this is the only industry that I've been in. And it's all been in some form of analytics or data perspective. And now having the ability to lead the data organization has been quite an honor.
Dana Gardner: And how about Coca-Cola Consolidated itself? How long has it been bottling?
Kevin King: Well over 100 years. We've been here a really long time.
It's a family-owned business, a third-generation family. So, we know how to do it pretty well.
Dana Gardner: And you're regional. You're distributing around what area of the United States?
Kevin King: Our footprint is on the East Coast, from the Carolinas up to Tennessee, Arkansas, Ohio, Indiana, also in the Washington, DC area.
So pretty much the I-95 corridor is where we operate.
Dana Gardner: I have to imagine that the data that you're looking to improve includes all the things around transportation and logistics, all sorts of things around procurement, and of course manufacturing and factory floor efficiencies.
And then there's always the back office typical for any business. Is it fair to say that you have a wide variety of data needs and therefore you have to pursue a lot of data resources?
Kevin King: Yes, you have it absolutely right. We have a large amount of data, different types of data where we have to figure out how to get all that in one ecosystem. Snowflake has been the data cloud that we've been able to do that with.
But not only that, it’s also external data. We have our day-to-day internal data, but as we move toward the future, there's even more need for external data to be smarter and be more relevant on how to serve our consumers and our employees.
Dana Gardner: As a 100-year-old-plus business, there's a maturity there, but you can always do things better, right?
And you're not necessarily looking to re-engineer your business, but to refine, improve, and automate. Is that fair to say that this is an exercise in refinement rather than wholesale change? Or maybe I have that wrong?
Kevin King: No, I think you have it right. Everything we're trying to do from a data organization is adding incremental new value, unlocking new potentials.
Because we've been an organization that have done things a certain way for a really long time, that creates a lot of manual processes, so types of inefficiencies with what our processes are and what we do with data. I think it's definitely an incremental approach with the capability to unlock new potential.
Dana Gardner: All of us are very familiar with Coca-Cola, the brand and the drink. I have to imagine, is there more that you're distributing than just the beverages or is it strictly the Coca-Cola beverage that you're bottlingand distributing?
Kevin King: Yes, it's strictly the Coca-Cola beverage. We do have some trademark rights for Dr. Pepper and a few other brands, but the majority of what we deliver is going to be within that Coca-Cola trademark.
Dana Gardner: We all take it for granted. We go to the convenience store, the supermarket, maybe just a machine on the corner, and there's the beverage. What is it about getting that there that people perhaps don't appreciate? Is there some part of being a distributor and a bottler that maybe we take for granted or don't understand?
Kevin King: Yes. When I started, I never realized that when you go into a retailer such as a Walmart or a Harris Teeter that most of the product that you see on the shelves is actually put up by a Coca-Cola employee.
It's very high touch from the warehouse floor to the transportation to get it to the outlet, and then someone that's putting it on the shelves. The unique thing is the amount of detail that goes into it. The brand order, the number of products that's on the shelf, the flow, the look and feel, and making sure everything is rotated.
There is a lot of science behind the scenes to make sure our retailers and customers have the best experience, which obviously drives more velocity for our consumers.
Dana Gardner: I can't think of too many better use cases for people, process, and technology, having to work tightly together and looking for the refinements across those domains.
Tell us about how data science and AI now are starting to drive these improvements. What do these technologies bring to the table that you couldn't have done before, for example?
Kevin King: We can understand the business in a different way. I mentioned how fast-paced the business is and having the ability to provide quicker insights, the ability for a salesperson to walk in the store and we've already synthesized all of the information for how they're supposed to set the store, how the store's supposed to look, how we price that store.
We can now synthesize all of that data to allow them to focus on selling to the customer and focusing on making sure we have everything set correctly. It's a new place that we're at versus previously the amount of time that they spent trying to get all of that data together and then to manually process all that data.
It's now real-time insights that allows them to be the best that they can be -- without worrying about all the other stuff.
Dana Gardner: I imagine that the more automation you can bring to your drivers, distributors - with all that touch along the way -- needs to bring the digital to the tactile.
How are you bridging better, by getting more people in that process to be able to act and work with the data? Is there some threshold that we're crossing in terms of making it conversational or using bots so that your employees can take advantage of the data to the best of their ability?
Kevin King: Yes, that's the North Star. We're not there yet, but the more information we can put in the hands of the sales team, the warehouse team, and the more focus that they gain the more they understand their key performance indicators (KPIs) -- that's critical. We can minimize the disruptions with the data outputs that we have.
In the future, the days of standard reports and push reports that they get every morning, you know, I imagine I walk into an outlet. It tells me all the key things about outlet. “Hey, I had a customer give me a call and they submitted a ticket because they had an issue.”
“Hey, I'm missing these key promotional activities. Hey, I have these targeted innovation things that I want to sell in. I don't have them sold in yet.” I think it's just synthesizing all that information in a very, very real time manner, which will really unlock the set of potential for the outlet.
Dana Gardner: That's that conversational question-and-answer query, and then pursue more information and knowledge. That I think is, as you say, a North Star for a lot of what chatbots and other types of large language models (LLMs) are providing. How deep are you into that? How much of the LLM side of AI do you use to try tobridge the gap between the data and the people?
Kevin King: We're at the very beginning of that journey. What I'm excited about are things like Snowflake Intelligence, which will allow for those LLMs and those bigger data models to be put in the hands of ouremployees so they can talk and converse back and forth, to have a dialogue.
Obviously, we're concerned with things like hallucinations, and the technology is getting better day by day. But that's where we think the sweet spot is, when we’re able to take advantage and can move away from our traditional CRMs.
We’d like to have a lot of information and notes and be able to have a dialogue, a conversation with the data. That is just a key to unlock as our true North Star that we're approaching.
Dana Gardner: How about the way you’re using Snowflake services and technologies now? How do you see that changing so that you can progress in this direction? What is it about Snowflake that's paving the path you want to take?
Kevin King: We've taken a very organic approach to analytics, AI, and machine learning. We didn't come out the gate trying to do major investments, getting a lot of third parties.
What Snowflake has allowed us to do is to be very natural. As they create more innovation, understanding their innovation, being locked together as a key partner, and allowing us to take our time as we're learning things like Cortex Analyst or Document AI, for example, to be able to scrub PDFs and then not even to think about all the structured data that we have, pictures and PDFs, and how can we scrub the data and then put that in data warehouse and then allow those individuals that deal with the data.
And to then be able to have those conversations that we're talking about. So, it's been instrumental. They're doing a great job meeting us where we are in our journey and the more they innovate, the more we're somewhat falling behind.
Another great thing is they also allow us to give feedback. What works? Where are opportunities, where are we dreaming at, and how can they meet us there as well?
Dana Gardner: You know, Kevin, thinking more about your industry, I have to imagine there's a great variety of different people that you're distributing the beverages to. That ranges from very sophisticated, like a Walmart, where they have end-to-end insight into their supply chains and real-time data and analytics that they can rely on.
But there are probably also a couple of mom-and-pop shops along the way -- and everything else in between. So that means you have all sorts of different interfaces and processes. That could be a couple of decades old. Tried and true, but old. So that to me means you have a lot of unstructured and different types of data interfaces that you're dealing with.
Is there anything about the way that Snowflake helps you with unstructured data that allows you to still bring this all into the digital domain that you need in order to do those analytics?
Kevin King: Yes. That's our biggest opportunity: That data strategy ability to understand all the traditional systems we have internally, how do we get all that data out of those systems, and manage traditional processes and future processes to make sure that data is a priority.
Many times, we have a lot of gaps because unfortunately as valuable as data is, sometimes leaders forget about the structure of the data and how important the downstream impacts are with the data and then adding the ability to share the data.
So, that's a new outlet that Snowflake enables us with, the ability to minimize these manual emails, PDFs, and invoices and then gain the ability to do a data share where I can just work with that company's IT team. And, you know, within days we have a data share, we're sharing data back and forth. That's an example of a capability that Snowflake is unlocking for us.
It's really making the world smaller as we converse with different retailers and customers. Size used to matter. The scale of the customer would matter around how we can do data share. But Snowflake is allowing us to shrink that disconnect that we historically had, no matter what size and scale the customer is. Which is pretty cool.
Dana Gardner: We talk so much about AI these days, but the cloud is still important, and we don't talk about it as much any more. But having the data both centralized and decentralized -- in a common platform environment that's cloud accessible -- is a pretty strong and powerful technology and capability.
How is that a benefit to you all? I should think that you have people out, up and down the I-95 corridor in the southeast United States, a very large geographic area. And you have drivers, trucks, and machines. Is the cloud model itself also powerful for you to keep all eyes on the same page, so to speak, when it comes to data and process?
Kevin King: Absolutely. That's the key. Going from a traditional data warehouse to a cloud data warehouse provides the ability to one, condense all our data in a structured way, including unstructured data; and two, gain the ability to minimize costs in how we do that. We historically were on SAP HANA, and it was extremely expensive to one, store the data; and then two, to be very performant.
Snowflake’s cloud-based approach has really done a great job of optimizing those capabilities within our data warehouse. Secondly, not only have they optimized it, but they're always thinking about how to make us more efficient with costs, which is quite unique.
That's how Snowflake makes money. But they're always bringing innovation, attempting to make us much better when it comes to our compute and our data storage, which is why I think it's just such a great partnership.
Dana Gardner: And there's probably no better person in your organization than you to go to ask about those metrics. How are those KPIs now returning on your investment? Do you have any quantitative ways of measuring how your approach to data and analytics -- and your data strategy i-- s benefiting Coca-Cola Consolidated?
Kevin King: Sadly, the answer would be no. And part of the data strategy principle we have, and we're actually about to do this. Now that we are four years into our Snowflake migration, we've done a lot of great things. Now we're in a place where we want to take a pause and go back and evaluate. Do we have the right queries? Do we have the right setup? Do we have junk out there that we just have lost attention to?
So, we're going to take a data dive and dive in and partner with Snowflake to understand our queries and our costs and what that really means. Because I think we just have not kept that as a top focus. We're going now and looking at the data strategy that's important, especially as we begin to build our analytical capabilities and we want to invest in making sure that our house is in order. It’s probably my top priority as a data leader right now.
Dana Gardner: That's not uncommon at all. For very mature businesses that are very large and complex, the priorities are always in getting the job done and keeping the customers happy.
And it's only when you become sophisticated and mature in your IT that you can then take that step back and not be fighting fires but actually get to know yourself really well. I think that's a very intuitive and fortuitous approach.
And it's probably the first time in the 100-plus years that Coca-Cola Consolidated has been around that you're actually going to get that real solid view of what's going on within the covers of your company.
Any other thoughts about bringing the people, process, and technologies together? We've talked a lot about product and process, but not so much the people. Are you getting a sense that you're going to have to adjust culture or educate people or get them thinking differently about these tools as they become available?
Once you get to know yourself, how do you then act on that? Because it's the people that ultimately will be a part of that payback.
Kevin King: As an organization we're in a paradigm shift. The business wants to go really fast. They want a lot of information.
We're bringing all different types of data into the data warehouses I spoke about earlier. We are in the period to educate and create a culture around data. That's why that data strategy is so important. We're also at a place where we want to increase the empowerment of our organization with accessibility to data.
Obviously, there are things you have to be concerned about; cost, obviously being first; security, the safety of the data; and the accuracy of the data, the quality of the data. Part of this journey is to begin to educate and sometimes explain data is difficult to do because everybody just wants it.
They don't really know or want to care about the downstream impacts to their decisions. And the reality is, most of the time when we have issues with data or inaccuracies of data, it's typically driven by a business process. They may not want to accept that, but I think that's just a reality.
We have to build a broad perspective that we have to educate. And then also for the people that we're empowering, to access data. One of the parts of our strategy is to create a data citizen policy. So how do we teach you how to use data? How do we teach you how to leverage data in Snowflake?
How do we teach you how to use the appropriate tool? How do we as an organization choose the appropriate tool for you to be able to do the things that you want to do for data? So, there's a lot that we're trying to figure out around this data strategy, which is mostly educating and creating literacy around data and what we do and how we use it.
Dana Gardner: Before we close out, I wonder if you have any advice for other organizations, perhaps mature businesses like yours, that are trying to get that data-first culture and benefits going? Looking back with 20-20 hindsight, what would you advise others to do now that you've been through the part of the journey?
Kevin King: The thing that we jeopardized along the way was that data strategy. In being very thoughtful in what we're attempting to do with data, not only in current state, but in future state. Where do we see it going? Because when I talk to peers and other folks in industries, the thing that the data scientists say all the time is that this is really bad day and I'm spending more time scrubbing the data than I am actually writing logic and doing statistics and analytics.
You have to have self-reflection, right? What are we doing well? What do we really suck at? And then what are the processes and procedures and strategies that allow us to come to the center? We did a lot of guerrillawarfare as we centralized BI.
And as we began to take over responsibility for more of the data assets for the departments, that was the right muscle that we needed to build on in order for us to be as impactful as we expect to be in the future.
The last thing I would say is it can start at any time. No matter if you did it right in the beginning, if you're reactive to it. The biggest thing is you have to start and define what your data strategy is, no matter where you are, and then take it very seriously and socialize it. We can't do it alone as a data organization. We have to do it with the support of our leadership team and support of the employees that we're trying to give data assets to.
Dana Gardner: That's an excellent place for us to end. Thank you so much to our latest Data Cloud Podcast guest, Kevin King, Senior Director of BI and Analytics at Coca-Cola Consolidated in Charlotte, North Carolina. We so much appreciate your sharing your thoughts, experience, and expertise with us, Kevin.
Kevin King: Thank you, Dana. It was my pleasure.
No comments:
Post a Comment