The Data Stack Show - 218: Breaking the Language Barrier Between Data and Business with Joyce Myers of Modern Technology Solutions

Episode Date: December 4, 2024

Highlights from this week’s conversation include:Joyce's Background and Journey in Data (0:39)  Technological Growth in Logistics (3:51)Leadership and Communication in Logistics (6:54)Impact of Data Quality (9:13)Significance of Data Entry Accuracy (12:05)Data's Role in Decision Making (16:01)The Cost of Adding Data Points (21:26)Real-Time Data in Logistics (24:28)Understanding Master Data (31:15)Data vs. Information Distinction (33:21)Navigating Change in Data Management (37:35)Career Advice for Data Practitioners and Parting Thoughts (41:10)The Data Stack Show is a weekly podcast powered by RudderStack, the CDP for developers. Each week we’ll talk to data engineers, analysts, and data scientists about their experience around building and maintaining data infrastructure, delivering data and data products, and driving better outcomes across their businesses with data.RudderStack helps businesses make the most out of their customer data while ensuring data privacy and security. To learn more about RudderStack visit rudderstack.com.

Hi, I'm Eric Dotz. And I'm John Wessel. Welcome to the Data Stack Show. The Data Stack Show is a podcast where we talk about the technical, business, and human challenges involved in data work. Join our casual conversations with innovators and data professionals to learn about new data technologies and how data teams are run at top companies. Welcome back to the show. We are here with a very special guest, Joyce Myers. Joyce, welcome to the show.

Thank you. It's nice to be here. I appreciate it. All right. Well, you are with MTSI, who is a defense contractor, but you have had a long history of working for, being inside and interacting with the military. So give us just a brief overview of your background. Absolutely. So I am an Air Force brat. I am retired from the United States Army as a soldier and also retired as an Army civilian with several little blips of being an army contractor. And now I'm fortunate to be the chief data officer on the corporate side for a defense. So the whole game. Yeah, that's great. So Joyce, one of the topics I'm excited about digging into is logistics and data and how closely those things truly are tied together.

Give us something that you're excited to talk about. Well, I'm actually excited about that. I spent my time in the Army and as the Army civilian, as a logistician. So I'm an accidental data person. I'm just coming to it from the business. So I'm excited about that. Great. Awesome. Well, let's dig in. We have plenty to talk about. Let's do it. Joyce, I am really excited because I don't think we've had anyone on the show who has

done the type of work that you've done within and for the military. And that is going to be, I'm just, I'm so excited. And I have so many questions, but would love for you to just give the listeners an overview of your, just an overview of your career. You gave us a little, you know, a short one in the intro, but what types of work did you do? You were a soldier, but then you did different types of work in the army and for the Army. So would just love to hear more of those details. So I've actually been very blessed to experience technological growth and process growth through my career, right? So I joined Army in the early 80s. 2% of the work of the soldiers were females. So it was a unique experience.

I was a logistician. I was a unit supply clerk. So the person that you went and got your paper and your pens and your toilet paper from, the person who made sure that all of the equipment was on hand or on order. So from the very beginning, we were doing logistics. We called it supply and nobody really thinks of it as logistics but you know how many boxes of toilet paper do i need to order for this many people and do i have the right budget to do that and do how do i order it we have our expendable supplies like that but

army says we're authorized 10 trucks do we we have all 10 trucks? So from the very beginning, right into that manual, all of that was tracked on paper. Wow. In the early 80s. So fast forward a couple of years, we get our first computers. We've got word processing. I personally used my first spreadsheet. My first Excel was graph paper that I would draw the lines on. And, you know, we didn't have the internet. I know I'm very much aging myself, but we bought a copy of D base three and created our first database to track and move equipment and manage that property. And so I really started

getting into the data and our business processes, right? What do we require? What are we authorized? What are we, where is it? What is the status? All of that's data, but nobody really called it data, right? I mean, it's just fields that we were tracking. And then we had army computers, right? Again, not connected. Each organization had its own computer. Our first one was, you know, sat on a picnic table, sized table. And we continued to grow and learn. And then we started connecting those computers. And now this unit could talk to that unit. And this commander could see the organizations below. And we could start to put that information together.

And then those became connected at an Army level. And as time progressed and technology progressed, we could see ourselves better right we could see the logistics not just from a small sliver but from our left and our right and our up and down and it was it's just really exciting to to one i've been able to watch that journey and still participate into it yeah um and be on the forefront of some of those systems being developed and providing some of those business requirements as a soldier, as a consumer of the end result. It was kind of nice to be on the back end. Sure. One thing we were chatting before the show a little bit, but I told you that I went to the 80th D-Day anniversary with my father-in-law. Both of us have,

you know, family who participated in the D-Day operation. So, you know, very meaningful. I did a ton of reading about that entire operation. And one of the most fascinating, I mean, you know, sort of inexhaustible, you know, topics that are, you know, going to be studied for decades and decades. But one of the big ones is logistics, actually. And it is mind-boggling how many people, pieces of equipment, everything they moved. I mean, there were considerations around secrecy. There were fake military operations being executed, ultimately to sort of achieve one of the largest, in some respects, the largest

sea invasion in the history of the world. So can you just reflect on that a little bit, having managed logistics in the army and seen that from paper all the way through to developing the systems and having been a soldier, can you just reflect on that a little bit? It's boggling, mind boggling, exactly. But I think it boils down to a couple of things, right? Leadership, communication, trust. So whether you're using computer systems or whether you're in a room with darkened windows planning the largest invasion in the world, the leadership has to have the vision. They have to understand what is the final outcome we're trying to achieve. Who are the stakeholders? Who needs to know?

Whose feedback do I need? You have to trust that the information you have is the right information. So we know there have been, I can't even begin to imagine the number of leadership lessons that have come out of, right? Communication styles from the leaders, from our general officers that were there,

the communication coordination between the different forces, between the Air Force and the Army, right? All of the forces. So those key concepts that made that so successful are still the foundations of the key concepts today we're using computers but the leadership still has to have a vision we still have to have that goal that understanding what that end goal is because if you just throw technology at it you don't know what problem you're trying to solve. You just have expensive technology. Communication, our leaders have to share what they want,

but we have an inherent responsibility as data, as employees, as whether we're an engineer or an analyst or a two-day officer or a logistician, we have a responsibility to say, this information doesn't make sense or did you consider that my purview should my viewpoint shows that so i think i've thought of this a lot too dda is one of my favorite topics and so regardless of what the data was and think about it was all on paper yeah it's all on paper and handwritten notes And the fact that you could bring that all together comes down to the human element, right? Yeah. The communication, the connection.

Yep. So I'm curious, so we're talking all on paper. Do you think, because you shared a story before the show about kind of the, almost like the opposite opposite of this of where the data quality was low fidelity. Do you think that there's maybe some aspects of, I don't know, a discipline or I don't know what the right word is, where when it was on paper, it had to be so regimented and structured. And like the handoffs had to be really clean. Whereas like when it becomes electronic, some of that's automated for you. But there's still components that are human components that are really important. Do you think that transition has potentially resulted in less high quality data?

I think in some instances that is absolutely the case, right? When it was paper oriented, and I am not in no way advocating for going back to math. Yeah, of course. You say that with personal connection. Yes. As the person who had multiple manila folders on my desk, having to manually review them, I will say that, no, we don't want to go there. But I think there's always been people who are going to slack off. There's always been people who don't do the attention to detail i think we magnify it some in our systems because now so many more connections are out there right before i would fill out my paper and i would hand it to a person and it had multiple checkpoints where it had they a lot of people call it human in the loop i like to call it a reliable human review right so it's the understanding the context the the experience or wisdom theoretically yeah

yeah so i think maybe you know then it was a very controlled environment very trusted leaders very just you will do this right again the military has a hierarchy and a leadership chain. So it's important. And men's lives are on the line, right? Soldiers' lives are on the line. So when we as people know what's on the line based on our results, I think the quality will be there. But if we don't understand what we're putting in the computer and how it impacts

the end goal, if we don't understand in marketing that our marketing command campaign grows our business, if we in HR don't understand that hiring, tracking the right data gets the best people on, it impacts our revenue. I mean, there's just so many. And then, of course, in logistics, you don't put the right stuff in, you know, when you get the right doesn't get to the right place. Can you share that story that you shared with us when we were chatting before the show about that, you know, sort of seemingly insignificant data entry issue, and then the impact? I think that's just such a good example of what you're talking about. Yeah, of course. So in a previous life, when I was working as an army civilian,

we gathered all of the army maintenance data and that maintenance data came in from computers all over the army, every location, right? So this data starts at the lowest level. It starts from the mechanic who is working on an individual truck or trucks. And so at that point in my career, in the life of technology, we were still tracking some of that information on paper. there, someone would walk around the truck and they'd say, this tire needs replaced or this mirror needs replaced or whatever the issue was, right? And then the mechanic would do the work and they would fill out and it took me however many minutes to fix it. So that amount of time is that allocation, right? That time allocation that knows that this particular maintenance activity takes an average of 50 minutes. Sure. So that's a standard and they can justify, right? And we know that when we go and take our car into the dealership, they know that it takes about this

much labor to do something. Right. It's no different. but if you don't fill that form in then the guy putting that information in the computer they don't necessarily know that but it's a mandatory field and then they're not the one who did the inspection they're not the one who replaced the tire they're not necessarily the mechanic they're the clerk putting the information in so easy for them they just go 0.1 0.1 0, 0.1, or whatever the case is, whatever they put in there, zero, because the field just has to be filled in, right? So that data gets aggregated from that organization up, from that organization up, from that organization up, ultimately feeding into the folks at the top who are doing statistical

analysis and looking at information. They're not replacing a tire. They don't know it takes 50 minutes or whatever the amount is, but they're seeing a trend across the army that when they put all the data together and they analyze it, that it really only takes like six minutes to do a tire. We that's not true but the data is saying and so at the top level they look and they say oh well i guess the organizations don't really need three mechanics because this stuff doesn't take as long as we thought it did yeah and so ultimately as they do those new documents to say you're authorized three mechanics maybe next year you're only authorized two there's a cost savings and while the military is not a revenue generating

they still have to be stewards of the resources so if we can save money by reducing positions that's a good thing we can move that money somewhere else. So now this person putting the data in has no clue that their simple 0.1, 0.1, whatever feeds all the way to the army. They may never even see that decision. They may be moved on to a new unit because sometimes it takes a while for that type of analysis and follow on. So I used to brief that to groups of people and they would go, wow, I never knew that what I'm putting in the computer goes all the way up to the army. So I think that's the same in our organizations, right?

In every role that we're in in what is the information that we're providing being used for what business decision is being made and then we have a little bit more of a understanding and responsibility of let's make it good quality yeah or at least the best that we can yeah and i think there's another avenue here because it's what is it being used for? And then there's what could it be used for? Because it very well could be they had been collecting that data for years and nobody looked at it. And somebody even told the individual that was inputting like, ah, don't worry about that. We don't look at that. And then somebody comes, oh, we need to have an initiative. We need to save some money this year. So then it's like,

oh, we're going to look at this. Oh, all that data so like there's all these scenarios where like in data there's not typically a good like we're talking about like master data dictionaries all that there's not typically a good measure of fidelity or quality for a data point like like in statistics there is but we're just talking about data of like hey you know this is like like we were talking on the show the other day in in engineering if you need a part for a spaceship there's this like level of precision that's required right within like 0.5 you know millimeters or something but in data there's not typically that I've seen any sort of accounting for like, okay, this is like measured from a very precise machine and automatically input here, where this was a human fills it out, like you said, on a Friday afternoon, with a very low like fidelity, you know, for the field. So I think that's a big problem for a lot of organizations. Well, and it's, you know, like you said, is depending on what the data is going to be used for. And oftentimes, and that's such a great example, because our information and our data either supports a process or is generated by a process. It's important. So for example,

you're doing business process reengineering. You hear the phrase digital transformation all over the place. People want to optimize their business processes and gain efficiencies. So some of that is you can just remove some steps. But what if in this process, there's this form and this form has been around since 1960. You've always had this form because in 1960, you needed that information. But that's so overcome by perfect example. When I was in the Army, we used to track a registration. Why did we track it? Because in World War II, it was required to drive the trucks on the German highways. Wow.

Fast forward to the 2000s do we really need that is it still a required element is it just this extra made-up number that somebody's coming up with right so we did a several year study to say why do we need it why was it several years because we had to look all the way back to why did we have it in the first place right it's that reason still valid do we still need this and where else is it being used because just because it's on one form data has a life as its own it's up in all these other places so that fidelity is not just impacting that one process. Yeah. Understanding. And that's where that optimization, we hear people, processes, and technology. I think you got to kind of know the information and data too when you digitally transform.

Because you don't always need everything. But like you said, what else could we use it for? You know, is it, can we drive innovation by knowing this information? What other questions can we ask that are relevant today with, I don't know, large language models or neural networks or artificial intelligence, the greater the problem, right? Is it valuable to us now now it would appear so because it appears that everything matters now right but do you want to put a bunch of junk in or do you really want to know what you're using the information for so i yeah that's a it's a great concept yeah we need it anymore so so apparently i'm really pushing paper today because I have another thought on the paper thing you can order a transcript of this show

we'll mail it to you if you fill out the form but there's another thing around the fidelity thing in paper it was much easier to think through cost with people and paper and capturing data like oh I need to capture these 10 more data points. Okay, we're going to need to hire more people and buy more paper. That is a lot cleaner than, okay, we need to capture 10 more data points.

Oh, storage space, basically free. Compute a little bit more than free, but not much more. Okay, great, there's no cost. Well, there is a cost and the cost is is kind of what you're saying is it's the well once we add this it's actually infinitely harder to remove it than add it because like you said you spent multiple years trying to like understand if we could remove this one thing whereas when they added that they probably spent a couple weeks i don't know but there's that weird ratio

of it's really easy to add but it's so hard to take away and when that when it when things appear free up front there's like oh well we'll just capture in case we need it but it can that can cause actually so many problems yeah if the fidelity is bad or or it's misused like we captured it with this requirement and like later like people use it for the wrong thing and that goes back to so many conversations over the years right when we started we didn't have databases we had the data that we had right we could see ourselves and then we could see our neighbors and then we could see our parents and our parents and our parents and all of the family. We could see it all with the right permissions.

But we could see more. And the more we could see, the more we wanted to see. Did we know what we were seeing? Did it matter that it didn't matter to us at all? But we wanted it and then once we had that data nobody wants to get rid of it right so archiving and retention and disposal but natural life cycle of data is at some point it no longer holds value right right or it no longer holds Right. gratification we want it we want it now we want to be able to log in and find it right but if you have if you have 80 years worth of data does that 80 year mark really bring value to your analysis if it's a historical project probably right so that's interesting too right

and then let's think about that we're talking. We've got how many folders on our desktop that we filed Word documents in or PDFs in, right? So those aren't in databases. That's unstructured data. It's our electronic version of paper. Of forms, of reports. So we've got all of that as well. And we're not measuring it to a finite

because everybody does it different. Every Word document is different every year. Right. So, so there's so much opportunity for us to continue learning and seeing what that next cool thing is that can help us get there. Yeah. One question I have for both of you, because John, you spent, you know, you worked for a logistics company. And one thing we were talking about on the show recently was how data is a way to, it's an attempt to describe what's happening in reality, right? I mean, the most real data that we can experience is what we're doing right now. I mean, we are exchanging data live and it's happening in real time, right? At the pace that time goes. But obviously, you can't run statistical analysis on that, right? You have to actually distill that into some

approximation of what's happening. It's interesting to think about all the ways that we do that, but in every case, you lose fidelity. And earlier we talked about how there's a process, or on a show earlier this week, we talked about how there's this process of you sort of have raw data and then it needs to be refined so that someone can model it. Then you model it and it loses more fidelity, et cetera. What's interesting about logistics to me, and if we think about logistics as data, you are really actually trying to get pretty close to physical reality, especially when you think about moving equipment or other processes like that. So I'd love to hear from both of you on, when we think about logistics as data, how do you control for losing that fidelity?

Because in many cases, if we think about, let's just talk about distilling this conversation we're having into a transcript or a summary, right? It's okay to lose a lot of the fidelity because I'm just sort of looking for a summary of the conversation because I want to write a LinkedIn post, you know, to get people excited. And that's great. You know, that it can be very lossy and that's fine. Right. But when you're trying to get trucks to, you know, it can be very lossy, and that's fine, right? But when you're trying to get trucks repaired into a certain place for soldiers that may be in battle, now we're talking about data that really needs to try to, you know, it really needs to map pretty closely. So, yeah, Joyce, why don't you speak to it, and then John would love to hear about your experience as well. So one of the phrases that is like, I don't know it doesn't cause trauma but it's definitely there so as we were gathering

the data from all the different places and leaders would say we want to see where the things are and we need it in real time real time so real time means like right now when it's happening. So we got them off of real time to near real time. So what is near real time? Within 24 hours, within 48 hours. So that's so important what you just said, right? Because the farther away you are from the actual event, the more likely that the data is not the same, especially as a truck is moving or on a shipping

status. It's been picked up at the warehouse. It's gone through six RFID tags, right? And now, you know, it's here. So real time, it might still show it at time, it will show it at the sixth RFID tag. Near, it might show it at the third. Or you might still see it at the warehouse because you're not tracking all of the tags. We used to see it at the warehouse, in shipment, and arrived. That was it. You were lucky if you got the middle piece. So I think that's a really important thing,

that farther away you are from the point of the actual creation of that data, you as a business, as as an analyst as an executive have to decide what is the acceptable timeliness of the data for your question that you're asking john yeah i i think i can really mirror that in my experience the thing that came to mind was warehousing. But it's very similar to what you're saying. There's typically an absolute minimum requirement of I need to know this high level, let's call it a status. Like picking, shipping, shipped, arrived. There's just a few of them. We absolutely have to have those.

And then the fidelity below that like is super helpful for optimization of business processes and performance potentially so i'm thinking specifically in warehousing in warehousing and one side note is it's actually interesting to like there's all these like changing of systems typically like if you're like if you've got like a third party involved and you've got a carrier involved that's moving it and then you've got somebody receiving it. Each person's system is different software-wise, but the data ideally stays in sync. So that's a whole other thing to think about. But assuming that you get that part right, then it's like, okay, so the shipment arrives on a dock somewhere.

It sits on the dock and say it got stuck on the dock. It got shoved behind something or whatever. Maybe it was a small shipment. Like that'd be great to know. And that would really help with troubleshooting when that shipment doesn't flip over to delivered as expected or somebody calls would be great for, you know, let's say customer service to know like, Oh, it's still on the dock. So those are the types of things that like when you increase the fidelity, it's helpful.

But there's challenges where you're actually typically adding work to increase the fidelity. So you can imagine if you work in a warehouse, like, okay, I'm going to scan it in when we first get it. And then I have to scan it in again here, because that's like a checkpoint. I have to scan it in again here. And there's some neat RFID technology that can make sense sometimes. But sometimes the RFID in the private space is cost prohibitive

compared to the cost of the item. If you're shipping very cheap unit priced items, RFID, the cost, it doesn't make sense. So then you have that problem. And then there's the labor problem of like each of these subsequent checkpoints. A lot of times you're adding labor to each touch, which you're trying to reduce. So it's this weird push and pull of like higher fidelity with, but keeping the labor costs and stuff low. It's the trade-offs, right? And that's where that

the different costs. And then as a leader, as the analyst, you can make those recommendations. Here's a trade-off space, right? We could add it in this part of the process. So working in the corporate job that I'm in now, right? So I'm inward facing, working internally for my company. And then my company supports external customers. And so it's a really interesting dynamic one. It's the first time i've

ever done this focused in so i am i'm learning all new sets of information and data that perhaps i didn't know before but the concepts remain the same right do we understand our process whether it's a logistics process or an hr process an IT process? Do we understand it? Do we know that information flow? And that goes back to, right, sometimes data on its own doesn't give you what you need. You need this data and this data. You need information to make the decision. It might come from different systems. It might come from different organizations. Like you said, sometimes it's third-party logistics. That's where master data and data quality starts to come into play. Whether you're supporting a customer or whether you're doing it internal for your own business, is everybody calling the customer by the same name is every employee

called the same name if i go to a database and i'm myers comma joyce or joyce middle initial myers or j.myers but they're all me right now there's three different entries for me it looks on the surface like there are three different employees. So that's where understanding master data, regardless of what the master data is, an employee, a customer, a vendor, whatever it may be. That's where I think organizations, especially from a logistics perspective, can really gain some ground by understanding, because regardless of what your end goal is, those master datas are going to be used across multiple angles. So I think we could always just really focus on identifying and managing our master data and having people understand what that actually means.

A lot can really bring some improvements. Joyce, can you speak to, you made a statement about data and information. You have data, but you need information. That, I think, is a really fascinating distinction because in many companies, you just refer to all of that as data, right? We need data or we want to be data driven. Can you speak to that distinction? Because data and information are not the same thing necessarily. So I'd love for you to dig in on that a little bit. So I'm going to put it in like a real life example that has nothing to do

with a job, for example. So I wake up in the morning and I want to know how to dress and I look outside and it's sunny. That's a data point. It tells me that the sun is shining, right? So I could look at my app and it shows me the little sunshine. That's a data point, but it's 33 degrees out there. It's cold. So now I've added the temperature to the thing. So it's a little bit more information. So one data point, sunny, I might go out in shorts and a t-shirt. I might make a decision on one data point, but now i've added the temperature that tells me it's cold okay but maybe i like it to be cold that's okay so i could make it that's information now i've had more than one data point i've got a

bigger picture but the wind is blowing 30 miles an hour yeah So as you bring these different data points together, you get a fuller picture. I get a information about the, right? Alexa, tell me what the weather is today. Today we'll reach a high of 52 degrees with sunshine and winds of up to 32. I have information. The weather, right? Yeah. Temperature is a data point. The sun is a weather point. The wind is a data point, not a weather point. But when I put those together, I have weather information. And so we all do that. We all look

at our phone, at our weather app. We all look at those types of things. So I think as an organization, people like to talk about data all the time. I can't find my data. What data do you want? Well, I need to know. I need to know something. Right. Well, is it one data point?

I need to know the temperature at two o'clock on Friday. Yep. Yep. Yep. So I think sometimes just understanding that because our reports, our Excel spreadsheets, our, our dashboards, that's information. It's a lot of data. And if you have so much data, too much data, old data, bad data,

then your information that you're making your decisions on is skewed. Yeah. Yeah. It just makes me think about it even just in, in meetings I've had this week, I wish I could go back and say, okay, is that data or is that information? You know, and just, I think it's such a helpful distinction well people use them interchangeably it's it's my part i mean that's sure i think it really helps people understand better the flow through like if i'm in hr and i onboard an employee and i put that name in there that employee is a data point but everything about it put together is the information about that. Yeah. Yeah. That maybe starts to be used in other places. Yep. Well, I know we're getting close here.

I have a question and then John want to leave time for you to speak to our listeners who heard about a multi-year process to deprecate a registration number and who maybe really identify with that and say, I'm in a company, we're swimming in data, we don't need half of it, but it just feels like a really big mountain to climb to actually make change and turn data into information to our previous topic. Can you just speak to that person after having gone through that? I'm sure in an environment where security and hierarchy probably tend to be a little bit more stringent than you know, than say, yeah, than the average, you know, than the average private business. I just love for you to give that person some tips and encouragement on how to face that when they go into work tomorrow. So I'm going to start with change is hard. All change is hard, right? And so that was changing not only a data element, but regulations and policies and cultural beliefs and years of history and emotional attachment to the people who managed it.

So keeping in mind, change is hard and people don't like to change so if you keep those two things in mind instead of starting with the big thing we're taking it away for the whole one right that's what we said we're going to research why we need this right and we broke it down why if you go with the who what where when how wise who uses it then you start to get a stakeholder base right what role do they play what are they using the information for what are they using it for where is it being used why did we start using it in the first place? When is it used? When you ask those types of things, when you're doing change, when you want to remove something, some of those will naturally help us bring out the, maybe we don't need to move it. Or maybe the trade-off space, like you were talking about earlier, it's not not the value of removing it is not worth the

of keeping it right so so that there's a lot of analysis that goes in but patience right just really if you can get into the mode of doing the who what where when how why for all of that kind of things, that really helps. And then what happens if we don't do it? And so that's what I would challenge, regardless if you're trying to remove old data, people do not want to get rid of their data. And so it's like, well, what will we use it for? When will we use it? When you start asking people the

questions and they realize they can't answer them then it opens their eyes right but just saying we're getting rid of it then people don't feel heard or seen what they're using it for feels negated that maybe what they're doing wasn't as important or right and i'm talking about one specific example but it took several years and we actually had a luncheon party at our we had a little like bingo chart on when we were going to get it approved and that's awesome i love it yeah it's one of my favorite memories actually but yeah i would just encourage people patience and remember people don't like change. Change is hard. But you don't have to change everybody's mind. Just the right people's mind.

And then the who, who, and how, why's. And I would say in everything that we're doing from a data perspective, keep those questions in mind. I like to ask this question because I know we have a number of data practitioners at all levels that listen to the show. What advice would you give somebody that's a data practitioner, very broad, more engineering, more analytics focused?

There's a lot of versions of that. But what advice would you give somebody that wants to do more leadership-focused work in the future and is currently doing maybe exclusively individual contributor work, what career advice would you give somebody? To know your business. So as a data person, whether you're a data analyst, a data engineer,

building pipelines, building databases, any of those, a data scientist, any one of those, any one of those is putting that data together for information to solve a business problem. When you know that business, when you can speak the language of the business, and it doesn't have to be the whole business, maybe you really like HR, or you're doing all the times a lot of analysis for your engineering department. Get to know the business reasons. Because what that does is it gives you a bilingual. You can speak the data and tech side,

and you can speak the business side. And that's a really big missing gap. We have a lot of people in the business who don't understand their data at all. And we have a lot of data people who don't understand the business. I would say, for me, I started in logistics and accidentally, like I'm not a data scientist, but I know how to put the pieces together and help others put the pieces together. Because I understand what this is. Yeah, that's great advice.

I love it. Well, it's been such a joy to talk to you, Joyce. I think this was, you know, throughout three years plus of doing the show now, one of the themes that just come up over and over again, and I think the show really in many ways is a great summary of it, is there are people behind the data and people who are going to consume the data. And so it really all comes back to the people element, whether it's paper or, you know, the advanced systems we have today. So thank you for a great reminder. Thank you for your service. And thank you for all of the good lessons that we learn and our listeners learn today.

Thank you so much. I've had a great time and I appreciate the invite. The Data Stack Show is brought to you by Rudderstack, the warehouse native customer data platform. Rudderstack is purpose-built to help data teams turn customer data into competitive advantage. Learn more at ruddersack.com..