Evan Wimpey: Hello and welcome to the Mining Your Own Business podcast. I’m your host, Evan Wimpey. Today, very excited to introduce our guest, Bill Shube. Bill is the senior manager for America’s supply chain operations technology for the LEGO Group. You might’ve heard LEGO. You might even be stepping on some and this post-Christmas season, as I know I am. Bill, thanks so much for coming on the show. Welcome.
Bill Shube: It’s great to be here. I’m thrilled to be here actually. I mean, yeah, thanks for having me.
Evan Wimpey: Fantastic. Hey, Bill, to get started, can you give us a little bit about your own background and where you sit today?
Bill Shube: Yeah, sure. So I’ve been with the LEGO Group now for about 11 years—pretty much always in supply chain analytics.
I’m also in the process of starting up my own supply chain analytics business on the side called Supply Chain Watchtower. I’m hoping to help smallish manufacturers and distributors gain better visibility into the health of their inventory. And so on that point, I do have a little bit of a fine print I want to hit on. I’m not here on behalf of the LEGO Group. I’m here on behalf of myself and my business, Supply Chain Watchtower. But of course, all my experience is with the LEGO Group at this point. So that’s what I’m going to be drawing from. So yeah—so that’s me. And fine for now to the way, let’s talk about data.
Evan Wimpey: Perfect—yeah. Sounds good. Always happy to talk about data, and we’ll make sure we throw a link to your website there—or any links that you’ve got—in the show notes. Can you talk about, you know, you’ve been there 11 years I think. Can you talk about your role now? And maybe just sort of give us a, you know, a lot of folks think LEGO, and they think as consumers and building and having fun.
Can you talk about what it is that you do in supply chain there and maybe an example or two of projects that you’ve worked on?
Bill Shube: Yeah, of course. So in the supply chain, this is kind of the nerdy side of the LEGO Group, of course. We’re in supply chain ops technology. You know, we’re really—my team is a team of citizen developers.
We’re all business users. And our job is mainly to provide data and analytic support to the rest of our supply chain ops organization that covers supply planning and demand planning, distribution, logistics, order management, that kind of thing. Thank you. And so to give you maybe kind of a better sense of what it is that we do, and I should call out, you know, I know you usually kind of have, data science folks on this podcast.
We’re not a data science team. We’re just kind of, as I said, you know, citizen developers—business users. And, you know, to give an example, I’ll maybe kind of throw out what started all of this for me, which was back in 2019, when I joined, the demand planning team and supply chain ops. And we had just gone through some reorganization in demand planning, and we didn’t really have anything in the way of kind of standard reporting for us.
We were kind of borrowing from other groups. We had one thing that was kind of clunky, but nothing really, not really what we needed. And so I had a reporting background. So my manager tapped me and said, here you go, and go build us some reporting. And I, in my previous life in reporting, I was one of those guys who would just, you know, try to stuff as much data as I could into Excel, write some VBA and hope it doesn’t break, right.
And those things, they always break, and it stinks. And I didn’t want to do it anymore. And so I tried to find, you know, a different way of doing it. I’d heard of Power BI, started tinkering with that a little bit—didn’t quite get what I needed to out of it.
And then I landed on Alteryx, and that changed everything. And Alteryx is basically drag and drop, no code, you know, analytics and ETL software. So it allowed me to build my own data pipelines without knowing the code, which I didn’t know. I knew it a little bit—a little bit of VBA—and that was pretty much it.
And so yeah, within a pretty short amount of time, I was able to build out a data model that was kind of a one-stop shop for our team. It had just about everything we needed in it. And I was able to write the data, you know, to write into Tableau because the dataset was too big for Excel.
And it was mostly automated. It was really just kind of a few button pushes here and there. So very limited data prep involved, and it was a real game changer for us. And leadership got wind of it and said, well, you know, how can we bring this to the rest of supply chain ops? And so that became kind of a side project that led into the, to our current team, which, as I said, you know, our job is to, to basically do that, to build, you know, data pipelines and analytics to help improve the data work that goes on in our supply chain ops organization.
Evan Wimpey: Awesome, Bill. That’s a great story. I love, you know, your leadership catches wind, and the answer is this has been successful. How do we scale this to more teams? How do we make this more impactful versus why aren’t we using this centralized it tool or whatever that we already have? Was there any flavor of that, or it was sort of a, this was already successful, great, let’s keep momentum?
Bill Shube: So from our side of things on the business side, there was no flavor of that because it was well known that our business warehouse that was our main source of data just wasn’t delivering what we needed, right. It’s kind of like the typical problem that a lot of organizations run into.
It’s very siloed data. When I built this pipeline, I was pulling from about a dozen different data sources to, to make it all happen. And I had been doing that kind of thing in Excel, and you know, with that many data sources, you just can’t make it happen in Excel very easily, and it gets messy. And, you know, so this new way of doing things really helped kind of like remove a lot of the barriers that you usually run into in that regard. Did get some questions from the IT side of things, and we can maybe talk about that a little bit later. But it was super helpful to have the support of my leadership because they were able to twist up some arms when it said, are you sure we should be doing this? And they said, no, we need to keep doing this. This is working for us.
Evan Wimpey: Awesome. Yeah, I love it. And you used the term earlier, citizen development—seems sort of like this specific use case of you identifying a problem, identifying a tool that can help and sort of building the solution.
Can you speak maybe generally what is citizen development, and is that sort of a theme?
Bill Shube: Yeah, so I don’t know that there is a specific definition of citizen development. For me, citizen development is where business users, non-tech users who don’t know code, use no-code and low-code tools to do work that otherwise would require involvement from a tech team.
Evan Wimpey: Yeah, and you know, when I think about other folks that have been on the show, a common thread is how do you know, thinking from the data and analytics team perspective, how do you organize? Are analysts embedded in a business? Is there some centralized team? And so citizen development, it seems to me like it’s sort of the bottom up most dispersed, like the people that are closest to the actual problem are the ones sort of driving the development solutions.
Bill Shube: Absolutely. And for me, that’s a really huge benefit that you get out of citizen development. There’s a lot of benefits that you that I found that you get out of it. And if you go kind of, you know, reading about citizen development, you’ll see a lot of things, especially, you know, the software vendors, they all tout all the, you know, the time savings that you can get, the cost reduction, the revenue increases, right.
Improved accuracy. There’s a whole bunch of kind of quantifiable metrics that they’ll talk about. And I do think that all those benefits are real. In my experience, we’ve, we’ve saved thousands of hours for supply chain ops per year. We’ve spotted millions of dollars worth of planning errors.
We’ve built reporting like what I just described that wasn’t even possible previous to, you know, prior to us doing this work. And so that’s all well and good. I think there’s really some great, very tangible benefits. But to your point, there’s kind of the softer benefits, like for me, that the hidden benefits like that, like, you know, bringing the end users, you know, the people who know the problem.
Intimately enabling them to actually do the problem themselves. Things like that really are for me, a huge, huge benefit of citizen development.
Evan Wimpey: Yeah, absolutely. Are you, you, you’ve mentioned this is sort of scaled. Are you, I suspect you’re not the only citizen developing. How does it sort of spread?
How does it work within your team? Do you sort of give access to tools and then say, good luck?
Bill Shube: I mean at first, yeah, that is kind of how it started for us, and I should say, you know, the way, the way we started our citizen development journey was very much a, you know, a bottom up movement where, you know, it was me, it was, you know, my colleagues, a couple of colleagues who introduced me to Alteryx and then, you know, a few other pockets of the organization all, you know, doing this kind of work and Alteryx, Power BI, Power Platform as well.
And that’s just kind of feeling our way. And really kind of taking it, I think a little bit by surprise. First they didn’t really know what to do with us. I would say, you know, that when I started doing this work, I had to like get in touch with the guy who was managing a SQL server and he’s like, you want, you want space in my SQL server, really?
Because you’re just a business analyst. What are you doing here? No really, I need this access. And then other it folks kind of periodically the first year or two, I was doing this would kind of call me up and say, can you, can you show us what you’re doing?
Because, you know, they’d act like they were, you know, just curious, but I always got the sense that there was some underlying concern and, you know, and I keep running into that. Actually, I found that citizen development does kind of make people in IT a little bit nervous. Part of it is just fear of the unknown and concerns over data security.
And I think they’re very right to have those concerns. But as I’ve talked to them more about it, I’m finding that it has a lot to do with governance. You know, and so when you mentioned scalability, it’s one thing to have a couple of users doing this stuff on the side, and you know, you can just kind of let them go and do their thing, but if you really want to scale this across an organization, especially a large organization like the LEGO Group, you really need good governance and, you know, and so when IT was showing these concerns, they were really basically, I think, concern that we were going to develop our own kind of wild west of data, you know, where we’d be making our own apps and data flows, people duplicating effort, building the same thing in different places, right.
Making things that are too big to fail, leaving apps running on a schedule when they’re no longer needed, no documentation, right. Exposing sensitive data, right. All these things that, you know, worries about, and it has. You know, governing principles already in place to, to manage their, you know, they’re worried that we’re going to go and do all that stuff.
And, and they’re very right to be worried because left unchecked we citizens will absolutely do those things. I have done almost everything that I just mentioned. I have not exposed sensitive data, but, but everything else I’ve done at least once. And the thing is that for citizen developers, we have to remember.
So IT kept coming to me and asking, what are you doing? And it got kind of frustrating cause it really felt like they wanted to shut me down. And maybe they did a little bit, but you know, but I I’ve come to realize that these tech teams have spent years and decades developing these best practices to, to prevent behaviors from like this, right.
To ensure data security and prevent tech debt, et cetera, et cetera. And so, you know, so come to real, to see those, you know, these governance initiatives as a good thing, you need governance to enable your citizens. Otherwise you are going to end up with a wild west scenario.
Evan Wimpey: Can you talk about sort of how you, you came back from those, you know, if you, you made a mistake and you’re, you know, you’re using. Bad data or you’re duplicating efforts. You know, what, what sort of the resolution there are you self correcting or are the calls from IT about specific instances or, Bill, what’s going on with us?
Bill Shube: Yeah, no. So for us, we are self correcting is, and, and that’s just kind of a function of the way that we started. Again, we started, you know, very much ground up. This was not. You know, we didn’t go into it intentionally saying we’re going to set up a citizen development regime, right. We, we, we went into it saying, I need to fix this reporting and demand planning.
And then, oh, by the way there’s this other reporting in these other parts of the organization that also stink. Let’s go and fix those. And then all of a sudden, a few years later we realized, oh, we’re citizen developers. Right. And as we realize that, and as you know, we move from just kind of me doing this to a little project team doing this to an actual official team with a leader and everything. We, you know, when we got to that point, we realized we needed to clean house.
And so we started, you know, as a team, we just started figuring out what is it we need to do. We need to go back and we need to document the hell out of all of our applications that we didn’t document and figure out what they’re doing. And we need to audit all of our stuff and make sure that our schedules are running optimally, and we didn’t leave stuff running that we don’t need anymore. Yeah, we realized, you know, we realized the hard way that we needed to do this.
Things just got too big. And so we put it upon ourselves. If you read kind of what’s going on in the industry in this, in this regard there’s a lot of best practices that I think you can, you know, when it comes to, it kind of comes along with governance, you can drop in there to prevent these kinds of situations.
In particular, one thing I’ve seen that actually we are doing in the LEGO Group, not more for in the power platform area where I don’t do so much work, but it’s set up a user community. And it’s kind of a partnership between, you know, the tech team, the platform team that owns the platform, and super users and then all the other users.
And so the tech team should, you know, should be hosting this, you know, you have a team’s channel and host monthly meetings to just kind of. show off functionality and, you know, hold office hours and that kind of thing. And it’s, you know, just kind of reading some of the literature out there, it sounds like a lot of organizations that do that have had a lot of success.
You know, in terms of getting people to follow the governance, the governing principles and, you know, adhere to best practices. It’s a good way to kind of communicate all this out to all the userbase. And make sure it’s presented in a way that’s easily digestible, right.
It’s not, it’s not just coming from the tech teams who maybe sound a little scolding, right. You should be documenting your workflows. It’s also coming from the super users who are, you know, just business users who just use it a lot, right. And so it’s, they can make it much more relatable.
Evan Wimpey: Awesome. Yeah, I like that a lot. And, you know, sort of typically thinking about it from the other perspective from the IT or like sort of data science analytics side that’s coming into the business. You know, we. We sort of think about that challenge to have. How do we communicate? How do we get on the same page here so that folks will use this and they will use best practices for whatever this output tooling is.
So it really seems similar. They’re just maybe more of the burden. More of the technical effort is from the business user instead of submitting a bunch of backlogs that The centralized team is likely never going to get to.
Bill Shube: Yes, yes. And actually I’m glad you brought that up because one of the other benefits that I see in citizen development is a shift in responsibilities away from our tech teams and onto the business teams in a very beneficial manner. So if you think about the way that tech implement implementations often happen and often go wrong—right. Like a lot goes wrong in tech implementations, right. And it has a lot to do with—shocking, right.
And a lot of it comes down to the communication that takes place between the tech teams and the business teams, right. They don’t communicate well. And so the way it usually goes is, you know, business team says, oh, we’ve got this idea. We need this, this new thing to be developed.
It has this technological component. They call IT, and they try to explain it there. You know, they’re. They’re vague. It’s kind of half baked. They’re not very good at explaining exactly what they need. They don’t quite know what they need because it’s still kind of abstract. And so IT then, you know, tries to gather their requirements and, you know, and has this whole routine where they, they, they get the business requirements and then, you know, And then business says, okay, well, that’s all you need to know.
Go and do it for us. Right. And then, you know, I feel bad for IT, right. They’ve got to go and find the data. They’ve got to figure out how to interpret it. They’ve got to figure out how to model it and how to build the pipelines. And then they’ve got to build the front end. And it’s, it’s a lot to do.
And they don’t have the business expertise to know exactly, you know, what it is that’s needed at the end of the day. And so, and this is where the, you know, the miscommunication come in and things get delayed and, you know, and they under-deliver or things fail entirely.
And so that’s kind of the old way. And, you know, and the reason that we have to do it that way, I think this is my theory anyways, that the code is just such a barrier to entry, that it’s, you know, it’s easier to say, okay, it, you need to understand the business rather than going to business and saying, okay, business, you got to learn the code to help them out, right.
It’s easy. Like I can explain I’m in supply chain. I can explain supply chain in three sentences. We make things, we put those things into boxes, and we ship them, right. And that’s being a little—it’s a little, it’s more complicated than that. I can hear, you know, supply chain folks squirming at that, but that’s like, that’s essentially what we do.
And, you know, the rest are details about like, you know, trucking and warehouses and, you know, manufacturing machines. It’s all kind of tangible things that, that, you know, we do. Most people can grasp, whereas the code is abstract, the data is abstract. It just becomes very hard for, you know, to get businesspeople to grasp it.
And so we just kind of say, okay, it has to go do all of this. And you can see where I’m going with this, right. When you don’t have to write the code, it becomes much easier to shift the responsibility onto the business. So, you know, so now the business can say, we’ve got this idea for this thing that we need, here’s the data that we need to make it happen.
IT go and expose this data, make it accessible to us. We will figure out how to interpret it and model it and build the front end. Right. And so from an IT perspective, that’s like right in their wheelhouse. That’s just data engineering. They know how to do that. They could do it in their sleep. From a business perspective, it’s not as hard as it sounds because, you know, interpreting the data. This is the data that they work with all the time in their business anyway. So there’s no, right. So they know what they’re looking at, right. They might not be used to looking at it, you know, in the tables, but they’ll figure that out and they’ll get used to it. And then the modeling without the code, they’re doing a lot of this modeling.
Anyway, they’re just doing it in Excel right now. So now they’ve got more robust tools that allow them to, you know, to do proper ETL operations and model the data that the way that they need it. And then, you know, on the front-end side of things, again, they don’t have to write code. They can just drag and drop.
Tooling into an application or, you know, or visual, whatever it is that they need until, and because they’re experts in the business processes they’re trying to support with these tools, they can just tinker and iterate until they find something that works. And so like I said, you get really much better alignment of responsibilities.
Everything kind of ends up in, you know, in each group’s wheelhouse much, much more neatly.
Evan Wimpey: Wow, that’s a very appealing explanation. Yeah. I absolutely—I buy into that and, you know, maybe, maybe supply chain or whatever the businesses is not always so simple, but at least at a surface level, you know, you can, with a couple of bullet points, you can, okay.
I can make sense out of this. I’ve got the data. I’m an IT person. I’ll figure this out, but the nuance and the actual understanding of the businesses is a lot harder to get. And so yeah, shifting to the business to be able to give them the tools they need. I think that’s really appealing.
Bill, have you, you know, I know your wheelhouse is in supply chain at LEGO. Have you seen citizen development take off in other functions in LEGO and other ops or manufacturing or marketing or sales or anywhere else?
Bill Shube: Yeah, we’re still pretty young—pretty new along the journey of our citizen development, you know, movement, I would say.
I have seen it in some of our manufacturing operations really done excellently. And then in BI too, the, the guy who introduced me to Alteryx in the first place was just a wizard in BI. But it’s really applicable anywhere we’re using data, right. Like it’s just, you know, it’s another way to work with data.
So it’s really kind of a question of time, when, you know, as this kind of starts to permeate the organization, our, our IT organization also needs to kind of get a little bit more, kind of wrap their heads around it a little bit better. They’re still trying to figure out how they want to enable this.
And once that happens, I can really see it taking off. And then the other challenge is that people have to just recognize that they have problems that maybe they don’t realize are problems and be willing and interested to try new ways of solving those problems. Right.
Because Excel is everywhere, right. And everyone can seem kind of happy with it, right. Because they don’t know any better in some cases, right. So like I talked to another manufacturer recently, and they are a midsize manufacturer locally.
They are running their S&OP processes out of Excel. And when I asked, I just talked to the VP, I asked him, you know, exactly what they’re doing. And he said, oh, we’ve got, you know, we’ve got this Excel lady and she’s written all this VBA and she’s pulling from a hundred different Excel files and it’s all automated.
It takes her four hours a month to do. And I’m sitting here thinking, okay, that’s good for Excel lady. But what happens when she retires, right. That’s, you know, like you can’t scale that kind of solution. But he was perfectly happy with it. And there was no way I was going to convince him that he should be doing something different.
I didn’t even, I didn’t even try. Right. Because if you don’t know there’s a problem, then you’re not going to try and fix it. So really there’s a mindset change that needs to take place that yes, Excel is a great tool, but there are now other great tools that you and the business can use to automate yourselves without the help from IT and make yourselves even better and more efficient.
Evan Wimpey: With a lot more structure and scalability than Excel
Bill Shube: Exactly.
Evan Wimpey: Yeah, Bill you made clear, you know, you’re not data science analytics, machine learning type of background. Do you still have those use cases that you need to go to another team?
I’m thinking like a lot of reporting, a lot of data modeling, but is there any like predictive or prescriptive or machine learning modeling that sort of your business would have an appetite for, and is that available in these tools or is that something that you would need to go elsewhere within the company?
Bill Shube: For now, it, that, those kinds of things go elsewhere, but we would love to start doing that actually, you know, Alteryx offers a whole suite of no-code predictive analytics tools, you know, machine learning tools, time series forecasting, all that kind of thing. And we’d love to do it.
The main reason that we’re not is that we’ve just found so much very valuable low hanging fruit in just getting, just getting our basic data foundations in order and supporting our team in the just the daily operational reporting that, that they do. There’s this whole, whole suite of shadow reporting that takes place because as I mentioned earlier, our, our business warehouse doesn’t fully support.
You know, our needs in the business. And so in the absence of support, people just kind of figured out other ways to do it. And with their limited skill sets. And I don’t mean that as a, you know, to disparage my colleagues. I’ve got great colleagues and they’re super smart people.
They’re just not all data analysts, right. This is kind of like a side gig that a lot of them have to do to get their job done. And so those things just keep coming our way. Eventually, as I said, we’d love to get into it. We have taken some training and, you know, I mean, I have, you know, prior years, I have learned a little bit of R, I’ve learned a little Python, right, so we, you know, we do have the capabilities, it’s just finding that, that use case that’s compelling enough that we need to prioritize it, but not quite compelling enough for our central teams to want to bother with it.
Evan Wimpey: Sure. Gotcha. Yeah.
Bill Shube: And that’s really, I should say that’s kind of like the sweet spot for citizen development in general is right.
You know, something that’s worthwhile for us to spend time on, but not worthwhile for our global IT, you know, BI data science teams to work on theirs. And there’s a lot that falls into that category of surprising amount. And that’s, that’s really where a citizen development really shines.
Evan Wimpey: That is a really good heuristic and way to think about it. Bill, I would. I would say maybe, maybe as a final question, let’s sort of ignore any of the constraints about, about timeliness or, or, or prioritization. And let’s just say it’s the Bill Shube Show, and everybody is aligned to your vision, and you get to pick, here’s what we’re going to do next.
We’re going to work on this and say, okay, Bill, that’s it. Where do you point, where do you point efforts?
Bill Shube: Yeah, you told me you were going to ask this question and I’m still having a hard time answering it, but I’ll try. So one of the things that we struggle most with is data access. Yeah, I mentioned earlier, it wasn’t quite ready for us to be doing this stuff and they’re getting there.
But, you know, but if I had my way and I could push people around like that, I would, you know, we’ve had some experience where we’ve actually had access to some, to some of our back-end SAP tables, and it was really kind of eye opening what we could do. Do with access to that kind of data.
We’ve never really had access to that kind of stuff before. And yeah, but we don’t have everything we need. And I would love to, to just get it, all of that stuff and build out a model or a set of data models that, that models our supply chain. And then take those, there’s some really great natural language querying tools out there is a thought spot is one Alteryx auto insights where you drop a data model in there and you can put it in front of the end user and they can just ask it a regular old question and, you know, and get their answer.
And, you know, and so if I had these data models available to me, we could do a ton of great analytical work on our own with that, but, but the ability to also. Put that in front of end users and just let them go to town. Kind of to your, you know, as, as we talked about earlier, you know, getting the people who are most intimately involved in the problem, giving them the ability to solve it themselves.
I feel like there’d be a ton of great stuff that would come out of that kind of setup. You know, it, it takes, it takes a lot of effort to get there. But yeah, I feel like you would. We’d be able to answer a lot of questions about our business much, much more quickly than we’re able to right now if we can do something like that.
Evan Wimpey: Awesome. I love that answer. It’s a, it’s almost an empower everybody to be able to answer the same question of what’s the most important thing for you, for your world, for your process and give them the ability to answer it.
Bill Shube: Yeah. I mean, that’s what citizen development is about. It’s empowering the average user, right.
Evan Wimpey: Awesome. Bill, that’s all the time. We have, thank you so much. It’s been really interesting listening to you and learning a little bit about citizen development there at the LEGO Group built. You’ve been at the LEGO Group. Remind me again, remind the listeners again, the name of your company.
Bill Shube: Oh, right. Supply Chain Watchtower. Yes. Don’t even have a website yet. I’ve got the URL. The website is, is nonexistent. Maybe it will be by the time this podcast goes live. But yeah, anyone who’s interested can also find me on LinkedIn, of course.
Evan Wimpey: Okay, perfect. Follow Bill. If the website’s not live now, it will be one day, and we’ll re-release the episode. Bill, thanks so much for joining us today.
Bill Shube: It’s been great. Awesome. Thanks so much, Evan. Good to talk to you.
Bill Shube | Guest
Evan Wimpey | Host