Evan Wimpey All right. Hello, and welcome to the first and inaugural episode of the Mining Your Own Business podcast. I am pleased to introduce our guest today. Gerhard Pilcher. Gerhard is the CEO of Elder Research, a data science and analytics consultancy. Gerhard is also the author of the Mining Your Own Business book. And if that name sounds familiar, it’s because it’s the same name as this podcast, and it is the inspiration behind this podcast, so I am very excited to introduce Gerhard as our first guest. Gerhard. Welcome to the show!
Gerhard Pilcher Thank you. Pleasure to be here, honor.
Evan Wimpey All right, fantastic. So we’ll jump into the book and sort of how that ties into the podcast and what we hope to learn from the show. But I think maybe this is a quick intro. Could you give a little bit of background about yourself and how you got to? Here is the CEO of an analytics consulting firm.
Gerhard Pilcher Yeah, sure. Yeah, my career has been a little bit circuitous. I would say I went to NC State Wolfpack, graduated from an engineering school and computer science was an engineering background. Let go engineering, and from there I was in the telecommunications industry for a number of years, building digital switching systems and transport fiber optic transport systems. Data was just coming on the scene at the time. In fact, we were arguing about what the protocol ought to be to carry data around, and I was arguing for IP. And some of the telecom people argued for a more telecom oriented, 64 bit packet thing that I said would never support streaming video and things like we’re doing today. A visionary. Yeah. Well, maybe lucky guess, I guess. But I left that I was the chief technology officer and executive vice president for worldwide, for a telecom firm at the time and was spending much time with my family. And I decided that wasn’t really what my mission in life was. So I left a really great job and a great bunch of people that I worked with and bought a road construction company and did road construction. But I couldn’t get away from technology. So I bought things like jeeps and laser guided grading and computers, and the equipment and predictive maintenance to the construction industry and the construction industry went through a hard time. In 2008, I decided to go back to grad school a little bit after that and analytics and statistics back at NC State again went back again. And from there I found research and joined twenty eleven to start their practice in the civil federal government. And I’ve been working here for about five years, and John Elder, our founder, said, Gosh, Gerhard, I really don’t want to be CEO anymore. I want to write and speak with you, take this job. And I said, Sure, I’ll take it for a while. I’m looking for that next person to kick me out so that I can continue to consult with our clients, but hope to retire here and working in this and having been in business for a while, including my own business, that was very cost driven and and thinking about analytics and how analytics could really have served me even better than the basic engineering statistics I was doing in my business, I felt like trying to help executives understand that a little bit better and write a book. I was really terrified to write a book to begin with, honestly. You know, I’ve seen some of the review’s authors have gotten, but I finally realized that I read so much and appreciate that the authors took time to write something down. Even though I might not agree, I rarely pick up a book and agree with everything in it that I ought to at least put some of my thoughts down and offer that to people who are wondering about how to think about analytics and maybe don’t have the same technical background that I do.
Evan Wimpey Yeah, that’s fantastic, I hope you haven’t had to read any of these super critical reviews. I’m sure they’re all just jealous. But yeah, for reference for those few who don’t know. But the book that Gerhard wrote and that is the inspiration behind this podcast, Mining Your Business, I think the subtitle is really informative. It’s a primer for executives on understanding and employing data mining and predictive analytics. So you walked, walked us through your background. You were an executive who, for a perfect, got my copy here as well. Thanks. Yeah, a very unique background is as a business executive. You’re right, a circuitous path. Probably not a lot of analytics CEOs from the telecom and then construction world into analytics, but I think that gives a great vantage point that you have here. And so we appreciate that you wrote that book. It’s been super helpful, and it’s also been roughly six years since it was published. And then certainly on the technical side, air data science, it moves really fast. Algorithms from six years ago are now very old. Is there anything that feels stale in the book now or anything that you wish that you would want to rewrite or rethink? Given that it’s it’s it’s now 20 22?
Gerhard Pilcher Yeah, it’s my most critical review so far and done a second edition in your industry as the industry is changing fast. And so definitely we have been thinking about and begun work on a second edition now, you know, as an aside, I really wanted to do it as an e-book and to get repositories. I want somebody to buy it. I could just go in there and make updates and they could download them and get the latest stuff because the industry is changing. But not only the industry is changing, but I’m a lifelong learner and I’m learning more and learning things that I would like to add. You know, this idea of a repository really came from Hadley Wickham and Famous. And ah, he does similar things with the books he writes. You can just download them from Get and he can update them in real time. I think that’s maybe the future of publishing, but I was overruled on this story. And so maybe the next book? I’ll write it that way. But back to your question, I think one of the things that I’ve already written in draft form to go in a second chapter is aimed at data strategy. And the reason for that is we talk a lot about analytics or machine learning, but the real focus from a business perspective is the information I need to make a better decision. And in some part of my business, whether it’s what’s the probability if I make this decision versus that decision of something happening, whether it’s because I have a complex set of information based on that, what does it mean? And therefore, what should my next actions be or what are their choices for my next action that I can use my human brain to decide? So use a machine to take out some of the complexity and then use what the human mind is? Gift it for making the connections between that and making a decision to go forward. But no matter what we’re doing where we’re really trying to. Get a more nuanced understanding of forces acting on the business, reduce that business somehow or intelligently automate some process in that analytics is just one step in the process to achieve those outcomes. And business leaders don’t typically think. About analytics, they think about what’s happening in my business and how to solve for that. And so I think maybe a little bit more around that in this second edition would be helpful not only for the business leaders, but for the leaders they’re hiring to think about in a different way to understand more, how those executives are thinking and to think more about how to communicate with them and break down the business problems they’re put forward into steps that have to happen analytically so they can have a better perspective on, Hey, it’s not going to happen to MA because we might need to collect data first and then we might need to run a model. Then we might need to figure out how to deliver that for somebody to consume it into a delivery process.
Evan Wimpey Yeah, that’s a really great point. Yeah, and I think I think your first book certainly geared towards the executive and understanding what analytics can do, but ultimately the goal is to to bridge that gap between those executives and those analytics practitioners. So at the analytic manager, analytic leader level, sort of a guide to how they can bridge that gap. I think that’s super useful. They probably also are more inclined to update to get Paul the updated version of the book somebody with some analytic prowess. So maybe there’s the opportunity for the book repository.
Gerhard Pilcher Yeah, man, thanks.
Evan Wimpey All right, so you’ve got the executive, you got an analytics leader, you’ve consulted across a lot of different industries. Given that you’ve got a good use case, if you’ve got something that you’re excited about and you think will work , it is a good place to employ analytics. You still got a lot of work to do to get it off the ground. Can you talk about what some of those challenges are or maybe what is typically the hardest challenge, maybe that people don’t think about? It’s not just building the model, it’s not just tuning some model, but actually getting a model, a predictive model off the ground. Sort of. What challenges do you see there?
Gerhard Pilcher Yeah, sure. You know, I always hesitate to say this because we’re an analytic company, but analytics is sort of the easy part. The date of time, the really the biggest challenges I faced are the beginning and the end of the process. What do I mean by that? The beginning of the process is really how we define the business problem and how we define it with enough clarity. And do we understand the value of solving that problem and how to break that down into the information or data that we need to solve it and how we might deliver that? Thinking about that from the beginning of the process, and typically that kind of information has to come from an executive or somebody fairly high up in the organization to make that connection between the value of what we’re doing and the value to the business. What I find is that it’s absolutely critical. If you don’t do that well, you may be fighting the wrong problem, not fine. Executives get frustrated with it one because their time is short too, because they think really fast and they’re frustrated when people don’t think at the same speed that they do. But thirdly, they get frustrated because they feel like you haven’t started, you know, like you’re not doing something. And so therefore you’re not making progress. But I’ve always found through all of my experience that if you take the time upfront to get that part right, whether it’s specifications or understanding that business problem, then at the end of the day, you’re directionally be in the right place and it’ll be a lot faster and a lot less costly than having to make course corrections toward the end and more expensive part of the process. And it’s not that analytics. This is not. The thing that’s a little bit different in mindset is that analytics is really a series of experiments to better understand the information and to see if it has a signal that you’re hoping for, the inference that you want or the clarity that you hope to achieve. And so you have to test that to make sure that it does. But you’re experimenting and as you’re experimenting, you want to directionally understand where you’re going. But also you might need to pivot and make adjustments to that. So this is the first part of the process. Is usually really hard, and I’ve got some examples where it was done well and the project went super well, and I’ve got some examples where it wasn’t done well and it didn’t go as well and so on. Statistically speaking, it is pretty well correlated that if you really take the time on the front end of the process, you get a better outcome. The other part that’s really difficult and you have to think about it up front, too, is where am I delivering this information? If you think about the whole purpose of analytics as we’re trying to take some information data we have, combine it with other data and gather some insight. But the insight isn’t. What’s important is how that insight drives some decisions that we’re going to make better. Maybe we ought to make that decision, or maybe we feed that into a higher level strategic decision that could be millions of dollars to a company. And so thinking about that delivery mechanism or that consumption of that to drive a decision. Most businesses are run around a set of processes or systems, which are processes and skill sets. And so if you’re adding some new piece of information thinking about how that’s going to be consumed in that process, and even if that process can actually absorb that new piece of information, if you can actually make that decision, that system is capable of it ahead of time. So you can begin that planning, you can begin that communication, you can begin to modify those systems that are going to need to change to consume that information ahead of time. But most companies don’t want to invest in that. And it’s amazing to me because over the years, you can look at bigger companies and, you know, all the systems that they buy, they think that software or technology somehow is the answer. But software is just a tool. It’s really how you apply that to within the business that matters. And yet we seem to forget that every time a tool is purchased, millions of dollars might be spent and put into place and there’s no follow up training or changes in process, and the tool just sits on the shelf or gets utilized maybe a tenth of the actual capability that you purchased. And it seems such a waste that we don’t think about that more often than up front. We just continue to believe that that tool is that easy button sort of technology is the cocaine business. People give me more technology and all my problems will go away, but it’s never really the case in my experience. But this is the second hard part, and I tell people it’s really the long pole in the tent. And if you’re not willing to invest in it, then it’s going to be very tough. I’ve just recently met with one of our really great clients and their acting chief data officer, but I’ve told them this several years ago about a product they were buying and technology they were buying. I said, Well, we need to understand the current process and then how that’s going to change, what the new process looks like, what skill sets? No, no. We’re just going to buy the software and the software was purchased and it’s good and it’s provided some value. But now they’re realizing, Hey, if we don’t really dig in and change processes around how this software, it’s going to be used, then we’re not getting the full value out of it. So now we’re going to exercise. You know, fast forward at that wisdom, maybe it would have been helpful to put in place in the very beginning. But, you know, everybody has to at some point internalize some of these concepts for themselves and see them in action for it to matter to him. But that’s why it’s hard. Some people just haven’t been through that and seen it. What success could look like by spending that effort and time up front
Evan Wimpey Yeah, certainly, certainly a hard lesson to learn, and it’s easy. It’s easy to see the siren or hear the siren’s call. Do you see the siren? You hear the siren call of the. Problem solving technology, what we have to do is buy this and then boom, we can implement this solution. Yeah, that’s a really, really good point here. And I’m curious, you mentioned putting the time in effort at the front end to really understand the business and the problem that you’re trying to approach here. You mentioned pulling an executive and who can really help drive the type of decision you’re trying to make. You’ve seen successes and you’ve seen failures aside from time to time spent on the front end of an analytics project. Is there anything else? Are there any commonalities there that you’ve seen in successful projects? Certain questions that they ask are certain things that they tend to focus more on.
Gerhard Pilcher I mean, it’s really different for every business and sometimes within the business, for every team within that business. Where I have seen the most clear. Success is when the team is willing to have their norms challenged, because a lot of times data is revealing something not intuitive to our experience. And that’s an interesting find. That’s one of our most interesting aha moments. Or we call it, Oh, wow. Yeah, well done, right? And so exploring why that is and why it’s counterintuitive to us sometimes leads to some of the biggest breakthroughs. But I find that sometimes the people who aren’t willing to be challenged can sometimes miss a great insight and opportunity there. So yeah, that’s that. Probably the next biggest thing to success as well and to be challenged.
Evan Wimpey Yeah, I think I think that’s a great point and probably true in more than just analytics as well. If you’re looking to make change, you’ve got to be willing to make change. And so I think you’ve given some really good guidelines here. This, again, is the first episode of this Mining Your Own Business podcast, assuming they don’t pull the plug after this one. We’ll have more episodes, and we will talk to some analytics leaders and some executives that have done just this and have tried to get some analytics initiatives implemented off the ground and driving some real business change. And as we talked to them in the coming weeks and months, is there anything in particular that you would want to glean from them what their secrets are or what their lessons learned are?
Gerhard Pilcher Yeah, I think there are probably three areas I’ve been really curious about. One of them is their successes and failures and driving adoption and doing the change, planning what we call change planning early in the history of our company. We initially just delivered an equation which was the model, right? That’s all the model is. It’s an equation that creates another piece of data. At the end of the day, the more complex one is just a page of equations instead of a single equation, but essentially an equation. And then we realized that wasn’t really effective until we started building software systems to help. Deliver that to help somebody consume that, and in a way that was intuitive for them, not from a. Quantitative point of view, but intuitive to the way they did their work, and that really helped us and got us to another level of adoption, and then finally, we realize that without some change playbooks, without some approaches to how change should happen, especially around new information and decision processes, we can be successful. So we’ve built that third leg onto a school on our virtual stool for elder research, along with the training platform that allows us to do the data literacy and training necessary to support change. So I would be interested in these leaders, what they’ve experienced and changed and some of the secret sauce, or some of the things that have helped them drive change and convince people of the need for this change in investment in that early so that you can get the most value from your analytics. That would be one second would be, you know, as they’ve experienced this and gone through this, why have leaders been resistant to it? What is it that we’re failing to learn or failing to communicate that leaders continue to resist investing in change? And finally, what they’ve learned about transforming organizations and to ones that continuously reinvent themselves, right? You know, I know that we have that as part of our service model, reinvent continuously and we’re constantly on a quarterly basis looking at the things that we can do to improve. But I’m curious about how leaders are helping other companies do that. One of the interesting things about being an interim CDO is a friend of mine. Dr. Peter Aiken wrote a book on data strategy, and he says most CDOs, the first CDO of a company, are usually fired because they have to push really hard against the organization, and they don’t make a lot of friends initially. So he said sometimes it’s better to have a consultant be the CDO because they’re easier to fire and they can lay the groundwork for the next idea to come in and have success. So all of that, what I just was talking about is sort of wrapped up in the reason that it’s difficult for CDOs.
Evan Wimpey yeah, certainly a lot of a lot of inertia carrying an organization in one way. And if you’re trying to make changes, you’re trying to continuously reinvent the inertia. It’s not always with you there. I’m curious, are you as interim CDO or are you the initial CDO? Or are you replacing a previous CDO?
Gerhard Pilcher I’m the the initial CDO so I’m in line to be fired here.
Evan Wimpy All right. Don’t take it personally. You were hired to be fired.
Gerhard Pilcher to be fired. Yeah, I’m pushing hard.
Evan Wimpey That’s really exciting. OK, we’re going to wrap up here shortly before I want to ask you. I want to ask you one more thing. And I do think it’s worth pointing out it’s it’s funny. We’ve talked about analytics and how the challenges people have and how we try to overcome those. And we had a completely non-technical discussion there. There are very few things that are technical focus that are real driving challenges. Sort of change management, having norms, challenging making, making organizational changes. So I’m curious if if you or or a consultant that works with you at Elder research, if they’ve they’ve got something, if they got a business use case and they can, they can press the magic button and all of the stakeholders, from the executive down to the practicing data scientists and data engineers, they’re aligned there. They’ve got the same vision they’ve got by and they’re ready to execute. So you don’t really have those change management issues like everybody is on board ready to go. Do you have anything that this is what I would want to tackle, but maybe it’s too much in the real world. But with the magic button push, then yeah, let’s use analytics on this one.
Gerhard Pilcher I’d love to have that magic where we’re looking for it in general. I think, you know, it’s what I mentioned earlier. I would like to shift the conversation to information away from analytics. Analytics is a very amorphous term and means different things to different people. But I think if we talked specifically about business and information that people need to make better decisions, what should managers do? If I only knew this, I could do that. And so really change the conversation in that way with analytics just being a tool to help derive that piece of information that they need in general. But if you want to get more specific, you know, and this may be my retirement, I don’t know if I could really have a project like that. I would love to apply analytics. And information, better information to education, especially K through three. I think if I look around the world and look at how the world has been transformed over the last 50 years or even 100 years, education really hasn’t been transformed at all. It’s really where it was. And I think especially in younger children, k through three, especially is what sets the precedent for whether a child, if they develop the math and reading skills at those ages, then they’re much more likely, significantly more likely to graduate from high school and succeed in our society. And so focusing on delivering those skills outside of what I consider artificial grade level eight grade levels is bad for a lot of reasons, one for really smart students. They tend to slow them down in the areas that they’re gifted in. Like for me, I would have accelerated in math and I would have been slow in English while artificially putting a grade around that forces me into just so much skill level. I could have used more language skills and or more time in language skills, and I could have advanced faster in math. Others would be just the opposite. And on the opposite side of it, slower kids today in education are passed to a new grade level without the requisite skill. So why not think about education two or three more like little pods of skill development and look at almost like a gig economy magic students with teachers letting them stay until they’ve mastered that skill? You can. Another analogy might be a martial art where you stay at a yellow belt or a white belt until you achieve a certain skill level. Then you’re a yellow belt and you stay there to achieve a certain skill level. But nobody fails. Everybody just moves along a continuum at their own pace so that we ensure that kids, at least by the third grade, have the requisite math and language skills that they can succeed later in their education, rather than just passing them along in that, in that way. And I just think we just need to reimagine education and we can learn from children and help them discover where they’re gifting is and therefore better match them to teachers and skill development so that they their success is ensured rather than uncertain as it is today, and information can play a huge role in making that dream come true. So I’m gathering people now, and it’s something I’m thinking about. I don’t know if we’ll be able to do it without elder research, but I’m certainly going to try to pursue it in my retirement one day. If I retire, that’s another question. So I’m gathering teachers and other interested data scientists right now and people who have experience in these things. But you can see that getting alignment across all the stakeholders and something like that is going to be huge, huge. Problem or challenge?
Evan Wimpey Yeah, absolutely. And sorry, you’re there, certainly no easy button on that one. Talk about some stakeholder inertia. There’s a lot behind education. I think it’s fitting how you say, if I ever retire, I’m just instead, I’m going to do so. Good. Good. Try to solve one of the world’s hardest problems instead. Nice. Nice, easy retirement. Kick the beat back. A very, very good answer. A very noble answer, and I think certainly highlights how important the stakeholder buy in and stakeholder alignment is.
Gerhard Pilcher That certainly does. Thank you for your time today, and thank you for inviting me on.
Evan Wimpey Absolutely. Gerhard, we’ve been very pleased to have you on the show, our first ever show. Folks can see Gerhard in the work that he and his team do there. Research Gerhard. Appreciate your time today. Thanks very much.
Gerhard Pilcher Yeah, thank you. Take care.