A Transformational Power BI Success Story at Medtronic w/ Luciano Miranda

Rob Collie

Founder and CEO Connect with Rob on LinkedIn

Justin Mannhardt

Chief Customer Officer Connect with Justin on LinkedIn

A Transformational Power BI Success Story at Medtronic w/ Luciano Miranda

This week, gear up for an episode packed with insight and excitement! Our guest Luciano Miranda, Vice President of Analytics and Insights for Global Operations and Supply Chain at Medtronic, took us on quite a ride this week. His globally dispersed team helps inform critical decisions across Medtronic’s far-reaching supply chain using the power of data. What began as a scrappy five-person team doing their best in Excel has blossomed into an analytics powerhouse delivering insights that assist over 90,000 Medtronic employees.

But it wasn’t easy getting there. From adopting new technologies to overcoming changes in organizational structure, Luciano kept his focus squarely on delivering value to stakeholders. By putting relationships before process and always taking the “learner path” over the “judger path,” he turned obstacles into opportunities for growth.

The future holds even more change as AI promises to amplify our capabilities. But Luciano sees a bright future where these technologies enhance the power of dashboards and the insight of analysts. He chooses to ride the waves of change rather than be overcome by them. The human element still reigns supreme.

As always, if you enjoyed this episode, be sure to leave us a review on your favorite podcast platform to help new listeners find the show. And be sure to subscribe for new episodes delivered weekly to your inbox!

Note: Everything we talked about is for non-medically cleared or regulated applications, simply internal analytics.

Also in this episode:

Medtronic: Microsoft Customer Stories

Formula 1 Ted Talk

“Principles” by Ray Dalio

Spotify and AI

NBA coach Doc Rivers and Analytics

Episode Transcript

Rob Collie (00:00:00): Hello, friends. We have what I think is an amazing conversation to share with you this week because Justin and I sat down with Luch Miranda, vice president of Analytics and Insights for Global Operations and Supply Chain at Medtronic. Now, Medtronic is a massive company. Most of our clients here at P3 are very squarely in the mid-market range, and Medtronic definitely is not mid-market. They're a medical devices manufacturing company, an international company, and get this; every second of every day Medtronic helps two patients worldwide. Every second. So that clearly is a global enterprise. And they are a client of ours, but when you compare them to our mid-market clients, well, there are some things that are different and some things that are very, very much the same. Here's how I would describe the difference between enterprise organizations and their data projects as compared to mid-market organizations and their data projects.

(00:01:00): If you rewind, let's say 10, 15 years or so, the enterprise market had an advantage. They had capabilities just because of their size, their economies of scale allowed them to tackle projects that the mid-market simply couldn't even really think about. But with advances in software, particularly from the Microsoft corner of the world, the mid-market is no longer locked out of these things, not at all. And so, paradoxically, the difference today between enterprise data projects and initiatives at mid-market organizations, is that enterprises have all of the same problems, all the same challenges that mid-market organizations need to overcome, but then they also have extra problems that are sort of uniquely brought on by the massive scale at which they operate. So yeah, that's right. It used to be that mid-market companies had a steeper wall to climb than enterprises did. That's kind of inverted with the advances in software.

(00:01:55): So we're sort of in the beginnings of a golden age for advancement in the mid-market. But it's not like things have gotten harder for enterprises. Things have gotten much, much better for them as well. And when you bring the right team and the right attitude and the right focus, you can get some truly spectacular results. And working in tandem with us at P3, Luch and his team have delivered precisely that level of spectacular impact. So spectacular in fact, that it was the subject of a recent joint case study by Microsoft and Medtronic. It's a case study primarily about their success with Power BI. Luch is quoted in this case study. I read it, I thought it was amazing. My only complaint with the case study was that they were very careful not to mention any consulting firms that were involved. We'll link to that case study in the show notes, by the way.

(00:02:43): So I mentioned that some things are the same and some things are different between mid-market and enterprise. Let's start with a few things that are the same. For example, just a few short years ago, everything that Luch and his team were doing, they were doing in this tool called Excel. Luch himself was and still is what I would refer to as an Excel pro, you know, the VLOOKUP and pivot crowd, my kind of people. Another thing that's the same between mid-market and enterprise is that what we at P3 call a faucets first approach remains the way, and we talk about that. Another thing that's the same is the importance of working backwards from one's stakeholders. What is it that the people using your dashboards, what is their daily workflow? What are the things that they are truly trying to do? What are the things that they should be doing?

(00:03:31): Don't just go forward from the data and build reports. Work backwards from the impact. Work backwards from the people, work backwards from the workflows. And that theme even ties in with last week's sort of short form pod that I did about listening to your stakeholders. That's a topic that's just going to be evergreen, isn't it? Of course, some things are quite different. For example, in order to air this podcast, I'm obligated to tell you the following. Everything we talked about is for non-medically cleared or regulated applications, but instead simply for internal analytics purposes. Another thing that's different, Luch's team's Power BI footprint reaches 4,000 active users at a rate of about 7,500 report views every workday. He also now has a 50 person team, which is of course an asset in that he has 50 people to help him, but also of course introduces scale and organizational challenges, which require a lot of deliberate thought planning and communication.

(00:04:32): And with challenges of that scale, I think as you listen to this conversation, you'll see that Luch very, very much is the right person for the job. Just such a deeply, deeply thoughtful individual. I can't begin to capture or even tease all of the things we talked about in an intro of any reasonable length. No matter what size organization you work at, I think you'll find this was a conversation packed with applicable wisdom. So let's get into it.

Speaker 2 (00:05:01): Ladies and gentlemen, may I have your attention, please?

Speaker 3 (00:05:06): This is the Raw Data by P3 Adaptive Podcast, with your host Rob Colle and your co-host Justin Mannhardt. Find out what the experts at P3 Adaptive can do for your business. Just go to P3adaptive.com. Raw data by P3 Adaptive is data with the human element.

Rob Collie (00:05:31): Welcome to the show, Luch Miranda. How are you today, sir?

Luciano Miranda (00:05:35): Fantastic.

Justin Mannhardt (00:05:37): Yeah, let's not ruin that.

Rob Collie (00:05:38): First of all, thank you so much for being here. I think you know Justin pretty well?

Luciano Miranda (00:05:41): I have the pleasure of meeting face to face and nice to see you Justin.

Justin Mannhardt (00:05:44): You as well.

Rob Collie (00:05:45): Luch, just tell us briefly what your current job title is. Let's get all the particulars out in front.

Luciano Miranda (00:05:50): I am the Vice President of Analytics and Insights for Global Operations and Supply Chain for Medtronic.

Rob Collie (00:05:56): Medtronic.

Luciano Miranda (00:05:57): Medtronic is the largest medical device manufacturers in the world, so we're very fortunate in what we do, we're about a 30 billion organization. My piece is global operational supply chain, which is about supporting 45,000 of the 90,000 plus employees that Medtronic has.

Justin Mannhardt (00:06:12): Small job.

Luciano Miranda (00:06:14): We have a stat that basically we help two patients every second in the world.

Rob Collie (00:06:18): Wow.

Justin Mannhardt (00:06:19): That's like a core KPI at the executive level, right?

Luciano Miranda (00:06:22): It is. It's written up actually on the wall to remind us of what we do, is when you're in analytics and so on, you're looking at numbers, you're looking at data, et cetera. But there's actually a real human being in our world, at least on the other side. Gives a lot more meaning to what we do.

Rob Collie (00:06:35): How broad is the product catalog? You make five things or do you make 75,000 things?

Luciano Miranda (00:06:40): More on the latter? We have a multitude of products on various different parts of the body. So the cardiovascular area, pacemakers really cool, like some pacemakers now that is smaller than your pinky finger. Pretty, pretty amazing stuff. So pacemakers, things like ear, nose and throat, neurovascular. Our neurovascular division for example, produces products that can, if caught early enough, can help minimize the impact of a stroke, which is massive.

Justin Mannhardt (00:07:10): Yeah, that's great.

Luciano Miranda (00:07:11): And all the way through to our surgical robotics. Our robot is called Hugo. He's really cool. Broad, broad range of products from very small, tiny devices to obviously the larger ones such as the surgical robot.

Rob Collie (00:07:24): External devices like machines that you would see in hospitals, as well as things that go inside people.

Luciano Miranda (00:07:31): Exactly. So they're external devices and internal devices.

Rob Collie (00:07:35): Is there any chance that some of the titanium that's in my body from injuries and stuff might've been made by Medtronic?

Luciano Miranda (00:07:41): You have something with your spine that you got an operation or so, it is a high probability, yes.

Rob Collie (00:07:47): Okay. All right. Well, I haven't had that yet.

Luciano Miranda (00:07:48): Good.

Rob Collie (00:07:50): So we all know people who have interacted with Medtronic products within the last week. So you answered that Medtronic is closer to the 75,000 product end of the spectrum than the five, right?

Luciano Miranda (00:08:03): Yes.

Rob Collie (00:08:03): And you said that you're in charge of analytics for the entire supply chain.

Luciano Miranda (00:08:09): It's more than supply chain, right. So there's global operations and supply chain, so that is supply chain. Which is supply planning, demand planning, inventory planning, transportation distribution and trade-offs. And we also support the supplier management organization, both indirect and direct purchasing organization. We support strategy group, which also includes enterprise risk and facilities. We also support the manufacturing organization. We support our global operational supply chain finance. And we are also supporting the global operation supply chain quality as well as IBP, as well as some analytics on how to use inventory with commercial information, et cetera. So the breadth is quite large. How we've grown from a small team of five folks that were playing around with Excel at the time, to now much more professional organization.

Justin Mannhardt (00:09:02): I hear you describe thousands and thousands of products, global organization, international operations. You work in what I could easily describe as a very complex environment. Complex organization, complex supply chain, complex technology estate, complex groups of people. I'm just curious, as a leader, how have you managed to simplify that, create the focus for your team so that you were able to grow and maybe tell a little bit about your story, about how you were able to build the momentum of success to get where you are today?

Luciano Miranda (00:09:36): It all started with a TED Talk. So this is going back maybe five years or so. And this TED Talk was about Formula One. I've always been really enamored by F1 and how they use it, the real time, how they monitor their car, how they have all these sensors. And this TED Talk talked about how they used their analytics in order to save babies' lives. It really got me thinking, "If F1 is doing this, what if I use that as a juxtaposition to the supply chain?" At the time, because we were just doing a piece of the supply chain, right? And saying, "If F1 could do this, why can't we do this?"

(00:10:13): Now, keep in mind at the time the group I was with was very, very immature in the sense of analytics. So it took a very long time even to get a basic inventory report. And our tool was Excel. We were not necessarily data scientists, we were not necessarily of IT background. We were folks that really cared about the business and wanted to make an impact. And that's how it started. And started making partnerships, started learning about the technologies, the amazing technologies that the IT team had purchased for Medtronic to use, and then started to think about how we leveraged them. When we first started team of five, and there's been reorganizations that happened within Medtronic. We have been big benefactors of those, right? Because what happened was it says, "Hey, you're doing some really cool stuff with Excel."

(00:10:56): So we always focused, we did it a little bit backwards to be honest with you, that from the traditional, a lot of the traditional approach, which I know people out here might not agree with me, and I won't even disagree with you, but you've got to have your backend data first. Data has to be clean, then worry about the front end. We did it backwards, and I'll explain why.

(00:11:16): Because we also didn't have much of a budget. We needed to show people, I love this line. I learned it from a sales course I once took and they said, "The best way to sell is show how you solve." And that stuck with me because I'm like, "If we want to get investments, we need to show people the value that we can give them." With, even if it's an Excel spreadsheet, the user doesn't care. What the user cares about is does it work? Is it accurate? Can I trust it? But then we run into a problem, because then we were getting too big. We learned that no matter how much memory you have, 32 bit Excel will stop working once you use two gigabytes of memory. So that was a big learning for us.

Rob Collie (00:11:57): I've been there.

Luciano Miranda (00:11:57): You've been there?

Justin Mannhardt (00:12:00): Everybody's like sending help desk requests like more RAM.

Luciano Miranda (00:12:03): Yeah, you need 64 bit Excel. We're like, oh.

Rob Collie (00:12:06): Did you ever apply the hack to the Excel exe that allowed it to address four gigabytes of RAM? You could set a bit on the Excel exe file that would tell Windows, "No, no, no, no. Let me access more RAM." And it would, but Microsoft wouldn't bless it. So you probably didn't do it.

Luciano Miranda (00:12:24): Didn't do it, didn't do it.

Rob Collie (00:12:27): And even that extra two gigs was a lifesaver.

Luciano Miranda (00:12:30): Oh yeah, because you were so excited. You were like, yeah, I got a million rows now.

Rob Collie (00:12:34): Yeah, feed a dose of VLOOKUP and it's all over.

Luciano Miranda (00:12:37): It's all over. And you see the thing spin slowly, slowly, right? It's almost like I'm going to date myself, when we had dial up internet and it was like, beep beep beep.

Rob Collie (00:12:46): Yeah, you were hitting the limits.

Luciano Miranda (00:12:48): It was a good thing, right, because it forced us to look somewhere else. Because we were all, I remember one of the folks I worked with and is still in our team, he always tells me, he goes, "Luch, every time we think we're good at something, we realize we're not." And every time we get comfortable with something disrupts us, and we have to learn more again. And we were getting pretty good with Excel. Good enough to know that we were not experts, but good enough that we could use a lot of the functionalities and do some pretty amazing things with it. But there is a problem right, now we have too much data, so what do we do?

(00:13:18): So what we ended up looking was start really assessing other tools that were there, and there was Tableau at the time and Power BI and Spotfire and all these other ones. And I remember having a conversation with one of our IT leaders saying, "Look, this is what I have today. I have a series of folks that are very resourceful, but they don't know how to code. They don't know how to do SQL. We just know how to do Excel and do those things that we do." And what he said, he goes, "Look, Luch," he goes, "Don't do it yet. Microsoft has come up with something called Power BI, but it's probably about a year away from it being really good where you can do all the stuff that you want to do." Again, if he's listening, I'm sorry, I did not listen to him. And immediately as soon as I hang up, I downloaded it.

Justin Mannhardt (00:14:05): That's the way, yes.

Luciano Miranda (00:14:05): I watched a little YouTube video and literally within an hour I have my first dashboard. And then I start Googling and I see how easy it is to transfer your Excel skills. I'm like, "Huh." Now I have another issue, right? Because there were tools at the time that were better than Power BI, no doubt. But the challenge was my budget at the time was tiny. To put in perspective, I think that the total budget for the year for me to do any consulting or anything was under a hundred thousand dollars. So you're not doing much with that in that sense. So I'm like, "Well, you know what? We have a series of Excel folks, let's convert it." So I remember going and saying, "Everyone, we're now going to start using this tool. Let's learn it."

(00:14:44): The first early models were very pretty in the front end, not so pretty in the back. And I'll share another hack that we did, because none of us were really UX UI designers. So we're like, "We need to make this look good." So what do you? You go to Google or whatever search and you say, "Show me the best dashboards in the world." And to start again, our backend data was not great because backend data requires investment. You're cutting that loop where it is like, how am I going to get the investment? I have to show value. To show value, I got to show something. Otherwise, there's just buying up a PowerPoint. It wasn't going to happen.

(00:15:17): So that's how we kind of grew. And to be honest with you, our focus has always been almost zealot about satisfying the people we serve, and then it's just every year it's just getting better and better and better. So then what happened was, and there's another reorganization, and now they basically, Medtronic was pretty good with analytics. You had a group, but they were siloed. So just for a group. And then there were other group that had just as good analytics or better in some areas, and the same with other groups and other groups. So there were two folks that we owe a lot of our success to that basically just said, "Hey, we need to have a central supply chain." This is literally only about two years back. They asked me, "Hey Luch, we want you to run this team. What we're going to do is we're not necessarily going to add people, we're just going to take the analytics people that are already in these other groups and bring them together as one. Your guys' job is to basically create analytics, but that are global in nature."

(00:16:13): When you're doing this, not everybody loves centralization. In essence, there is a sense in the organization at least that, "Hey, I'm going to lose my ability to create my own analytics. I'm going to lose my autonomy." A lot of these problems that we deal with, sometimes they're very basic human problems, or human issues, and I think you identify them. So we worked our tails off to create the supply chain analytical ecosystem. Now the other problem was how does things look and feel? At the time, I think Medtronic had over 70,000 dashboards.

Justin Mannhardt (00:16:51): That seems low for a company of that size.

Rob Collie (00:16:55): It's one per product.

Justin Mannhardt (00:16:56): It's one dashboard, yeah.

Luciano Miranda (00:16:59): You look at this and you're like, "Okay, well here's another problem." Doing sales service analytics is not that difficult. Anybody can build a dashboard. Now, the quality of it, a different story, right? When we came in, there were 12 back order analytics. All pulling from the same source, but all of them making a different join, a different relationship or a different assumption. So the numbers were close, but none of them were the same. And we're like, "How do we solve this?" And then how do we tell people that the analytics that they're looking at, it's coming from this group, that is supported, et cetera? We said, "We need to brand this."

(00:17:35): So we created a brand, like internal brand. We basically said, "Anything we build", we came up with all kinds of names, "Let's call it Insights." Now the other thing that has happened now is as you amalgamate the groups, is the skill levels that came together. My group was more on the business side. There was another group from another area that were legitimate coders. So now we're all on the same team. We all just got so much better together than we could have ever been on our own. So we come up, we talk to Microsoft, we're like, "Hey, we're obviously the customer of you guys, can you help us with the UX UI? You guys have some beautiful analytics. We just want to talk to one of your designers." So I can tell you it was extremely humbling. We thought we were good. Right?

Rob Collie (00:18:20): Yeah.

Luciano Miranda (00:18:21): So after we recovered, now, we basically created a template. So what we did to come up with the view and what this was going to look like, because people get really enamored about how their dashboards look and feel. We need to solve this, right? So how are we going to do it? Well, the best way, I had just read the book by Ray Dalio called Principles. So I'm like, "We need principles." So we wrote 15 principles on what this needed to do. One of our principles, we still to this day follow is that when anyone looks at any Insights analytics, the look and feel is exactly the same. We just wanted to give people a standard experience and then at the same time build the brand. Through that year, we built probably over a hundred analytics to the point that we almost burned ourselves out. I mean, not almost, we did burn ourselves out.

Rob Collie (00:19:07): Yeah.

Luciano Miranda (00:19:08): Meditation really works, right?

Rob Collie (00:19:11): Yep.

Luciano Miranda (00:19:12): So exercise really works. And having an amazing family that supports you as you're doing these crazy things. Then there was another reorient to say, "Hey, this worked really well. Can you now do it?" So at the time we were just doing well, I say just right, we were doing supply chain. We want to do the same thing, but now for global operation and supply chain. So we had a CEO that came in who obviously we owe a lot of our success to. His name is Greg Smith, and he basically just said, "Hey, one best way. And you guys are doing some good stuff. So from an analytics standpoint, let's create a one best way for GOS analytics." Because the executive team wanted to stop, you would go to some of these meetings prior and a big part of the meeting was discussing about which number is right, versus-

Rob Collie (00:19:58): Totally.

Luciano Miranda (00:19:58): Right? If you're having let's say profitability issues on a product, somebody may say, "Well, mine is 89.3." The other person said, "No, no, no, it's 89.7." I say, "Who cares? The number starts with 89. Call it 89.5. Move on. Talk about what is the real issue." So part of, I think the success that our team has was also our executive team, when I look at it, because they basically started saying that anything that comes in for them to discuss has to be Insights. So if anybody brings a different number, they won't even look at it. For us, we have to be humble enough that if somebody makes a better analytic than we do to actually bring it into the ecosystem and create that innovation. Because there's obviously so many people that are doing some self-service analytics.

(00:20:42): But the results are pretty cool. When I look back, this is from day one from supply chain, right. There were about 15,000 views a quarter. When we look at now, we get 450,000 views per quarter. Just under 4,000 active users, and we're trying to grow that. So I think we're at 35 to 3,800 right now, active users. So it's been successful. Creating that holistic end-to-end view where you have analytics that can be summarized at the executive level, but that same analytic with different views can be brought down to the lowest of levels to take action.

(00:21:21): We started with a different approach that we switched from that now, in the sense we will take, we call it technical debt. It's most viewed purposely, and we actually log it so that we say, "Hey, we are taking technical debt as we're doing something faster than we should because our backends are not ready." But as we did do that, obviously our funding increased because people were willing to invest in us because we're giving them value. So yes, it was a little bit unorthodox at the time where we said, "Hey, we're going to start actually with the front end first and the backend later." Now, have you ever opened the hood of our backend?

Justin Mannhardt (00:21:53): No, I'm good.

Luciano Miranda (00:21:56): No bueno, right? No good. But the users didn't care. There weren't massive models at the time. Today we have one model I think has just reached a billion records. There's no way you get away with a bad backend. Now we have the backing, the funding that is needed in order to produce professional analytics. Like right now, we can't get away with doing what we did three years ago. It just wouldn't work. But when you're getting started, just start, and build something, show value. It's much easier to sell and get the necessary investments, because you do need a solid backend, from a position when you've already showing value versus a position of what it could be.

Rob Collie (00:22:35): Hey, Justin, what do you think of this unorthodox approach?

Justin Mannhardt (00:22:39): I don't know if you caught Rob and I smiling at each other because we think this is nailing it. This is the way, for people listening, the way Luch is describing, and we call it faucets first. Start where the people need water, what's the value? And then Luch was describing, then you can go back and say, "Okay, they want more water. I need some money. I need people. Then I can run a bigger pipe, to make this scalable." And I think your story there, Luch is a great example of things companies struggle with, especially things like adoption and governance.

(00:23:10): You've been successful in those areas. You talk about the extremes from the completely federated self-service model, all the way to the other extreme of a very centralized model, and sort of the way you've had to find the balance between that and what solutions you take on and take over, and what you learned from the community. You've had success with that, but it's been a result of y'all focusing on the value and as those teams come together, putting the right people in a position to be successful, to keep the snowball rolling. So that's a pretty cool story, man.

Luciano Miranda (00:23:44): You have to be proud of it, right? And again, it is not only because of what we did, but also how and the people, the group of people. I mean, I've been so freaking lucky to get the leadership stuff and the staff holistically that we have. That has that mindset of they really, really care. Doing analytics, obviously if you're an analytics team, you have to develop good analytics. But it's how you develop it and how you interact with your stakeholders, and the service that you provide them that ultimately is what they will remember. People won't remember what you did for them or how you did it, but they will remember how you made them feel.

(00:24:20): Especially in the world of analytics, we are going to make mistakes. In a complex environment like ours, with the amount of data and products and different manufacturing plants, it is possible that sometimes you do something very fast and it's not a hundred percent accurate. So having that good relationship with the people you serve, it's so important, because they will have your back when things are not great and they will help you, right? They will tell you what it is.

(00:24:45): So now obviously we have a playbook. We're obviously a more professional organization from our humble beginnings, but it didn't start that way, and we can't forget that. We always say we have really good processes, but the process is not the customer. The customer is the customer. The process is there to help the customer, but don't get so enamored with your process because you'll start satisfying the process and you start forgetting who are the true people that you're serving, right?

Rob Collie (00:25:11): It's so easy for something that's a means to an end to become its own, its own end, its own goal. Especially when it's heavy and intense. This is the old story with data warehousing. It was a means to an end, but it was so difficult that it became its own culture. It was easier to get further and further removed from how the data was going to be used, and more and more focused on getting the gleaming monolith correct, and not really thinking as clearly about where the people come in. When you were still just the five person team in Excel, roughly what year was that? Our organizations met and started working together I think in late 2018. Where were you relative to that date? How do those timelines line up?

Luciano Miranda (00:25:55): It's probably about one or two years back. It's not that far. The growth has been pretty fast from the day we started. Microsoft actually recommended you, and I think we had budget for just one of your people. Obviously we have more than that now because you've earned it. It was one or two years after that.

Justin Mannhardt (00:26:12): I don't know if you know this Luch, my first client assignment was to come over to headquarters and teach a class.

Luciano Miranda (00:26:22): Oh, I remember that! That's right. We hired you folks as well to teach our developers.

Justin Mannhardt (00:26:27): That was the very first thing I did for P3 Adaptive was drive over to Medtronic and teach a class.

Luciano Miranda (00:26:33): That's amazing. How that was, because I remember the feedback was fantastic, that people were really impressed with how you guys work, and obviously the knowledge they got. So you were a really good fit to our culture. Because we try to make the people we serve better and you're trying to help us be better. You know we treat the folks from your team the same way that we treat our own employees. We don't make any distinctions between our contractors, because while they're playing in our team, they're part of the team. That was really good. You did a good job, Justin.

Justin Mannhardt (00:26:59): Not bad. It was my first one. So it's fun like the parallel journeys, so you're hearing about how you guys have grown over the last five years as well. But it was fun, and I remember various people in the room from different areas of the company, and I think your story is, it's a perfect example of what happens when you put a really good tool in the hands of really passionate, curious people that want to help. And their eyes light up and their mind starts spinning, it's like, "Oh, how can I do this? Or how would I do that?" Right? And then you see the fruits of that over 5, 6, 7 years. Pretty incredible.

Luciano Miranda (00:27:33): At that point, we had gone from a team of 22 analytics people or software supply chain, to now a team of over 50 Medtronic employees. So a pretty large one. Again, no new folks that were coming, that were already doing analytics. So I remember the biggest concern at the time when we were building the organization was the more you centralized, the more you potentially get away from the people that use the products you use. So I know that there's a lot of write-ups around this, about centralization, decentralization, hybrid model. One thing that's working very well for us is almost getting the best of centralization because all of these minds that are coming together. Folks that maybe were in one area as a lone wolf, and now you bring them with 10 other lone wolves, that are wolves diverse in thinking, but similar in passion. And you watch these folks get better. It's amazing.

(00:28:31): We had, last year, we had a bunch of issues with data quality as an example. That opened up to us now running over 3000 automatic data quality checks, as data moves through our system. I can tell you if we had not had amalgamated and gotten the power of the centralization, there's no way we get that. But this is all innovations that are coming internally. Now you have to be careful because a lot of folks still want the rapid response, et cetera, because again, analytics is not difficult to do on your own. You can grab a bunch of data, download, put it together and create a dashboard. It takes one to know one. That's what we did when we started.

(00:29:11): So we operate under what we call a squad model. Dedicated sub-teams of different groups within my organization that go together and are dedicated to inventory, to supply planning, so that you can still give the customer that reality. That they have a team that's dedicated to them. And I mentioned this in case there are anybody here that's thinking about that. That is absolutely a challenge, because as you centralize, you can't get too far away. So whatever you do, just make sure that whatever design you come up with, you keep close tabs and you give those functions that you're supporting almost as if we are embedded into their teams as if we're part of their team. It's a little bit different than they have in Spoke, but has been very successful for us in our journey.

Rob Collie (00:29:57): Just all the context switching, if you have a global team, a centralized team, and they don't have a center lane that they're assigned to. If everyone's just sort of like a pool, and they're talking to a different corner of the business every day, you don't get that opportunity to build up the tribal knowledge, to build up the relationships that it takes to do a good job. So you need people with some consistency. Even though they're on a centralized team and they can share ideas with each other, you get that cross-pollination you're talking about that is so valuable. It's similar to Hub and Spoke, right, in that you've got a focused area for each squad? If I'm understanding it properly.

Luciano Miranda (00:30:33): A hundred percent, a hundred percent. And you have to have that, right? Because part of the value add of the team is not the customer comes to you and says, "I want you to build me this. I want you to build me the KPI here, with a trend chart and a bunch of summaries." We actually have a playbook that is very detailed in all the steps, but the first step is, "That's great. Maybe that's exactly what you need, but tell me what business questions you want to answer." And we have a group within our structure that's basically their main job. We call them our architects. These are the folks that if you're building a house, you start with an architecture of, what is it that you're going to end up building? And in order to do that, you need to understand your customer and what do they really want?

(00:31:17): Because they may say, "Well, I want five bathrooms and seven bedrooms." And you may say, "Okay, but you're a family of three. Do you really want to clean five bathrooms? Is that really what you want?" "Oh, no, I just want to make sure that I have access to them." And say, "Okay, how about if we place them in this way, and you end up with three?" So, that's using a construction analogy, but that's the power of asking the right business questions from the beginning. And the truth is, even as a team, we realize ourselves, when an analytic is presented in our sprint review, so we also work in an agile environment. In the sprint reviews, you see an analytic and you're like, "Hmm, what business questions is this really asking?" And when the team doesn't really have them, you can tell it directly into the quality of it.

(00:32:02): And sometimes I think as developers and just wanting to get things done, you want to skip those steps because you just want to build something, right? You want to build something, you want to look at it and then enamored in your new baby that you just created. And you're not talking about weeks, right? You're talking maybe a few hours, if that, of really asking the right open-ended questions to really get, what is it that the people you're serving truly need?

(00:32:29): And you're not going to get it right all the time. We don't get it right all the time, right? A lot of times you build something and then you realize we got it completely wrong. But it increased the probability of building the right thing from day one. Anybody that tells you that, in my opinion, that it's guaranteed, probably hasn't done this. It's all about increasing the probability so that you minimize rework, and ultimately the customer's happy, right, because they get what they want from day one, versus then going through all kinds of redevelopment cycles, which are obviously more expensive and also very frustrating.

Rob Collie (00:33:00): And you also have an opportunity to not burn credibility. If your first result for them and you go "Ta-da! Check it out, dashboards!" And they look at it and they don't understand at all how this was going to help them. Whether they blame themselves or blame you, it doesn't matter. That's going to damage the credibility of that relationship. And I've experienced this, by doing it the wrong way many, many years ago. By taking the data that was given to me and sort of going forward from the data to produce visualizations insights for a client. And this happened to be just some family members of mine that I was helping out on the side. It was a really safe place to learn this lesson, back in like 2012.

(00:33:40): I gave them things that they didn't know what to do with. And they did, they blamed themselves. But in hindsight, if I'd worked backwards from what their decisions were, like what are their actions every day that they're taking already, that they're taking with low confidence, making decisions on gut instinct, what are the things that they're missing? Things that they should be acting on, but they're not. When I worked backwards with them from that, I produced a much more useful set of dashboards for them. That first round was almost the end of it. Because they thought, "Oh, well, now I've seen what this looks like, what this whole dashboard thing looks like that Rob's been talking about forever. I've seen what it looks like, and it didn't help me, so I'm out." That was a real possibility, a real danger of working forward.

(00:34:22): There's even a story in the NBA about this. When the analytics folks were sort of first breaking into the NBA, they told one of the coaches, Doc Rivers, he was talking about analytics. He says, my analytics team came to me and said, "Hey, did you know that Rajon Rondo hits like 20 miles per hour running in the open floor?" And Doc was like, "Get out of here. What do I do with that? It's not useful at all. That's why we have you? Is to tell me how fast a player runs in the open floor?" Think about how that translates into any of the decisions that an NBA coach needs to make to optimize their team, right? It means nothing.

Justin Mannhardt (00:34:54): Yeah. Rondo's fast. I knew that.

Rob Collie (00:34:56): You want to avoid the 20 mile per hour Rondo dashboard whenever possible.

Justin Mannhardt (00:35:01): Luch was describing this experience inside of a company, and it's remarkable whether you're a consulting firm, inside of a company, working as an analyst, the frequency by which someone comes to you and says, "Here, build me this. I've already drawn it, defined what I want." This is advice to anyone that's out there, ask those questions. Because as much as they think they know exactly what they need and what they want, you'll always find something where it's not quite on the right. The worst thing we've ever done is said, "Sure, we'll build that thing for you." And we go, and we build it, and it's on the nail to the spec, and they're like, "Ah, this actually isn't, I thought it would be cool, but it's not." Right? And so the power of those questions, it doesn't take much. Just sit down for a little bit of time, chat about it, you get to a much better place much faster.

Luciano Miranda (00:35:48): I think sometimes it's almost like our own mental block. Sometimes I think about it like, I want to go for a run, and the worst part is putting my shoes on and going out. And then you go out and it's actually not that bad. But in this case, it's not what I do, I want to build stuff, right? You'll save yourself so much more time, and it's not that bad. If anything, you're just understanding your customer way better. And to be honest, I mean kudos to you guys because that's one thing we really like about your team is that, even though we have wire frames and all the questions and so on, your team still challenges us. And we welcome that. So it's like, "Hey, are you sure you want to do it? Did you know that you could do this, this, and this instead?" "Oh, we didn't even know we could do that. Okay, fine. Let's do that."

(00:36:28): And then the other piece, and this took me a while in my career to get through it, because, you know when you build a dashboard, you have an emotional connection to it.

Justin Mannhardt (00:36:36): Oh, yeah.

Luciano Miranda (00:36:36): Right? And sometimes for us that are passionate about it, you are literally working late into the night. Sometimes it's just because you want to finish and you are so proud of it. And then a user who basically doesn't have anything anywhere close to it, and you know that, but they'll look at it and they're like, "Oh, yeah, this doesn't have this." And you're like, "Are you kidding me? What is this?" So what we did to protect against that, because in essence it's not the user's fault as humans, we're wired that way, right? I mean, if you're doing a review, by default, the review is likely going to tell you what are the things that are missing?

(00:37:11): So to protect us, I think I did it selfishly for myself and my own sanity, but I started calling those reviews, "Call my baby ugly parties." Tell me what you like or don't like about this. And it's something as simple as that just gives you that psychological safety. One for the user to tell you, because sometimes their feedback, it's spot on, and you missed it in your design. It just puts a touch of humor into it and lightens the mood. Also for the receiver, it's just to coach them, because otherwise, what you don't want to do is get defensive.

(00:37:45): There's a book I just finished listening to, so I listen to a lot of audiobooks. Unreasonable Hospitality, and this is about the rise of 11 Madison Avenue in New York City too, I think it was the number one rated restaurant in the world in 2017. And the person talks about how they not only focused about the food, but they started really focusing on the service. And how they would treat people. And there's this really amazing story. It resonated with me so much, and it talks about they were very proud of what they were doing. Similar to all of us, right? We're proud about what we do., But this particular guest ordered a steak medium, and the book it talks about, I guess most people don't know that medium is actually a little bit less done than what we mostly think medium is. So the guest goes to the waiter and says, "Hey, this steak is under done." Right? "This is not medium. I ordered a medium." The server was just trying to defend the pride, and to say, "No, we did not make a mistake. This is, as per the guidelines, medium."

(00:38:48): But one of the things that they did at the restaurant was they always did this retrospectives and talked through these things, and when they looked back at it, the guests felt terrible. Because you're basically, think about what you're saying is, you're telling the guests, "You don't know what you're talking about. This is medium. Fine, we'll cook it more for you, because you don't know what you're talking about."

(00:39:06): Think about it how many times we do that from an analytics standpoint, when somebody gives you feedback. If we react in that way, in essence, we are putting them down. And if we want to be not just an analytics organization, but also a service organization, we really have to think about it, and then just take a note of it. That was a really cool story that really resonated, and to be honest with you, guilty as charged, right? I've done that many times. Because I'm very proud of the work my team does. So I want to defend it with tooth and nails, but it is not the right approach if we want to be the organization we want to be. So I thought that was a cool story to share with you guys.

Justin Mannhardt (00:39:45): Whenever there's a conflict between the human being and the software, usually it's something's not right with software. People say, "Oh, the answer's training, we need to train them how to..." It was like, okay, a really good dashboard that's aligned with the right business questions and has good UI, good UX, it should be easy. There should be no friction between the human being and the business problem that they're using the analytics to help them with. As soon as the friction comes, party's over.

Luciano Miranda (00:40:10): Yeah, a hundred percent. Like for us now, as we continue to grow, there's more people coming in and there's different personas coming in. So there's a frustration happening that these new personas, the dashboards are really not designed for them because they might be, some of them are too detailed. So we now need to change some of the designs, not really change, but just create different views, for this type of persona that is coming into these analytics now. And we see the same thing. It's like, "What are you talking about?" The one person says, "The data's not there." And the other person says, "What are you talking about? The data is all there. We have data everywhere. We have this beautiful connected system." But the bottom line is we need to look to ourselves and say, "Hey, the data is there, great, but clearly there's a design issue with this persona. That this persona doesn't see that the data is there."

(00:40:58): And I think that's such a huge change in mindset, that kind of look at ourselves and say, "Okay, wait a minute. How can we do this differently? So they miss this filter pane. Should we put a popup when they come in with a big arrow that says filter pane here?" Or whatever it is. So that's what this year, for example, we instituted Gemba walks. So Gemba walks comes from manufacturing, you walk the floor to see firsthand, so how things are going. So for us now, what we're doing is that this team has to, within their goals to do Gemba walks, to watch people use the analytics without judgment. And they're like, "When was this updated?" And I'm like, "It's right there! The date is on the corner, it tells you!" But the bottom line is they're missing it. So if they're missing it, it's on us. And we need to figure out a way on how do we make that more visible, et cetera, et cetera.

(00:41:45): So that's a huge power, and there's been amazing stuff that we've seen where I had someone do a Gemba walk with one of our functions, and it's probably going to save them now, with relatively minimal effort, probably three, four weeks of work. We're going to save a whole group of people two days of full work. Again, this persona did not know that you could do these things, but our team that is more technically inclined and obviously is what we do, said, "Wait a minute, why are you doing that manually? We could totally automate this for you."

(00:42:14): The Gemba walks are incredibly important, but I wouldn't say very difficult, but they are challenging to do because you cannot solve it as you're doing it. If you're watching them, let them show you what they do versus you saying, "Oh, you want that filter? Just open the filter thing it's right there." You want to see that, and that helps you then improve your designs. And to be honest with you, your designs are never finished, in the sense of there's always ways that they slow down in the amount of change, but there's always things that you could do to improve them. In my opinion.

Rob Collie (00:42:44): I agree. That is the way to do it. When you first start, this is a journey, right? Everybody who gets into this business of dashboard building, data building, analytics. Eventually goes on this journey, of realizing that they're building software and software is built for users, and the users win. You don't want the software to try to make the users change. The software needs to bend to the human. I went on that journey myself at Microsoft, learning to design Microsoft software. In my early days, I was the nerd who was like, "Well, we're going to design this system and the humans are going to love it." No, you're, they won't. They're not going to love it it turns out. They don't like your nouns, Rob. And so I'm going to give you a three word phrase, see if you have any feelings or reactions to it. The three word phrase is, "Export to Excel."

Luciano Miranda (00:43:39): Do I have a feeling to that? I do.

Rob Collie (00:43:41): From the smile on your face? I believe you do. Yes.

Luciano Miranda (00:43:44): When we started the journey with Power BI, we had seen other softwares try to take that away. We observed the quickest way to upset everybody and to fail is to take that away. Because I mean, I think there's like 30 million, or who knows now, maybe my numbers may be a little bit dated, but 30 million plus users of Excel. It's a phenomenal tool. Let's not kid ourselves. So how do I feel about it, now? We obviously always had a design protocol that takes you to details so that you eventually could download to Excel.

(00:44:23): Now, what we encourage our users is don't download the whole thing. If you're downloading everything and then doing lookups and then bringing information and filtering and so on. That's a failure on our design and we need to do it better. So what we try to encourage them is to say, "Hey, let's build a flow where you get to the exception, then download those." Because Excel is a fantastic tool to really quickly look at the exceptions so on. It's just that the exceptions may be, I don't know, a few items to let's say a thousand items, whatever. But it's way better than a million items. You don't need to look at that. You need to look at just the exceptions, a small part.

(00:44:59): So I am definitely on the side of allowing the export to Excel. I think it's a phenomenal feature, but again, it can get abused. If someone's trying to download too much, then I think that it defeats the purpose of the animation and efficiencies that you can get through a dashboard. But I wouldn't take it away. Let me put it that way.

Rob Collie (00:45:18): I love that. I think that is the right answer. It's not just because that's my answer as well. But both parts. Don't take it away because the literal last mile, or the long tail of analytics tasks that are going to be one-offs, that aren't truly repeated. You're not going to want to build dashboards for each individual long tail use case of like, "I need to drop these 20 records and then go do something else with them." But somewhere along that half bell curve of the long tail is a line where you say, "Okay, but anything above this, we need to take into account in the design of the reports themselves. We need to address that use case, is a repeated recurring use case. We don't want the 150,000 row download to excel as the answer to that use case, for so many reasons."

(00:46:14): It's not just like, "Oh, because the software doesn't like to drop that many rows", and all that kind of stuff. It's also because this is exactly what your team exists to do, is to make that process more reliable, lower the labor cost of it. And so when you recognize a recurring heavy export need, I would expect that'd be the place where you go and take a look and say, "Okay, this is exactly one of those places where our design needs to evolve."

Luciano Miranda (00:46:39): And that's where the Gemba walks are so good, because you see that, we just saw one with one of the people we serve, where this individual was downloading and doing all this other stuff and we're like, "Whoa, wait a minute. You don't need to do that. Where are we failing?" There's another book. I'm not going to get philosophical, don't worry, we're not going to get deep, but it's called Change Your Question, change Your Life. In a nutshell, it's about the way you ask questions, even to yourself, will change how you feel and how you make others feel. The author, right talks that there's two paths, the learner path and the judger path. And we're all judgers, I think. And he says, "What is wrong with them? What is wrong with Rob that he doesn't see where the filter pane is Rob? Or what is wrong with them that they're downloading all of this information?"

(00:47:26): Now you can ask the same thing in the learner path. Which is, "What is it that they need? What is causing them to do this? What am I not delivering to them that is causing them to do that?" Eventually it's like I'm going to jump to the judge's pit with two feet forward, but make sure that you go through that learner path first. And to be honest with you, it's not just positive to the other person, but it's also positive to you, because of the feelings that, again, in analytics, believe it or not, it's like you have to have a pretty thick skin. In my opinion any of those things that you can do just keep you in a better mindset, which then keeps you in a more creative mindset, which then leads to a better performance, which leads to a better analytics, which leads to a better team. So I know I'm making a lot of connections, but I do feel that they're definitely connected. It's not the only thing, but they're definitely connected.

Rob Collie (00:48:17): I don't mind you making lots of connections.

Luciano Miranda (00:48:20): I love connections, love connections.

Justin Mannhardt (00:48:22): That's a title candidate, by the way, "On the Learner Path with Luch."

Luciano Miranda (00:48:27): And the other thing I would say for folks that are building organizations and so on is really look at your people, and look at what are their hidden strengths. I was at a conference, like 2000 people I think in this auditorium. The speaker said, "Stand up if you think that the company's not using a talent that you have that is being underutilized right now in your jobs." And the entire auditorium stood up. As leaders, we have the same thing in our teams. And we have to ask ourselves, "What are the talents that people have?", and adjust to it.

(00:48:57): For example, the organization that we started with is not exactly the same as what we have now, because it's how do you look at everybody's strength and then appreciate it for it? And also sometimes we think we're experts in everything, and we're not. So how do you lean on your teammates when they know that area more than you do, and just be humble enough to acknowledge it? There's a lot that goes into a team, but I would say that would be another thing to look at, make sure that you're close to the customer. And two is be very aware of the talent and the skills that they may have, that completely may surprise you, and could be of huge benefit. We all want to give value in the world, and most people anyways want to do something good.

Rob Collie (00:49:39): That's the kind of validation that human beings want the most, I think. Is to help and know that they did. Really, really, really powerful force. We tend to under count that I think in the professional sphere. We define success by a number of criteria, but we tend to put a bunch of others ahead of that one, and I think that one's just as important as all of the others.

(00:49:59): Changing gears for a moment, what do you think of Microsoft's direction in the AI and fabric and all of that? There are ambitious new moves. Following that story?

Luciano Miranda (00:50:10): Yes, very closely. Earlier on I talked about how every time we think we get pretty good at something, everything changes.

Rob Collie (00:50:17): Yeah.

Justin Mannhardt (00:50:17): Here we are.

Luciano Miranda (00:50:18): Here we are. It's changing again, right? What I feel good about with our team is, I talk a lot about the how and how you make people feel and so on. I think if you're resourceful, we will figure it out. I remember when I started using chat GPT about December of last year, I was like, "What is this?" And so on. And I remember looking at it, I had a moment of panic. Because I'm like, "Oh my God, what is this and what is this going to do to what we do?" And I took a breath and I'm like, "Okay, there's going to be an opportunity." Our IT team did a phenomenal job of allowing us to use it, with guidelines. Obviously you couldn't put anything confidential or anything like that, but I said, "Start using it." Because in essence, start training ourselves on how to become prompt engineers, and how it works.

(00:51:00): So we also have seen some of the work that Microsoft was doing. So we've been keeping tabs with them, so what are these folks investing in, and how will we have to evolve in order to best leverage the technology that Microsoft is coming up with? Because we believe it's going to have some impact in how we work, and it's going to change potentially some of the rules. If we do it properly I do feel it has the capability to make us even better. It still probably has a ways to go before it has the impact that we think, but you can see where it's going, and I'm willing to bet that they will continue to invest and it will get a lot better. Some of the stuff maybe is not as great as we want it yet. However, I'm not too worried about that, because I've seen it before and they're going to get it to the place of where it needs to be of the full demo videos that sometimes you see, but we're looking at it through, how is that going to potentially adjust how we work, as well as how we serve the people that we do?

(00:52:01): One example I can give you Rob, is we don't know yet, but we know that putting generative AI on data is going to be a game changer. So what the team has already done is we created what we call a KPI hub or a metrics hub. We're hoping we haven't put this on that we can then unleash this into this model that is pre-connected, pre-joined. We have about a hundred KPIs already in this model. People can do more natural language kind of conversational with the model, right? Because what we're worried about is you can do the conversation, I mean that technology's been around for a long time, but is the answer going to be the same? That's the thing that we are worried about. Is how are we going to get the same answer? Am I going to go back to the time when there were 13 different back order reports, because it's pinging or making different connections and so on?

(00:52:49): So we're trying to put some guardrails around it from everything we've learned from Microsoft so far and what we think it's going to need. So we're very excited to use that technology further and just continue to test it and fail forward and figure it out how to use it. But we hope that the investment we made has given us a leg up, to leverage it faster.

Rob Collie (00:53:11): A few thoughts that ran through my head listening to that. So first of all, a new term that we're trying to make stick. Can you help us spread it around, which is not FOMO, FOBO. Fear Of Becoming Obsolete. It is a thing, but it's a cousin of FOMO. So when I look at fabric, I don't really get much personal sense of FOBO there. Very strangely for me because I tend to be a skeptic about most things software. I look at all that and say, "Oh, that's just places where my existing skillset becomes more valuable. There's just more ways to consume the value that people like me produce." I still see that most people, I think, are reacting to fabric in general with a sense of FOBO because they don't understand it yet. And they feel like it's replacing things that they know as opposed to extending the powers of what they know.

(00:54:00): On the flip side though, the co-pilot stuff, the generative AI stuff, that stuff actually is, it's a little scarier. It's not really that scary when I calm myself down and think rationally about it. But that one does sort of get at the primal brainstem a little bit. There's an influencer culture on LinkedIn in particular that kind of thrive on just click bait news has become such a dominant force. Headlines and things that are shared on LinkedIn are oftentimes deliberately constructed to inflict FOBO, by overstating things.

(00:54:33): So one of the things that I was seeing for a while, I don't know if it's already gone out of style, which was saying that dashboards are yesterday's thing. Dashboards are going away. That's a pretty weird thing to say. If there's something that we know that I need to know about all the time, and be able to interact with a few simple clicks, I don't think that's going away, any more so than a spreadsheet is going to go away. Excel is not going to be replaced with a generative AI chat bot. How do you see the relationship between the natural language, I can formulate a question, or ask it to build me a report. How do you see the relationship of dashboards versus perfectly tuned generative AI interface to data models?

Luciano Miranda (00:55:17): So, with the information that I have now, this is what I see. I do believe that there's still going to be a requirement to have a curated analytics that are trusted. Because imagine this, right? Imagine if everybody can build their own dashboard, and everybody's asking questions and so on. You go into an executive meeting, that could become a little bit chaotic. In how things look and feel. You have to constantly figure out, and I'm sure there'll be ways to control that. Where I do see it being a big opportunity is potentially in building some of them and giving us a head start. I can assume most of you have used chat GPT. It doesn't get you a hundred percent of the way there, but it gets you a good 85, 90%, a lot faster. And one thing we always see with analytics is that every time you build an analytic or a dashboard, you start using it and you're like, "Oh, I should have done this." Because you learn something new and you keep exploring and exploring and exploring.

(00:56:11): I think that that journey is going to be a lot faster with this technology. Because we'll be able to potentially build analytics and build front ends a lot quicker. We started with the front end and the back end. The back ends are going to have to be pretty solid in order for us to leverage it. So what I feel is, I think that there will continue to be dashboards that are required. Here's an analogy that comes to mind and tell me if I'm totally wrong. Spotify versus the radio station. You have your liked songs, and in the beginning it's awesome right, and you're getting it, and it's giving you what you want and it's just doing it for you. But then it gets monotonous. And you're like, I want someone to curate this for me. And then I might go back to the radio station a little bit. And then maybe I get tired of the radio station and I'll go back to Spotify, et cetera. And I'm sure Spotify's putting a lot of AI, they have an AI DJ now, et cetera.

(00:57:00): But again, I still believe that there is that need for standard analytics and standard dashboards. However, those dashboards are going to become way more powerful because potentially the interaction with those dashboards will become supercharged with generative AI where you can have, instead of having somebody from my team sitting down and saying, "Oh, what do you want to know?" And remember we talked about where's the filter and so on. I think that now that interactivity between the user and the dashboard you created will be supercharged. I'm assuming it'll be similar to a chat GPT experience. You'll be able to almost have an artificial assistant at your side, answering the questions that you want to answer of that analytic. So I see it more as a supercharging our dashboards, and potentially very much expediting wireframe to concept provided the data is available, that will likely be faster. Those are the two areas. I don't see dashboards going away with what I know today.

Justin Mannhardt (00:58:02): Think of the friction that happens in the process of developing something, and you're like, "Oh, I've got to do this thing and I don't know how to do it." And so you're Googling and you're looking for an example, you call Luch. It's like, "Hey, Luch, have you ever done this thing?" And to say like, "Oh, I just have an assistant that's like, hey, how do I do this?" I think sometimes we fantasize that, oh, I'm going to be able to say, "Give me a dashboard that...", and we're going to get all these fireworks and bells and whistles. I think about the friction coming out. Not only for the builders, but for the end users like you were seeing Luch.

(00:58:30): So you curate a dashboard, it's got to look and feel, they're looking at it and all of a sudden the end user goes, "Oh, Luch has given me this great product. I have 5, 10, 15 new questions that I want to know the answers to." Today they come back to your team or to the squad and say like, "Hey, what if I adjusted this for weather patterns or..." Whatever they're having these ideas, and they can have a creative element at the end. It's kind of like edge computing to me in a way.

Luciano Miranda (00:58:56): Our vision when we started with Excel, and I have an article to prove this that I wrote seven years ago, and I had no clue how we would do it, to be honest with you. I knew the steps that we needed to do, but I didn't know the technology, et cetera, because at the time when I wrote the original version of the article that we only had Excel. At least the vision was always create an artificial assistant, where you can have that augmented experience. Someone that can look out for you in a supply chain and just be your person that's actually looking at all the analytics. Because I mean, one of the challenges we have today, that we're solving, we've also now created a new problem where now we have too many analytics. One of the things particularly that we get from newer users is that they find it difficult to find the analytics that they're looking for.

(00:59:44): Totally fair. When we first designed, we had launchpads and everything is connected, et cetera, but when they were first designed, they were not designed for the amount of analytics that we have. So we're revamping all of that. And putting again, an official assistant to help you. For example, we're looking at a couple of technologies like power virtual agent, maybe watsonx, that you can actually ask questions of what you are looking for and start answering. I did a test. Every analytic has an information page, and these are PDFs that basically tell you what the analytic is, the one that I was testing on our IT team, again, huge kudos to them, they created something called Medtronic GPT. So now we can start putting not super confidential data, we can put some of internal data in it, so we can load our PDFs and start asking questions.

(01:00:27): So I loaded the PDFs and I said, "Hey, can you explain this to me as if I was a fifth grader?" "Think of it as you're baking a batch of cookies and you basically burned the entire batch and you have to throw it out. Now you have to build a new batch of cookies. The cost of poor execution is the fact that you just threw out a whole batch and you have to build another one." So the fact that this technology can do that, it's interpreting the information, the words, and then putting it in a language where you can understand it in the most basic of terms. That's a huge supercharger that no basic training is going to get to. So that's how we see it as an opportunity. And again, for us, the big opportunity is how do we make things simpler for users to find our data, to find the analytics they're looking for, and just remove as much, I like your word, the friction. Remove as much of the friction as humanly possible, and this just continues to get better.

Rob Collie (01:01:20): Yeah, I read something on Twitter a while ago, I don't remember who said it. They basically said that software development made it such that any process whose steps could be written down, could be replaced with automation. Really, really broad oversimplification. That's the revolution that we've all been living through our lifetimes. Digital transformation. Just the fact that everything has gone digital and systems have gone online and we've used software to run our businesses as opposed to pen and paper and all that kind of stuff.

(01:01:50): But the thing with generative AI, the twist on that is that now if you can describe the end state that you want, the software can do that. Without knowing the steps. Here's a really simple example that I will not miss once it gets to the point where I can do this. I will not miss laying out my visuals on the canvas and formatting them all the same so they all look the same, and making sure that they're all lined up with each other. I am not going to miss that tedious process. That doesn't enhance anything about the result. In my head I know kind of what I'm going for. Imagine just saying, "Hey."

Justin Mannhardt (01:02:29): Make pretty.

Rob Collie (01:02:31): And make it conform to our design scheme, the Insights brand guide. There's an example of giving generative AI a gigantic hug. Thank you, thank you, thank you, thank you. People are afraid it's going to replace them. I don't mind that part being replaced. That's okay, I'll be first in line for that.

Luciano Miranda (01:02:48): A hundred percent, and I think it just makes us more towards truly the value added activity that you're putting in, that you're looking at and so on. I choose to look at it as a huge opportunity. Because I can go dark too if I look at it from the other place, and I can tell you, being an immigrant, having been born in Argentina, which by the way won the World Cup last year. Still, for the next four years I get to celebrate. So coming to a new country, having to start from the very, very bottom and then obviously getting all the opportunities that the North America gives you.

(01:03:24): You do end up having FOBO, I love your term, Fear Of Being Obsolete. Now we can use that as a positive. Because FOBO keeps you from getting complacent. So now FOBO if taken too far can also create stress and drive you nuts, and I've been there too. But learning how to use it, if anything, the reason why I immediately got a paid license to chat GPT and everything because like, it was FOBO. I'm like if I don't learn this, this thing will obsolete me, potentially down the road. So I'm not even going to give it a chance, I'm going to try to figure out how to use it and how it can enhance us and so on and so on. But a little bit of fear is okay, a little FOBO. Too much FOBO and it's like I said before, no bueno.

Rob Collie (01:04:06): Some people are capable of running at high speed on autopilot. I'm not one of them. Emotion pretty much is my motivator. If I'm not afraid of something a little bit, or excited about something a little bit, I'm probably not going to be moving too much. And so being excited about being pulled forward into something new, that gets me up in the morning. Being a little bit afraid just a little bit right, keeps me up at night. But you're right, complacency is absolutely a killer.

(01:04:36): And so I was even saying this, I think in a recent LinkedIn post, you've got to get that perfectly titrated amount of FOBO. Not zero, but also not crushing, crippling like, "Oh my God, there's no hope." I think it's very easy for it to run away into that upper one. And this is why you also see some people just completely going the other way saying like, "Ah, it's all a fraud. It's stupid. It's just the latest fad." No, it's not really a fad. This isn't going away. It's going to make an impact. Don't kid yourself.

Luciano Miranda (01:05:06): There is a tsunami coming, and you have two options. You can either stay at the beach and watch, and look at the beautiful weather because nothing's happening yet. Or you can start swimming, and catch it at the middle of the ocean, which at least gives you a better chance to maybe ride it. I totally agree with you, Rob. Little FOMO, is okay, sorry, FOBO. When you go too much to the FOBO side, then I have the Oura Ring and it gives me horrible scores because I woke up several times earlier.

Justin Mannhardt (01:05:36): Whenever I feel the FOBO, I just look backwards. Even your story about your team, it's like, you did that over a long, five years is not a long, long time, but it's a long period of time. And you adapt and change, you learn how to, I always say a career in analytics this is a long series of wishing you had learned about something sooner. You just keep going and it will change. I think if we go in the time machine five years ahead from now, the way we're going to build these solutions is going to look a lot different. Look at what's happened just in the last decade, guys. So it's going to change, and there's still people out there that aren't even doing interactive visualizations. It's not like we're all going to get obsoleted tomorrow, but you do need to get in the slipstream.

Rob Collie (01:06:17): Back when I used to teach classes, to get people sort of motivated at the beginning, I would show these two bar charts to the class. And this is one of those things where the audio podcast is not a good format for this, because it was a chart for a reason. But basically what I showed them was like, here's what your day looks like today. The hours in the day, and the amount that's asked of you. The bar of what's asked of you is just so much greater than what you have time for. You're just absolutely drowning. And this is talking to them from the perspective of, again, this is going back a ways now, is now using Excel as your primary analytics tool. And when I was teaching classes, the vast majority, I think still probably today, the people who are taking a foundations Power BI class are coming from Excel still. This is what you should anticipate.

(01:07:03): So here's your day-to-day. You're being asked to do 10 times as much as you possibly can. And in the future you're going to be asked to do one and a half times as much as you can do. You're not going to be ahead or even really caught up. There's still going to be more asked of you, that's just the way the world works, but the ratio is going to become a lot more sane. Then the second thing I would tell them is that the really amazing thing though, isn't that that ratio gets better. It's that of the things you're doing, the proportion of the things you're doing today in Excel, is mostly manual grunt work. And a very, very, very tiny amount of thinking, and every little bit of thinking you do generates a tremendous amount of grunt work. So it's almost like you're dis-incentivized to think. Because every new valuable thought produces all kinds of labor that you have to perform. When you switch to Power BI, that ratio inverts. You're spending more of your time doing valuable thinking and less of your time doing the grunt work.

(01:08:04): It just occurred to me during this conversation that this AI revolution, all this stuff, is really just taking that to the next level. The ratio of the amount of valuable thinking you're doing is going to go up again, and eliminating all the grunt work even more so. Who's going to complain about that? Really no one. I think this might honestly be a turning point moment for me where I kind of step out of the uncertain zone, into the, "No, I'm really looking forward to this." As opposed to being 50/50 unsettled about all of it. I wouldn't say that, if I wasn't actually feeling it at the moment. This has been a very, very valuable conversation I think. Connecting it with that old metaphor of mine, and the Excel people could have been terrified, right? "The things I do all day, you're coming for those?" "Yeah, that's right. You're not going to be writing VLOOKUPs anymore and waiting years for them to run. You're right. I'm going to take that away from you."

Luciano Miranda (01:09:01): What I love about this podcast is it's not only about analytics, but also about psychological wellbeing. So I love it.

Rob Collie (01:09:07): Yeah. I mean, we subtitled it data with the human element, and you walked right into that didn't you?

Luciano Miranda (01:09:14): Love it.

Rob Collie (01:09:14): With all of that reading that you do.

Luciano Miranda (01:09:17): Yeah.

Rob Collie (01:09:18): Audio book listening. Yeah. We should have you on a lot just to talk about all the books you're consuming and how they relate back to analytics. Like medium cooked steak.

Luciano Miranda (01:09:27): Quarterly feature? Book corner with Luch.

Rob Collie (01:09:31): It's the book corner, yeah.

Luciano Miranda (01:09:33): It's funny because I remember listening that most CEOs, or you look at people that are really successful, and I'm like, they read a lot. Now, my problem is I don't have as much time to read and English is my second language, so I'm not the fastest reader, so you just figure it out, right? I'm like, "Well, I got to walk the dog, right? Put on an audiobook!" You got to drive, right, et cetera. So to be honest with you, there's so much good content in these books, and again, you pick just a little bit of each one. And we talked about connections and then how they connect with each other and so on.

(01:10:04): So that way, I think for all of us, you always want to continue to be growing, at least for me. Once you're not growing, and then it helps you psychologically. We talked about all of this stuff that's happening with gen AI and these are threats, potential threats that are real. But I think it just puts you into a position where you can say, "Okay, hold on a second. Wait. This is an opportunity." And just that shift of perspective takes you from inaction and almost like a helplessness, to now go look for it. It's amazing what the human mind can do when you actually start looking for things.

Rob Collie (01:10:37): I agree. To use a soccer football metaphor, you can shoot around the goalie or you can shoot for the open net. One of them is a defensive mindset, and one of them is a creative one. Let's go do something good, as opposed to avoid the negative outcome of the ball hitting the goalie. And I'll tell you, it's really easy to avoid hitting the goalie if you just kick it completely wide of the goal. If you're optimizing for the wrong metric, it sounds mathematically the same, shoot around the goalie or shoot for the open net. One of them produces results and the other one doesn't in practice.

Luciano Miranda (01:11:15): Yeah, a hundred percent.

Rob Collie (01:11:16): So at the beginning when we first started talking, I was like, Medtronic is by any metric, is absolutely one of our largest clients, if not our largest client. A lot of customers we work with in the mid-market, and even some of them are even small businesses. In the beginning we were talking about all the organizational challenges you'd necessarily have to navigate such a large enterprise. You're doing the centralization game and all of that. And I was thinking to myself, this isn't going to translate to a lot of the people who are listening, because they don't have these sorts of big problems. Two minutes later, we were talking about all of these psychological touch points and principles about everything about your job. I was like, "Oh, I feel a hundred percent. This is hyper relevant, to everybody." I did not anticipate where this conversation was going to go, but oh my gosh, Luch. What a thoughtful approach.

(01:12:10): I love the blending of the human factors with the tech. That's what it's all about. And as someone who learned the hard way, like I was a software nerd. I was produced by computer science pedigree, and I went to Microsoft with that pedigree. Really painful way of unlearning. Inflicting my math, inflicting my logic on millions of people. Respecting my design, of whatever. And coming around to work backwards from what the audience needs.

(01:12:41): The hardest won lessons are the ones that you internalize the most. Having walked that path to hear so much wisdom in parallel from your journey, I can absolutely nod with full vigor like, "Yes, this is the way." I'm really grateful that we got to have this conversation.

Luciano Miranda (01:12:58): Thank you, Rob. I'm grateful to you guys. I mean, you know how highly I think of you and your company, what you've been able to build. I've always been a little bit jealous too, because I always wanted to be an entrepreneur. I realized in one of the books I read that I am actually an intrepreneur because my last eight jobs did not exist. One comment on the fact that you say for smaller companies, because often I talk with friends that have small companies, they're like, "Oh, you have all of these resources and all these things." And it's like, "Well, when we started, we didn't. It was almost like a smaller shop." And what I was telling, it's like a Power BI license, I don't know what it is now, right? But when we started, it was like $10 and 99 cents a month. And you could do magic with it.

(01:13:37): Obviously you can't do the 3000 data quality checks and maybe some of the other stuff that's not applicable, but that doesn't prevent you. You know what I mean? So to me, this is available for everyone. I love what you guys do and obviously I'm a huge fan of, and I want to give you a compliment, right? Because you can see the leadership and how you take the human factor because of the people that you guys hire, and how they work. So thank you for doing that because you definitely make us better, and it's also a really pleasant experience to work with you guys.

Rob Collie (01:14:06): I really appreciate that. Thank you so much. Working with y'all is also great, thoughtful people to work with? Yeah, let's do that wherever we can.

Speaker 3 (01:14:13): Thanks for listening to The Raw Data by P3 Adaptive Podcast. Let the experts at P3 Adaptive help your business. Just go to P3adaptive.com. Have a data day.

Check out other popular episodes

Get in touch with a P3 team member

  • Hidden
  • Hidden
  • This field is for validation purposes and should be left unchanged.

Subscribe on your favorite platform.