Creating a Theater of Excitement in Analytics, w/ Austin Senseman

Rob Collie (00:00:00): Hello, friends, and welcome to week two of 2024. Today is one of those rare episodes where we welcome back a prior guest. Yeah, that's right. Making his second appearance on the show is the one and only Austin Senseman. In his first appearance on the show, Austin was an ex-P3-er who had left to found and become CEO of an internet of things startup called Conserv. That was a great episode, and I encourage you to look it up in our back catalog. As you might imagine, we talked about data quite a bit in that episode, and staying involved in the world of data over those many years that Austin was running Conserv led to him eventually experiencing a reawakening of that hunger to be directly involved in the space that he was when he was with P3 years ago. Having gotten Conserv into a strong position, Austin started feeling like his work was done there in a way, and yeah, that's right. Austin is now back with us for a while.

(00:00:59): Now, in last week's episode with Kristal Searle, we talked a lot about the jumpstart process, which through one lens, you can think of it as just the way that P3 starts a consulting project with our clients, but it's actually a lot more than that. That up-tempo, results-oriented, impact-driven, compressed process really epitomizes the entirety of how we operate throughout all of our client engagements whether they're brand new or long-sustained. It also epitomizes the ways in which you can work with data today that you couldn't work with it before, and it's really important to zoom in on that to distill that spirit because it is absolutely still possible with today's tools to work the old way and to not capture all of the value that's possible today.

(00:01:55): So the three of us getting together, myself, Austin, and Justin, we're all three veterans both of the jumpstart process and also, that overall spirit, that overall methodology of how to approach data projects today, and so we got very reflective about it. While in that reflective mindset, Justin busted out what might become one of my favorite three-word phrases of the year, "The theater of excitement." Naturally, you'll hear me say that phrase in a monster truck voice during the episode, but it's also very serious. Getting to that point where the intersection of data, tools, and stakeholders feels exciting like an electricity in the air type of excitement, that's actually a standard that you can shoot for. That's a standard that you should shoot for, and it's attainable.

(00:02:47): Now, yeah, there are always boring times in every project. Sooner or later, you find yourself just formatting charts, right? But at key junctures in the process, especially near the beginning and also, at key moments when a deliverable appears on the screen or is handed off for the first time, if you're not feeling an electric excitement, there's almost certainly significant value being left on the table. In that vein, we talk about how there really isn't a data strategy that's separate from business strategy. The excitement isn't about data or tools. The excitement is about the improvement that is happening before your eyes with your business.

(00:03:27): We also talk about how much opportunity is still out there. There's a really powerful metaphor that comes up that involves just walking into a random office and the opportunity that you are almost certain to discover there. We talk about how in a recent jumpstart that Austin performed, how a problem that the client had had two full-time employees struggling with for six months and how that problem was solved by lunchtime of the first day. So, shall we step into the theater of excitement?

Announcer (00:04:00): Ladies and gentlemen, may I have your attention, please?

Announcer (00:04:04): This is the Raw Data by P3 Adaptive Podcast with your host, Rob Collie, 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 (00:04:31): Welcome back to the show, Austin Senseman. How are you today, sir?

Austin Senseman (00:04:35): I am great, and I'm glad to be back.

Rob Collie (00:04:37): Well, you're back in multiple ways, aren't you? So, first of all, you're one of a small elite group of multi-appearance guests. Now, you're somebody. How's it feel?

Austin Senseman (00:04:48): It feels great. I'm pretty sure I'm the first insider-outsider-insider.

Rob Collie (00:04:51): You have been, in a way, there and back again Bilbo Baggins style, although that's the story that Bilbo wrote at end of life. This is just your point in time where you're at right now, right? You're an adventurer at heart.

Austin Senseman (00:05:03): I am. I'm a variety person.

Rob Collie (00:05:06): Yeah. You, Austin, were instrumental in the early days of P3, one of the merry band that was around when we went from 1099 to W2 many years ago now before the earth was cool kind of thing. You got bored, you went, and you ran a startup. Tell us a little bit about that.

Austin Senseman (00:05:24): The very short version is we left Power Pivot Pro at the time. This was before the rebrand.

Rob Collie (00:05:29): Power Pivot Pro, the P3, the source of the P3 and P3 Adaptive.

Austin Senseman (00:05:32): The source of all good things, P3. Then, in the beginning of 2019, we started Conserv with my business partner, Nathan McMinn, and we build sensors that measure things like temperature, humidity, light, and vibration, but we sell them to a very specific group of people, people that are trying to preserve physical objects and libraries, museums, archives, and they have to create these perfect environments. So that's textiles and paintings, other things that last for thousands of years, and so that's Conserv. So we build sensors and analytics software that sits on top of that that help them understand that data. The number one question we get is, "How did you get into that?" because it does seem like a strange detour in life. Nathan's mother was a conservator. We got introduced through her. That business is doing really well, and I went and climbed Mount Kilimanjaro last summer, 2023. I went with my dad and my brother.

Rob Collie (00:06:27): Not figuratively. You were talking about startups, and you say, "I just climbed Mount Kilimanjaro."

Austin Senseman (00:06:33): Yeah. That was a thing. Right, right. This is not a metaphor.

Rob Collie (00:06:34): Yeah. You were actually out in the world climbing the physical mountain.

Austin Senseman (00:06:39): We did that. Yeah, that's right, and there was something about that experience where I got back home and after five years of doing Conserv and just thinking, "I'm ready to do the next thing." Fortunately, we had a really good leadership team at Conserv, and so we promoted someone internally to be CEO, and I'm still on the board there. I'm still very involved in the company, but we have a different leadership team there for this next phase of that business. So that's where Conserv stands today, and then I had this question of like, "Well, what do I want to do now?"

Rob Collie (00:07:11): Yeah. First, I have to say that I owe Al Gore an apology because years ago, before he ran for president, one of Al Gore's publicists after Al Gore climbed Mount Rainier, which I don't think is that as big of a deal as Kilimanjaro. I've never climbed either of them, even though I lived in Seattle, and apparently, Al Gore was so transformed by having climbed Mountain Rainier that his publicist said he went up that mountain a senator or whatever, a man, and he came down a president. I've been making fun of that for so long, right?

Austin Senseman (00:07:51): I went up Kili as a CEO and came down just a regular man. It was the flip side. Yeah.

Rob Collie (00:07:57): Yeah. I don't know if it has the same ring, but the point is it's got a transformative power that I underestimated.

Austin Senseman (00:08:02): These are inhospitable environments where humans are not meant to be, and it's a weird vacation because you deprive yourself of all material comfort. You're basically in a Martian type landscape. You put yourself in situations where people lose their lives occasionally and all that, so it's not exactly fun. People talk about having a mountaintop experience. That's a real thing. There is something about going up, and doing that, and coming back down not feeling quite the same.

Rob Collie (00:08:26): I've tried ice axe and spike on boots repelling one time.

Austin Senseman (00:08:33): Yeah.

Rob Collie (00:08:34): It did not go well.

Austin Senseman (00:08:37): Yeah.

Rob Collie (00:08:37): "Get me out of here. Hoist me," is the verdict. You were part of our startup for a while, right?

Austin Senseman (00:08:45): Mm-hmm.

Rob Collie (00:08:46): Then, you went and created a brand new one, basically, from whole cloth, grew it for five years, and climbed the mountain, came down. I need something new and different. So here's the loaded question. New and different. You came back to the places that you were before. Isn't that boring? From the outside looking in, someone might go, "If you're looking for something different, why circle back to a place you've already been?" So what was different about coming back here five years later that was exciting and interesting?

Austin Senseman (00:09:14): Well, I was certainly different, and last time I was here in 2018, I think there were maybe a dozen people here. Now, there are 60 plus folks here or whatever, so it is a different organization. I was excited to be back here because I still felt the vision for the business was still very fresh and compelling, and Jeff Bezos has that line where he talks about it's always day one, that kind of feeling at your company where you're always at the beginning. Whatever you think of him aside, it's like I still have that feeling about this business and this space that we work in. There's still so much important work to be done, and it felt right to come back and get involved in that again. Then, also, the next few years are going to be fun.

Rob Collie (00:09:59): Yes.

Austin Senseman (00:10:00): Not just because of what's going on here, but the world is going to be changing a lot with things like Microsoft Fabric which we've been talking about. Generative AI is still unclear. We know it's going to change a lot of things, but how much? So all of that is just icing on the cake for it's going to be fun to be in here with y'all shoulder to shoulder working on these things. Yeah, it just seems like an exciting time.

Rob Collie (00:10:22): At Conserv, you are also still active, at least partially in some of these toolsets. I know we talked a long time ago that some things... You didn't use Power BI for your core customer-facing visualizations in the Conserv product. When you're going to be that dialed in, you want total 100% control of the visualization experience, and so it doesn't even make sense necessarily in that situation to use Power BI embedded, but I get the sense that you were staying current with what was going on. So were you using some of these tools in a back office capacity at Conserv? How are you still staying plugged in?

Austin Senseman (00:11:03): Yeah. Well, the first version of the product we shipped did have Power BI embedded in it. So there was a good six-month period in there where we were using that because it really facilitated customer discovery, and we got a little more dialed in on what we wanted. Then, we went and built it.

Rob Collie (00:11:14): That's perfect, actually, like the faucets-first direct philosophy that we use at P3. It's exactly how we would approach it. Don't go code from the get-go, your visualization and dashboarding experience in custom implementation, only to find out that you're way off the mark in terms of what customers want.

Austin Senseman (00:11:32): Yeah.

Rob Collie (00:11:32): Stay in that rapid iterative environment until you know what you want, and then going and coding that one thing, that one path that you've finalized on, that's not very expensive. Coding, and recoding, and recoding, and recoding is very expensive.

Austin Senseman (00:11:45): It is, and every organization is trying to increase the tempo and, "How quickly can you learn? How quickly can you iterate?" Yeah, Power BI was great, and so I think certainly, by the end of 2019, we were fully in the software mode. We weren't using Power BI. Also, at that stage of a business too that early on, things are changing a lot where we were really managing our business in Excel for quite a long time there too, and we would get Power BI out when we had very specific special ops things that needed to be done. We need to build a customer churn model or something you could do in Excel, but it was like, "No. We want to be able to click on all the filters, and mess with the parameters, and explore this thing." So we have three or four models there that they were very specific, tightly-scoped projects that helped us solve fairly complex things. That's how we used it.

Rob Collie (00:12:37): Okay.

Austin Senseman (00:12:37): Here's how I think about it. When people talk to me about, "What did you learn doing that, doing Conserv, doing a startup?" the one thing I key in on is that simpler is better, less is more, complexity is the enemy of execution, that whole mindset. I mean, we would do things. I would have learned this lesson, right? I would go and build something, and I would tell myself, "Ooh, I need to make this 10 times simpler," and I would do that. Then, six months later, I'd be like, "Mm, I should have made it 10 times more simpler than that, right?" It was just endlessly like that. Every business, you have more stuff than you could do, than you can do, and so a startup is a very extreme version of that, or starting any business from scratch is an extreme version of that.

(00:13:17): Just having a spreadsheet, there are five numbers in that kind of environment with all of that chaos turned out to be really, really powerful. It was just incredibly simple. So that's at the core of how we manage that business still. Could it be a Power BI dashboard? It could, but that's a team of 10 people today, 10 or 12 people. A lot of that slicing and dicing around, "Well, how is this region doing?" and getting deeper into it and trying to understand we have one product that we sell, and the business model is relatively simple in a lot of regards, and so that works for us.

Justin Mannhardt (00:13:46): I'm curious, Austin, you say that all the time when you and I are working together, "Simple is better." It seems like that's something you learned and discovered. How did that come about for you?

Austin Senseman (00:13:55): There are definitely lots of things in life that you really have to learn the hard way. I have children now too. That's something that changed the last five years, and so this has really sinks in where it's like I can tell my kids this, but they're going to have to go learn this, and I can try to help them avoid it, but sometimes that's what it takes. So it was definitely something where I always thought I was doing a better job of this. It's the same thing where you write some code or you write it in Excel or something. You look at it a few months later, and you're like, "Who was that person that wrote that? Oh, it's me."

Justin Mannhardt (00:14:21): Right.

Austin Senseman (00:14:22): So, yeah. I definitely was always getting better at it. There's a friend of mine, Janet King, who's an entrepreneur I've been working with, introduced me to a podcast where they introduced this idea. So if you had to teach a monkey on a pedestal how to read Shakespeare, how much time would you spend working on the pedestal? A question you might ask someone who is doing a startup just to see what they say. I mean, the right answer is you would spend zero time on the pedestal. The hard part is teaching this monkey how to read Shakespeare, but working on the pedestal is nice, and simple, and safe, and easy, and-

Rob Collie (00:14:56): It would be really convenient if the pedestal were adjustable, if it could come up to my height. Some days, I want to stand up and teach the monkey. Some days, I want to sit down. Why not have a hydraulically-adjustable pedestal?

Austin Senseman (00:15:08): You got it. Exactly, and there's a thousand reasons why you could improve it, right? You could work on that for three years and make zero progress on the most important problem in front of you, and so part of the "simpler is better, less is more" thing is like, can you simplify things down to the teaching the monkey problems? It's really easy to do a hundred things and not solve the most important thing. So those two things are very related, and I can just tell you I spent a lot of time. This is true for a lot of people who are in startups. You feel like, "Oh, I spent a lot of time working on that, but we didn't solve our most important problem, and so we're not making the kind of progress we want."

Rob Collie (00:15:44): You got a great pedestal.

Austin Senseman (00:15:46): Yeah.

Rob Collie (00:15:46): "We researched on the market, and there were adjustable pedestals you could buy off the shelf, but they didn't really quite meet our standards, so we engineered one from scratch. We're on Version 17."

Austin Senseman (00:15:58): Yeah.

Rob Collie (00:16:01): I'm familiar with this mindset.

Austin Senseman (00:16:02): I didn't get introduced to that idea until just a few months ago, but as soon as I heard it, I was like... It resonated with me, so.

Rob Collie (00:16:07): You said so many things in life, you have to learn the hard way. I have a name for myself for that category of things that you have to learn the hard way. I call it everything. So you were watching the developments in the Microsoft world?

Austin Senseman (00:16:22): Yeah.

Rob Collie (00:16:23): Keeping up to speed with it a bit through those special projects. I'm sure you had multiple ideas in mind, multiple thoughts as you got back down to the base of Mount Kilimanjaro. You probably already had some thoughts in mind. Let's get really direct into it. Why back into not just P3, but the Microsoft ecosystem that we're so focused on? You probably weren't aware of, for instance, the whole Fabric thing. That was still tightly under wraps.

Austin Senseman (00:16:50): Yeah. I hadn't caught any wind of that.

Rob Collie (00:16:53): But you did see things like Copilot coming, and there were clearly things you saw on the horizon or sensed that were telling you the next phase is going to be different. At the same time, you've also hinted to me backstage before we started recording there's also something about the world that is very much the same. The world's needs are still very similar to your first stint at P3, and this is something that's really easy to lose sight of, especially as a citizen of LinkedIn. Everyone is always talking about the next big thing over, and over, and over again. Constantly, constantly, new, new, new as if the old was conquered and the basics... I keep saying the next big thing is the basics, getting the things that should be table stakes done and done well for the first time ever. Do you share that view of things?

Austin Senseman (00:17:44): Yeah. When I left full-time employment for the first time in 2015 to do consulting work, that was my feeling that you could walk out of any building you were in, walk down the street, walk in the first business that you saw, find the person who was in charge, and ask them do they feel like they could do better with their analytics, "Could you have better numbers to make decisions in your business?" and that you were going to get a yes, and you could walk to the next business and get a yes, the next one and get a yes. That's the state of things where everyone feels like they could be improving in this area. I don't think that changed one bit.

Rob Collie (00:18:16): Yeah. That's a really powerful thought experiment. Let's even dial it up a bit. Walk out of an office building, walk into any other office building. Not only if you ask the questions, "Could you improve with your blah, blah, blah, your data, and all that kind of stuff?" not only would you get a yes, assuming they're not terrified of you sizing them up as a meal, the other question that drives it home, I think, even more is walk in there, "Could we add tremendous value to this business with the data stack that we have access to?" The answer is going to be yes. Objectively, yes.

Austin Senseman (00:18:50): I absolutely feel that way because analytics is downstream of the things that leaders care about. The real question is, "Are you still trying to make your business better?" The answer, of course, is, "Yes." "Okay. Could you use better information to help you do that?" The answer is, "Yes." I've been back here since the middle of October, and I did a jumpstart a couple of weeks ago. It was the first one I had done since I've been back. Got in there. Two full-time people had been working on something for six months, couldn't figure it out. By lunch on the first day, we had solved that problem.

Rob Collie (00:19:20): Yeah.

Austin Senseman (00:19:22): I think it was confirmation for me like, "Oh, yeah, this is still it." Right?

Rob Collie (00:19:25): Yeah.

Austin Senseman (00:19:26): It felt personally satisfying. By the way, the jumpstart is three days, so we're 16% on the way through it, and we solved the main problem they have, and now we have, basically, four more days to do more. So that was cool, but it was also really satisfying for them because they immediately saw what was possible, and it changed their whole outlook on what the next two years of their business is going to look like. For context, a jumpstart is for folks that are trying to get off the starting line with Power BI where we get leaders from the business and technical people in the same room. We often do it in person, but there's also a virtual option, and we really compress months and months of work into a few days of really focused effort where a working model that solves a real problem by the end of it, and then a roadmap too for the organization to think about the next 90 days, 6 months, 12 months, so they can walk out of that room with something that's working, and makes an impact immediately on their business, and have a good sense of where to take things.

Justin Mannhardt (00:20:22): Everything Austin said is true. I always add this additional understanding about jumpstarts that it's very important for me and helped me... informed my own experience executing them as well. When a customer is interested in a jumpstart and you tell them, "At the end of it, you're going to get a working version of something that solves a real problem. You're going to get a roadmap," there's very tangible takeaways, very tangible deliverables, but what you also get, you get this theater of excitement. Right? You get this theater of possibilities that you've just refused to entertain. You have to deliberately say that at the beginning of these things. It's like, "I know that if I don't say this, you won't bring things up because you've just trained yourself not to ask anymore."

(00:21:05): So, to me, the most important thing that happens in the jumpstart isn't that thing you get at the end. They're very important. The most important thing that's happens is that theater of excitement. You realize there's a better way to do this thing called analytics, and you are excited that you can move faster. I would much rather leave a jumpstart with that theater of excitement and go like, "Oh, yeah, and we're totally going to throw this model away because we're going to build a better one." Right? We know we can do that in a few hours. That's my spin on it.

Austin Senseman (00:21:35): Also, yeah, there's the self-interested version of that too like in a sales situation. What happened at this jumpstart in the afternoon of the third day, we got on the call with the four most senior people at the company. I'm not talking about in their department. I'm talking about all the way at the top. We showed them all this stuff, and you're just seeing, "If you're the one trying to get this done in your organization, this will be good for your career." Right? Like, "If you just approach this totally selfishly-"

Rob Collie (00:22:01): I'm hearing the monster truck voice coming on, "Step into the theater of excitement. We'll make you a hero."

Austin Senseman (00:22:10): Yeah.

Rob Collie (00:22:10): "Sunday, Sunday... No, it's not. We don't work on Sundays. Tuesday, Tuesday, Tuesday."

Austin Senseman (00:22:19): The theater of excitement.

Rob Collie (00:22:21): There's pyrotechnics going off, right? It's just like a KISS show. Ironically, I've been to a KISS concert.

Austin Senseman (00:22:32): Nice.

Rob Collie (00:22:33): Every KISS concert in recorded history starts with the same announcement. I don't know who it is, but they've got an announcer that comes over the PA right before they come on stage and says, "You asked for the best. You got the best." Can you imagine if that was our walkout routine, right? It was the theater of excitement, the jumpstart, right?

Austin Senseman (00:22:56): Yeah.

Justin Mannhardt (00:22:59): I feel that way when I do these things. I get amped.

Rob Collie (00:23:02): Yeah. I mean, it's the theater of excitement. I mean, you can't walk in their board.

Justin Mannhardt (00:23:08): When I was a consultant and I was out doing these, I had this mental clock on that first day, and I was like, "Okay. I'm going to do the whiteboard session. We're going to get clear on what we want to do, and then I'm going to get plugged in the data. I want to move really, really fast, and I'm going to put something really scrappy and ugly up on the screen as soon as I can, and we all watch it happen together." I remember Matt Selig. He and I had this experience. He looked at it, and he was like, "What is this? It's kind of... Oh, oh, oh, oh." Right? Then, it's like, "Okay. Now, we'll start making it pretty or whatever," and he's like, "I had no idea that was in this." Now, we're in a totally different clip and a different ball game, and it's so fun. It's just so fun. I need to go do a jumpstart. That's what the conclusion of this. I need one so bad.

Austin Senseman (00:23:54): You do.

Rob Collie (00:23:54): I'm essentially doing a pro bono hockey jumpstart right now.

Justin Mannhardt (00:23:59): Yeah?

Rob Collie (00:23:59): Yeah. It has truly been re-energizing me.

Austin Senseman (00:24:03): I love to hear that.

Rob Collie (00:24:04): It's so good.

Austin Senseman (00:24:05): We tell this stuff to people. Right? These are the messages we put out into the world, and I had done a lot of that in 2016, 2017, 2018, and there was a guy in the room who was an accountant type person, which is my background, who was their Excel guy. He had never seen Power BI before, so he was also in there. He hadn't been working on that project that had been going on. He was just there to understand what's happening, and to watch that guy's face in just those first four hours, you just knew his life was really, really different from that moment forward in a really positive way, and I love that. It's like what got me here in the first place, and it is still what the work is like today. It hasn't changed one bit. That guy had never heard of Power BI, and that's not rare. That's still the case. You leave your office, and you walk down the street and walk into a business. That's still the case in that business, and the next one, and the next one, and the next one. So there's just so much more work to be done here.

Justin Mannhardt (00:25:02): I remember when you got back from that jumpstart, even just hearing you talk about this even today. We've been talking about Fabric and AI. I personally get lost in the maturity curve. I've been doing this for a long time, and I've been working with the same customers for a very long time. You can easily forget that there's thousands, and thousands, and thousands, and thousands of businesses that are still ready for those face-light-up moments like you were describing. Incredibly powerful and transformative for the business, for the people involved. It's fun to hear you talk about it again because I was like, "Hell yeah." Yeah, we've got new tools, and we've got new things we can pull out of the bag. The fundamental shift that happened with the toolset many years ago is still there. It's still super powerful.

Austin Senseman (00:25:50): Yeah, and Rob has been talking about if people feel anxious about getting off the starting line, and if you're listening to this and you are one of those organizations where you feel like a little stuck or you have some instincts around like, "I don't know. This could be going faster," if you have the resources to bring one of our people in for three days, get some of your leaders together, get your data people together, put them in a room for three days, it will totally transform what you think is possible in your organization. It will put you on a different trajectory. From those experiences, we work with people afterwards. Sometimes we get people in a place where they decide they want to invest in their own internal capabilities. However it ends up working, whether we're involved after that or not, we have changed some people's lives, but it's very often the case that we come back. We continue accelerating. We're back again, and again, and again, and it just keeps pushing the envelope for them about what they can do. The jumpstart, it's fun for us, but it is one of the most impactful things that we do for people.

Rob Collie (00:26:51): I still just love the concept. Jumpstart is just, at its core, is, "Let's sit down together, virtually or in person." Here, we do a little bit of planning in terms of really lightweight, sticky note level of what your goals and ambitions are. We don't get very heavy with that though because what it's really about is, "Let's load data, and let's start moving towards that goal," and you're there for it. The iteration happens in front of you. You can steer the conversation, and those kinds of stories, like two full-time employees working on it for... What was it, six months, a year?

Austin Senseman (00:27:24): Mm-hmm, six months.

Rob Collie (00:27:26): Yeah. So one person year and not done.

Austin Senseman (00:27:31): Right.

Rob Collie (00:27:31): So one person year and counting, and what, done by lunch the first day?

Austin Senseman (00:27:37): Yep.

Rob Collie (00:27:38): Now, probably, they look at it and go, "Oh my God, we can take this so much farther than we ever expected we would or expected we could." Their ambitions grow. You pass the original finish line that they never got to, by the way. You pass it without even noticing it. You're beyond the original conceived finish line before you even know it.

Austin Senseman (00:27:56): Yeah, and there's another element of this. For me, it's really important that the morning of that first day... or maybe we oftentimes we'll do a meeting the week before just to set the stage. Get the people highest on the org chart that you can get into a meeting in a room and tell them this, "When you first started here, you were really curious about the business. You had a thousand questions, but people told you that they didn't have the data or they didn't know how to solve that for you, and I am going to implore you to get back to that feeling of wonder and curiosity because we're about to show you that much of that is possible with the right tools, with the right skillsets," and to have those leaders tell you, "What do you really need. Not what do you think you can get, but what do you really need to make your business work? Where are you be in 12 months, and what's going to unlock that for you?"

(00:28:49): Then, we come back to that at the end of those three days, and then we can build a roadmap with them on how to get there. They see in real time what is actually possible, and then they can start asking for those things again. It's really powerful for leaders who felt a little stuck and not having the information they need, and they can see a clear path forward, so it's fun.

Justin Mannhardt (00:29:07): It reminds me of a convergence of several conversations I've been having over the last few weeks, and one of them is an episode we did with another customer who was talking about this idea of data strategy and how we came in and did the jumpstart and everything. It was like, "This is data action." When you describe that story, Austin, of the leaders getting together and getting back to that level of curiosity, I just think about companies that are out there thinking about their data strategy as being something wholly apart and separate from their business strategy. So there's this team somewhere off in a corner doing this data strategy thing while this other team over here is trying to do the business strategy thing. Really, what's happening in that first day is, "Let's just bring these two worlds and slam them right back together."

Austin Senseman (00:29:58): They just start to collapse on each other, don't they?

Justin Mannhardt (00:30:01): That's how I think you should absolutely be thinking about that, and I think they're just maybe not enlightened yet. They're stuck trying to solve this important problem with the tools that aren't going to get them there, and they're wrestling with this complexity of whatever it is they're trying to do. The accountant you mentioned, you enlighten those type of people, and they're just... The floodgates open, like you told the team like, "I'm not doing anything fancy. I'm just sitting here clicking buttons." Like, "Oh, yes, I'm very good at this. I know what I'm doing," and it shows like, "And you can do this too." That brings those two worlds right back together. I think too often, data strategy gets described as these architectural elements that are required, and it's like, "Mm, no." It's really just what you want your business to do, and obviously, information is a really key ingredient to that, so.

Rob Collie (00:30:52): Yeah. Data strategy. It starts with architecture. Data strategy is business strategy.

Austin Senseman (00:30:58): I will say that the folks that were working on this were not employees of the company. They were vendors, and they were working with Power BI. So this is something else that I have recognized that's really different from when I was last here. Power BI came out when? 2015?

Justin Mannhardt (00:31:12): Official year. Yeah.

Austin Senseman (00:31:13): So, 2017, there weren't a lot of people working on Power BI compared to now, and I think these numbers aren't right, but I think the scope of it is. If there were a hundred people working on Power BI now, whether those are individuals or companies, then there's like 10,000 now. So I think going out into the market and trying to buy Power BI consulting is a lot more complicated now. If you searched for DAX in 2013, 2014, 2015, you showed up on our blog. There was a small number of people that were talking about it, and that is completely different now. When I got on LinkedIn, there are thousands, and thousands, and thousands of people talking about this. So if you're an organization going on the world, you're like, "Hey, we need someone to help us build our capacity. We need someone to help us accelerate." You crack open your computer and open the internet. That's very daunting now. Y'all lived through this over the past few years while I've been off doing something else. How is that feeling?

Rob Collie (00:32:07): It's a mixed bag, right? On the one hand, competition was not something we really had to worry about for a while as a company. Like you mentioned, Google DAX in 2014, 2015, and there was one zealot who had been writing about it twice a week for five years saying this is going to change everything. A bunch of other people who were hedging or hadn't even discovered BI yet, but the ones who had discovered BI were hedging that maybe this new thing wasn't going to be such a big deal, and there was one nut who was all in on it. It was me.

Austin Senseman (00:32:40): Our kind of nut.

Justin Mannhardt (00:32:41): He's our nut folks, and we love him.

Rob Collie (00:32:42): It was the first time I'd ever been sure of anything like that in my life. It wasn't a risky bet. Even in hindsight, I look back and go, "No. It was obvious." Once you'd seen the things I had seen, which most people hadn't, it was obvious what was coming, so I wrote about it forever and practiced it forever. That gave us the first-mover luxury, and we always knew that we would inspire copycats, but of course, it's harder to compete for market share when you have copycats who, from the outside, look the same or can look the same, especially in a digitally-marketed business like ourselves, right? That definitely increased the difficulty of us getting our message out. On the flip side, who can complain about the market rapidly expanding?

Austin Senseman (00:33:26): Right.

Rob Collie (00:33:26): Our customers used to be wild-eyed revolutionaries back in the day. Again, you had to be a special kind of "crazy" to bet on DAX like in 2013, right? We didn't even have Power BI. All we had was Power Pivot. We didn't even have Power Query yet at certain points in 2013, and we had clients who could see it already, and we did amazing things with that, even relatively incredibly primitive version of relative to what we have today, right? Now, it is just the responsible choice. Donald Farmer, when he was on this podcast a couple years ago, said like, "Power BI has become adopted by default rather than adopted by choice." That's a little bit of a shame, right? Power BI gets adopted a lot these days just because it's cheap.

Justin Mannhardt (00:34:11): This is something we talk about, the increased frequency by which we encounter somebody that's already given it a go. This is what you're describing, Austin. They have this mentality or a mindset that Power BI can't help me. I tried it.

Austin Senseman (00:34:29): Yeah.

Justin Mannhardt (00:34:29): It's hard to pick on this, but there is a dilution away from why Power BI was special, why the technology was special, and why it opened up this way of doing analytics more effectively. I mean, I've been in situations where we come in and basically say, "Well, you're just not approaching it the right way, and you need to use this approach." There's someone who's like, "Oh, well, we can't tell our leadership that because they think it's just going to be just easy to take Power BI and sit it on top of where we have this other reporting tool." So it's a different type of education because back in 2017, 2018, it was new. It was bleeding edge still, and there's a lot more of that. It's a reminder like if you had a false start-

Austin Senseman (00:35:10): Yeah, call us. Yeah.

Justin Mannhardt (00:35:12): Power BI can absolutely help you, but you need the right skillset, and the right mindset, and the right approach. It is different, but it's not necessarily hard. It's different. People don't get that always right away.

Rob Collie (00:35:23): First of all, screenshot of your Power BI schema.

Austin Senseman (00:35:26): Yeah. Show me the model.

Justin Mannhardt (00:35:29): Yeah.

Rob Collie (00:35:30): Does it look like a bunch of separate tables that are all islands? Does it look like one big, wide table that's-

Justin Mannhardt (00:35:37): A continent?

Rob Collie (00:35:38): Yeah. It's exactly like... Is it Tangia, or is it Archipelago? Not like Stockholm Archipelago with lots of bridges, okay?

Justin Mannhardt (00:35:52): We got an email today from someone who I assume they're maybe like an analyst or something in a company, and they said, "My company is switching from Power BI to another product. Please help me explain why this is a bad idea. I know that the models aren't there, and there's just an approach issue that is very solvable and addressable." Once you solve that, then it's like you stop comparing these products. It's like, "Oh, yeah, they all just make charts and graphs."

Austin Senseman (00:36:20): Yeah. Well, and if they're struggling with that, with this tool, they're going to struggle with the other one too. I think that's the challenge there is the tool choice isn't going to just magically fix that.

Justin Mannhardt (00:36:29): Right.

Austin Senseman (00:36:29): I feel for that guy or girl who sees that we're going to fix a pedestal here.

Justin Mannhardt (00:36:33): Yeah.

Rob Collie (00:36:36): Here's the thing, right? All of these tools are marketed as if they're magic. You buy one, you try it out, and you don't get magic, so you naturally conclude, "They were lying. This wasn't magic." Right? So you're going to go get a different tool from a non-lying software vendor, right?

Austin Senseman (00:36:54): Mm-hmm. Right.

Rob Collie (00:36:56): So you're probably going to get some things with the other tool that you weren't getting with Power BI just by accident, but that's the test. If Power BI hasn't felt like magic, you haven't seen its best. You really haven't. I just put this on LinkedIn the other day because it was just blowing my mind. All these years later, I sat down with my roller hockey league's historical PDF scorecards. Think about that. They're PDFs of old games, and I'm dropping them into a folder. You think about just how dumb, and ugly, and raw, and primitive a folder full of black and white PDFs. You open them, and they're mostly error. There's just a handful of tables, but they're spread across different. It's just so ugly. You could never get anything out of these PDFs.

(00:37:43): No one would ever print one of these. There's no point in it. They provide no value. I don't even know why these PDFs exist. They're just a download format that no one downloads. Why would you download this thing? From that, I'm producing just crazy visualizations about things that have happened in this league over the years. It is. It's magic. How did you conjure this output from such an ugly, boring, dumb input? It seems almost like creation out of thin air, and that's with an ugly, ugly, ugly data source. There's not a whole lot of insight to glean relative to what there is in a business, but yet there's still things that are laugh out loud, "Oh my God, I can't believe that," trend. So, yeah. If it doesn't feel like magic, you really haven't seen the best of it yet. If a lot of these other tools don't feel like magic, you probably are seeing the best of it.

Austin Senseman (00:38:32): I don't give them so much grief

Rob Collie (00:38:34): Really?

Austin Senseman (00:38:34): This is probably an area where I differ a little bit. Yeah. I approach people like this. We live in a golden age of analytics tools. I've seen people with really great tools, just like we're talking about right now, do very poorly. I've seen people with very poor tools do very well. Clearly, I've made a choice. I work on Power BI. I've had a strong conviction about it for a long time now, and I do think it is a very good choice for people. Probably the best choice, but you can pick another tool and get a lot of mileage out of it. I just think that's a very honest and authentic place to start because it diffuses the notion that the tool itself is going to get you 100% of the way there. We've just been talking about people that are, "Oh, good. I've got Power BI. All my problems are solved." Right? Then, they've got these crazy data models that are slow and don't work, and it's like, "Oh, well, I need another tool." So I come at it from a little different angle.

Rob Collie (00:39:19): I don't think we disagree on anything of actual substance there. It's just a choice of emphasis.

Austin Senseman (00:39:24): Yeah.

Rob Collie (00:39:25): We absolutely disagree on the emphasis because I'm a maximalist like, "Why the hell would you ever, ever, ever use those other limiting tools?" Right? It's like the old Frasier Crane like, "Oh, less is more? Well, think about how much more 'more' would be if you can use that tool and be successful." There's another level, but yeah, it's not the differentiator. It's not the old philosophy. It's not a sufficient condition. Choosing the right tool is not a sufficient condition for success, and it's not necessary either.

Austin Senseman (00:39:58): I am getting closer to your more maximalist opinion here as time goes by because you see, if you have a tool like Tableau, which was ahead of the curve in the early 2000s, I think a lot the Power BI exists because Microsoft saw what was going on with Tableau and wanted to be more aggressive in that space. Tableau got acquired by Salesforce, and the ecosystem around Tableau doesn't really do much for analytics folks.

Rob Collie (00:40:21): Yeah.

Austin Senseman (00:40:22): Right? It's like there's a CRM system sitting there that's seen better days, and there is Slack. That's the Salesforce world. For people that make investments in Power BI, oh my goodness, that ecosystem around it supports all sorts of other aspects of analytics, continues to expand and evolve. All the things that are happening in AI are going to sit right on top of that. The Fabric story I've been talking about, it's an investment that just by sitting around and doing nothing becomes more valuable over time because Microsoft is making smart investments and large investments around that product. You just don't see that same level of commitment.

Justin Mannhardt (00:41:00): You just hit a nerve center in my brain, and this was around when Fabric was initially dropping into preview. This is different. This is a different level of pace and innovation, and then you see generative AI come around. This is not normal. The level of things that are happening and the fact that it's not purely additive, but it's additive and also tightly integrated, it's like, "Wow, that's a different thing," and you see even some of the other changes that are happening in the Microsoft landscape specifically. It's because they're realizing because of this AI game, we need to be more tightly integrated, so all these things are really going to work well, and you're just like, "Holy smokes." I think that's where the FOBO theme starts coming out, but that's all very exciting and compelling. You can still do so much just by the things and features that have been there for a decade, right?

Austin Senseman (00:41:59): It is a thick level of delicious icing, but it is, for sure.

Rob Collie (00:42:02): It's such a fundamental difference in lineage between Power BI and basically, every other tool on the market. There were the old Cognos and Business Objects tools which were fundamentally data modeling, OLAP, multi-table schema interrelated, meant to be a comprehensive view of your business type of tools, but they were slow, antiquated, very, very inflexible, hard to use. Then, there were the reporting services of the world like FORMAT, your SQL query as HTML, and those were wildly successful. Just wildly successful in the market in terms of number of seats sold and all that kind of stuff. Even Microsoft's reporting services was by far their most valuable BI product in terms of revenue for a long time.

(00:42:51): Power BI comes from the lineage of interactive, all-up flexible view of your business like Cognos and Business Objects, and also, the old analysis services, but they got it right. For the first time ever, they were the first time that that kind of tool was built correctly and was built for real-world conditions, not clean-room academia and all that kind of stuff. So it really is, even to this day, one of a kind, and most of the other modern tools that compete with Power BI have a little too much grounding in that reporting services' SQL lineage, and it's very, very, very, very difficult to shake that lineage if that's the place where you started. It's not just a degree of execution like some software vendor was doing a better job than the others. It really is a historical thing.

(00:43:37): In a series of articles I wrote a few years ago, I finally started to put this in historical context. Tableau did us all a tremendous service making graphical interactive dashboards, turning those into table stakes. Tableau, more than any other tool, demonstrated to the world the value of that and was like the meteor that killed reporting services type tools as the BI tool of choice. By the way, even at Microsoft, we knew that reporting services was a dumb tool compared to what analysis services could do, right? But it's like, "Well, people want it. Sell the hell out of it." Right? So making the world demand that kind of interactivity, they move the world, they move the market in a very, very, very important direction. I give them a ton of credit for that, but you know, now, time to pass that torch. In fact, I think the torch have been passed.

Austin Senseman (00:44:34): Microsoft is in an interesting place. I loved watching that Satya Nadella interview when ChatGPT had come out and Microsoft had that partnership, and he said, "Oh, we made the team over at Google dance. They're going to know that we made him dance." The look on his face. This is a guy who started working on Bing 20 years ago, but stuck with it, and lived through, and advance that organization. Microsoft, a core part of their strategy has not been to innovate, right? It has been to be fast-followers on things and win market share. They have great distribution. They become innovative in the later stage of the lifecycle of things.

(00:45:10): So, for example, they didn't invent Tableau, but they invented Power BI, and now Power BI is clearly ahead, and they do that with a lot of product categories. It's very effective. They are very innovative right now in their AI tools, and it's like an interesting space for them. Satya Nadella has taken them to a different place now where they are not fast-followers on things that got launched four years ago. They are now launching the most innovative product in a category, in a technological wave that is potentially as big as the internet itself.

(00:45:41): So it's really fun to watch, and y'all have talked about FOBO a lot for the past few months. I love it, and you said this a minute ago, Justin. It's so great to come back to being grounded in, "Hey, we can help companies do the fundamental stuff well," and that is incredibly meaningful, and it's what most companies need right now. It's not AI. There's that maturity curve. It's not AI. It's all these other things. It is like wrangle your data and your business strategy, collapse them into one, get going, get moved, and get off the starting line.

Justin Mannhardt (00:46:12): There was a project Microsoft has that's in a beta stage, and I spent a whole day just playing around with this thing, and so it's this generative AI experience where you can load up some data. So I grabbed some of our dummy data, and I uploaded a CSV, and I just started asking it questions. It started laying out this mind map of visualizations and suggested next questions. This is still very beta. There were some things that I was like, "Yeah, it's gimmicky," or, "That's not useful or valuable." But you see stuff like that, and you wonder, "Okay. Is that where we're headed?" But I always come back to like, "But if I can't get the basics, I don't get a found model representation of my business, none of that stuff is really going to matter that much." I attribute it to this loss of features we've had for a long time like Q&A in Power BI, or like the Quick Insights feature, or like... None of those are really all that exciting if you don't have that interesting or well-conceived model to begin with, and it's like I get a lot of comfort in my own FOBO. It's like, "Yeah, this is still going to matter. Maybe we'll click less buttons making the line graph."

Rob Collie (00:47:18): Generative AI solutions that take my data model that I've already built and do all of the chart layout so I'm not futzing around, aligning things, setting the fonts to be the fonts that I want over, and over, and over, and over again. Right? I eventually give up. There's an asymptote of font change that I'm forever approaching. Done. I'm never quite there. Zeno's font paradox. I could really dig being out of that business, you know? Austin, you were saying before your business might not be in the place where it really needs AI yet. Maybe it is, but good news is that either way, the way that Microsoft is architecting what they're working on, the first step is getting the Power BI stuff working right.

Austin Senseman (00:48:08): Yeah, a good data model.

Rob Collie (00:48:10): Yeah. What are they calling them now? Semantic models? Okay.

Austin Senseman (00:48:13): Right.

Justin Mannhardt (00:48:13): Thanks, Christian.

Rob Collie (00:48:16): No. You got to love Microsoft. Let's mess with the nouns. Right? That's the pedestal.

Austin Senseman (00:48:23): Yeah, it is.

Rob Collie (00:48:24): Clearly, we need to change the name to something really fundamental that we've already had for a while.

Austin Senseman (00:48:29): But not "Dashboard." Let's leave that one alone.

Justin Mannhardt (00:48:32): "Put the monkey on a red pedestal." A few years later, "Put it back on the blue one."

Austin Senseman (00:48:39): I mean, the way they've architected Fabric, it's all roads lead through the data model and Power BI there. Fabric can be perplexing. It's easy to be lost in that world if you get in there for the first time, and I think it is an important message that it's like the semantic model. The Power BI model is still right at the center of all that, and a lot of things hang off it in an important way.

Rob Collie (00:48:59): I've really been coming around to this, and I might've even mentioned this in a previous podcast. It's like the way that Microsoft names and hypes this Fabric thing is understandable from a marketing disruption standpoint, but it's almost counterfactual the way they talk about it, at least obfuscating. The simple way to talk about it is that, A, Power BI is becoming a lot more important because most of Power BI is building these data models, building these semantic models. That sounds techy and cheesy, but no, these semantic models, the model that underpins a Power BI report is like this first time you've had an authoritative source of information about your business in a comprehensive end-to-end way.

Austin Senseman (00:49:49): It's a question-answering machine.

Rob Collie (00:49:51): Yeah. All the good words for this have been taken, right? Oracle speaks about a company, right? Although it's ironic because most data that goes into an Oracle database never comes out. Right? It's a black hole. It doesn't answer questions. It eats it. That data model becoming useful for so many things other than just building reports, great, but also, the Power BI philosophy, the underpinning philosophy, the way they've approached all of this is now being taken to everything else as well. "Move fast." The citizen developer thing, that target audience, even if you end up with high-end people operating these tools, by building for these "citizen developer audience," it keeps them honest. It keeps them in a place where the tools are going to meet the real world where it is, not in some scientific academic clean room. It yields a toolset that moves at the speed that you need. If that's all Fabric was described as, Power BI itself is more important and the Power BI philosophy is becoming more important, that would calm everybody down.

Austin Senseman (00:50:58): It would even if it was just the first part of what you just said too.

Rob Collie (00:51:01): Yeah.

Austin Senseman (00:51:01): I think for me with Fabric, I am in wait-and-see mode. I mean, I say that as we are actively working on projects to do migration, so that's happening, but I think professionally, I'm just more of this stage in my life like a show-me person. So when we talk about it, I find myself nodding. I listen to you guys on the podcast, I'm like nodding along. I agree, but I still want to go see it out in the wild knock some of these projects out of the park, and we're going to do that, but until I have really seen it, I do reserve just a certain level of caution that's made it fun to do these projects because customers who are going into Fabric right now are those same kind of customers in 2015 who were like, "Hey, we want to do DAX." It's like if this just came out of GA, we want to do it right now, and it's fun. You like working with those types of folks because they're in it, they're ready. I feel very good that we're going to do well in those projects, but we're rolling something out for the first time.

Rob Collie (00:51:56): We've reached the point. Austin is now more skeptical about software than I am.

Austin Senseman (00:52:00): Yeah. Yeah, yeah, yeah.

Rob Collie (00:52:01): Something flipped in the intervening five years. I'm positive I'm still more skeptical about software than you are.

Austin Senseman (00:52:06): I don't know if you are at this point.

Justin Mannhardt (00:52:08): It seems like you did your job, Rob.

Austin Senseman (00:52:09): I don't know. That's just one of those lessons that you have to experience enough.

Rob Collie (00:52:15): It's sort of that everything category?

Austin Senseman (00:52:17): Yeah. That's right. That's right.

Rob Collie (00:52:20): I could go all Yoda on you and say, "Listen, I planted the seeds while you were here the first time, but then I've realized you needed to go out into the world."

Austin Senseman (00:52:27): You did plant the seed in the sense that you have a very forceful, concise opinion about that. I think last time I was here was the first time I had ever worked or been around someone who felt that strongly about the software skepticism, where at that time in my career, I was like, "Oh, yeah. We just need a new tool to solve this problem."

Rob Collie (00:52:47): Your batting average was pretty amazing with that though.

Austin Senseman (00:52:49): Yeah. Well.

Rob Collie (00:52:50): That's how you end up digging yourself a hole. Having a good batting average on something like adopting new software is how you really get deep, right?

Austin Senseman (00:52:59): Yeah. You have the wrong fundamental model, but it just happens to be working.

Rob Collie (00:53:03): Yeah.

Austin Senseman (00:53:03): Yeah, and I think it's the "simpler is better, less is more" thing which I keep coming back to. A lot of times, new software is the pedestal. "Oh, Power BI is networking. We just need to click." Right? It's like, "No." New tools can be an easy dodge working on the real fundamental things that are going on. "Oh, this new tool will fix the problem." In Conserv, I did that enough, and I was the CEO. I was the person making all those decisions, so it wasn't like there's anywhere to hide. It's like, "Oh, new tool this, new tool that." It's like, "No, this is... Put these five numbers in Excel. We're going to look at that. How simple can it be? Let's focus on the real problem." So, yeah. I've come around in a significant way there, Rob. I think we're pretty close to each other at this point.

Rob Collie (00:53:45): I just have a sense, an instinct, and a history with these people, the people who are behind Fabric where I can see like, "Oh, oh, they're locked in that way." Right? I just go, "Oh, it's lights out." I have a special branch in my software skepticism tree for what they're doing, and it has a lot to do with the people. The thing is, is like even if they don't get it right out of the gate, they're going to get it right, and they're going to get it right in a reasonable timeframe. I can quiet that part of my brain for the if part, and it just turns into a sliding scale of when. That's where I'm at now that I have that simplified explanation of what they're up to.

Austin Senseman (00:54:35): Right.

Rob Collie (00:54:36): You know?

Austin Senseman (00:54:36): It's good to hear you say it that way because I think Fabric... In the same way, if you could go back a decade, it's not that you were writing about DAX and Power Pivot at the time, but it was also you knew how to take the messaging and translate the things that Microsoft was putting out and make that make sense for people that were doing analytics work for leaders that were trying to make a decision about. You knew how to talk in the language of the business that people could really pick it up. A whole wave of people that work here now because they heard that and it resonated with them. I think we have an important role to play in doing that again for Fabric.

Rob Collie (00:55:17): I agree.

Austin Senseman (00:55:18): Microsoft doesn't yet know how to talk about it. Yeah. I'm looking forward to hearing some Rob translation services here to help, and getting in there, and doing the projects, and working on it. We're going to learn.

Rob Collie (00:55:29): It's not just that Microsoft doesn't know how to talk about it. It's that they're obligated to talk about it a different way.

Austin Senseman (00:55:36): Mm, yeah.

Rob Collie (00:55:36): They have to talk about it in some vague, scary, disruptive way. If they just said like, "Okay. Really, let me level with you. What we're really doing is this, this, and this." Right? It wouldn't make that compelling launch splash. You would never swear thousands of people to secrecy about what you're up to if you were going to market it like that. They want the word "Fabric" to strike fear essentially, right, in everyone.

Austin Senseman (00:56:09): Yeah.

Rob Collie (00:56:09): It nets out positive for them, but there's a lot of negative along the way, and you could get a mirror to sit down, and go on the record and say, "Well, okay. Really, what we're up to is pretty simple. It's this, this, and this." Right? But those simple things are so powerful.

Austin Senseman (00:56:22): Yeah.

Rob Collie (00:56:23): Yeah, you need a punchy, splashy word, and it's important to choose a word that already meant something else in the space as we like to laugh about.

Justin Mannhardt (00:56:30): Fabric has a data mesh in it.

Austin Senseman (00:56:32): Yeah, yeah.

Justin Mannhardt (00:56:33): I was thinking about... We do a lot of different things for our customers. Obviously, it all revolves around analytics, but two very different types of things that Austin has experienced in his role being back, we talked about this, jumpstart and you mentioned here briefly this Fabric migration that we're working on with one of our customers. I mean, this is advice to companies out there maybe that are trying these things and maybe feel stuck or they don't feel like they're getting the value is the substance of the work that's happening. What I mean by this is when we do something like a jumpstart, the very first thing that happened isn't the laptops open, and Power BI is there, and we're like, "Okay. Point me to your data." The first thing that's happening is someone standing in front of a whiteboard with a marker trying to get clear on what business problems and questions.

(00:57:19): So when we're doing this Fabric migration, it's a substantial effort. Right? This is an enterprise scale situation. The first thing that happened, there was a technical deep dive involved, but that was... Honestly, that's the easy part. It's easy to go rationalize, "Oh, you've got existing data warehouse A, existing data warehouse B, existing Power BI things, C, D, E, and F." The hard part is going around to all these functional areas and saying like, "What are you struggling with? Where do you want to be at the end of next year?" That's the guiding force for this type of project. Yes, there's all these important technical features and important capabilities that we believe are going to bring value to those problems, but the guiding force wasn't this pretty architectural diagram that someone saw and said like, "Oh, I should probably have that. That's a good data strategy."

(00:58:08): That's what I mean by data strategy is business strategy. Too often, that gets conflated like, "Oh, you need a storage layer, and a staging layer, and a serving layer, and a resting layer," and one of the things we're big believers in, "Let's only build what we need to solve that problem." Right? So if we need a great lakehouse in Fabric, yeah, we're going to go build that. If we don't, let's not. If we get a new customer, they might come at us because they're interested in Fabric, and they see what's happening, they want to get in on it. It's very possible to say like, "You just need the data model thing." That's the first step. Right? So I think that's really important. If you are doing some introspection on where you've been, where is the substance of the effort, and how close is it to the pain or opportunity you're feeling with the business, not with the state of your technology?

Austin Senseman (00:58:58): Yeah. It's like the P3 double-whammy kind of because I feel very confident. Technically, we have some of the best people in the world working here, and it's a big differentiator for us, but all of those people are also very skilled at helping understand why we are trying to do things and help you navigate the world of getting to a place where you actually want to be, and stack those two things up together, and it's wildly powerful. The jumpstart experience isn't just solving technical problems. That would be important, but by the end of it, it's like, "Hey, we have a roadmap for where we can take this that's going to have significant improvements in our business." It's a really special thing. I think you just described it really well, Justin. It's different. I do think we're still trying to sort out how to best communicate that to the market, and since the market is more competitive, there's lots more noise out here, it's like I have a very high conviction, that's why I'm back here, that we have a significant role to play and just having way more impact on the world, but it's a little harder to get that message out now.

Rob Collie (01:00:02): I have this pithy thought in my head. You call it data strategy, and you're already off on the wrong foot. Now, you've got a means to an end, data strategy is a means to an end, but you've named it as your strategy. So, now, you're going to go and build the pedestal. You're not going to be Shakespeare at all if it's not serving the business. I mean, even just calling it a data-driven business strategy would be a massive upgrade. If you just change the name of what you were talking about, you're going to be more successful because it moves the bullseye in a very tangible way back to the true north, magnetic north. There's so many metaphors. We should just call this The Metaphor Podcast. Let me tell you how this relates to modern philosophy which ironically starts with René Descartes in the 1600s.

Justin Mannhardt (01:00:51): We should do a metaphor rundown though. Rob, that might be a backlog project.

Rob Collie (01:00:54): Oh my God.

Justin Mannhardt (01:00:55): So we can shove all the transcripts to the AI machine and like-

Austin Senseman (01:00:59): Yeah, every metaphor.

Rob Collie (01:01:01): Oh, that's my whole 2024 spoken for... That's what I'm doing, The Metaphor Inventory Project.

Justin Mannhardt (01:01:07): A good day with ChatGPT, and you'll get-

Rob Collie (01:01:09): You think so? I defeated every podcast. There you go.

Austin Senseman (01:01:12): Yeah.

Rob Collie (01:01:12): That's it.

Austin Senseman (01:01:13): Yeah.

Justin Mannhardt (01:01:13): Yeah. We're doing this next Thursday.

Rob Collie (01:01:15): Then, can we train it up so ChatGPT could host the podcast and see how well it does standing in for me?

Austin Senseman (01:01:28): Yeah.

Rob Collie (01:01:29): That would be so hilarious.

Justin Mannhardt (01:01:31): This is a fun idea, guys.

Austin Senseman (01:01:33): You could do both sides of it, Rob. You could be talking, but you could also read its responses or... Right?

Rob Collie (01:01:38): Oh, wow. Have a conversation with my AI-generated self?

Austin Senseman (01:01:42): Right.

Rob Collie (01:01:43): It'd be one of those things where I came away afterwards like saying, "That AI is all wrong. It's wrong. This, this, this, and this." It doesn't get me right in any of these ways, but someone else is looking at it going like, "Yeah, Rob just described himself. That thing was right on."

Justin Mannhardt (01:01:58): One of our people did a pet project. One of the things you can do is you can deploy GPT-4 and Llama 2, a bunch of these large language models in Azure, and so we did that, and we fed it... I don't know how many. We fed it a bunch of your blog posts from that 2015, 2018 era and started asking it questions like, "What does Rob think about how to handle," I don't know, "value above replacement for fantasy football?" for example, right, and see like, "This is going to get away from me if we don't book the day where we're going to sit down and do this because I think it'd be really fun."

Rob Collie (01:02:33): The podcast transcripts, there are many, many, many, many, many words in those things. There's a lot of training data.

Justin Mannhardt (01:02:40): Oh, yeah.

Rob Collie (01:02:41): If we can lab-grow a version of me in AI, if it can be done, we have the data to do it.

Justin Mannhardt (01:02:47): It seems like a very sound business priority to me.

Rob Collie (01:02:51): Yep. Totally. That's exactly what we should be working on. That and Indy Inline Hockey statistics. Clearly, business critical. If it reminds me of that magic, right, it's totally worth it. It's just so amazing. Seeing the members of the league respond to it and get geeked up over it, it's priceless. It's so good.

Justin Mannhardt (01:03:13): It's not just the magic of like, "Oh, look, I did this seemingly impossible thing with all these PDFs," but once you got all that information modeled up together, you're not just doing like, "Oh, look, I made a bar chart of the scores of these games." You're doing cool things that like, "Oh my gosh, I never thought you could show the streaks of how many games I scored goals in and doing all kinds of cool stuff like that." That same magical experience is what happens in the business. It's like, "Well, of course, I can show you your GL report, but I can also do all these other interesting things," and that's that opening of possibilities that you should experience.

Rob Collie (01:03:50): I did the 2D scatter plot bubble chart type thing of goals and assists in the league. There's these three clusters of dots with these wide no man's land deserts between them. Guess what? I'm in the cluster of dots at the bottom left, right? The ones near the origin.

Justin Mannhardt (01:04:09): No assists.

Rob Collie (01:04:12): Very, very, very little production, and then there's these really solid players in the middle. You'd expect there to be a continuous transition between these three clusters, right? But I mean, there are no dots cluttering the transition zones. Spotless. It's so weird to see how sharply clustered everything is. To me, that will be probably the most interesting thing that I find out of this dataset, but of course, I put a play axis on that just so people can see dots move. Right? I don't know if that's worth much.

Justin Mannhardt (01:04:52): That's pretty cool.

Rob Collie (01:04:53): Yeah.

Austin Senseman (01:04:53): It still captures people's imaginations.

Rob Collie (01:04:55): Yeah. I mean, getting engagement, right?

Austin Senseman (01:04:57): Yeah.

Rob Collie (01:04:58): Even if the play axis isn't really particularly an animation, isn't going to be capturing it, but I've got an animated bar chart where we can see the season points race. But again, it's only that upper cluster who's going to care because no one else is in the running.

Austin Senseman (01:05:14): This is a topic that I did want y'all's opinion on. It came out of the jumpstart I was at a couple of weeks ago, and if you have a good data model, it's really easy to build reports for people. It's almost too easy in the sense that it is really easy to build a lot of noise.

Rob Collie (01:05:31): Yeah.

Austin Senseman (01:05:32): Mm-hmm. Yeah, and you see this with software products too. This was relevant to me because I experienced this at Conserv. It's like, "Oh, well, if we just add one more feature into this, we're..." and then you do that for a couple of years, and you're like, "This is a mess." Right? But you have analysts being product managers now because they have to think of what they build as a product they're going to put in front of people for them to do some kind of job. Again, it's a non-technical thing, but how do we help them learn how to navigate that well? Because for this jumpstart, where they started is an artifact of... What these vendors had built was probably 10 pages of Power BI reports that on careful examination and talking through to the leaders, it was like, "Yeah, this really isn't helping us." They had produced some things, right? But it was like, "We're not using this." That's a real trap.

Rob Collie (01:06:17): Yeah. It's dangerous. I call it going forward from the data like, "I've got this data. I've built this data model. Let me show you some things I can do with it." This is natural because you want to educate people on the art of the possible. You want to help them regrow those curiosity muscles, right? The atrophy of curiosity. You want to show them that the walls of their invisible prison have been knocked down, but along the way, you end up showing them because you're fishing for something that you don't know yet about what they need to know. So you end up presenting things to them that aren't useful to them. That's a very dangerously disenfranchising behavior. Now, in the hockey example, I'm absolutely going forward from the data. I am 100% just saying, "Check out this novelty. Look, a balloon animal."

Austin Senseman (01:07:07): Yeah.

Rob Collie (01:07:07): Right? Right? Some of them are hits. They're not all going to be zingers, but yeah, that is a delicate balance in a real-world scenario where people don't know the art of the possible yet.

Austin Senseman (01:07:22): You have a lot of experience as a product manager, Rob, in that environment where you're trying to think about features, the overall experience, and different kinds of users. How does that translate to Power BI?

Rob Collie (01:07:34): A lot of the lessons that I, again, learned the hard way while at Microsoft designing software translate to Power BI, but also, things that I had to learn outside of Microsoft working with Power BI would translate if I went back. I had an old friend, Eric Vigesaa, on the podcast, and we were laughing about how we both had grown up in an era at Microsoft where all features and all product ideas started with this sentence, "Wouldn't it be cool if..." which is just the worst way.

Austin Senseman (01:08:06): Right, right.

Rob Collie (01:08:12): It's like this data forward thing, right? You're just trying to make a magic trick with the technology as opposed to working backwards from what the world really needs. It's hard to work backwards from what the world needs. Much harder, I think, at Microsoft than it is in our shoes because we get direct contact one audience at a time. I had a manager that used to say, "What we do is trying to order a pizza, one pizza that a hundred million people are going to like. Good luck."

Justin Mannhardt (01:08:44): Yeah.

Rob Collie (01:08:45): Power BI, everything we do is software development.

Austin Senseman (01:08:48): Yeah.

Rob Collie (01:08:49): But it's faster moving software development than C++.

Austin Senseman (01:08:52): With lots of feedback.

Rob Collie (01:08:54): And it's aimed at an audience, right?

Austin Senseman (01:08:58): Yeah, yeah.

Rob Collie (01:08:59): So we have a tremendous advantage as "product managers" in our role here at P3 relative to what the people in Redmond have to do. It's crazy hard.

Austin Senseman (01:09:09): It's something that we... I don't know how much we talk about it, but it's something that we unintentionally teach to organizations when we're in there working with them. It's a side effect of just sitting down next to someone at the table and working through it.

Justin Mannhardt (01:09:23): I pulled something up that I had some notes on when Austin was asking this question about the role of an analyst becoming a product manager, and I started writing some thoughts on this. It was really informed by my own experience working with people on the same problem. Three things I see people trying to build, and there's more than three, but the three things are what I call like a Swiss Army knife. They're trying to build a report that answers every conceivable possible question and every conceivable possible combination of questions, and what you end up with is this unusable thing. If you find yourself building a Swiss Army knife, you've missed an opportunity to empower people to work with the model in the more exploratory ad hoc avenues. Right? So if you find yourself in that scenario, that's a cue to like, "Okay. Maybe I need a timeout because I don't need more reports."

(01:10:12): Swiss army knife is one. The other thing is... I call it a precision instrument, and these are really valuable like the things we're going to look at all the time like our executive dashboards, our team dashboards, things that are very tightly wound to the action loop, Rob, like, "This thing serves a very specific and clear purpose." Then, the other thing is... I call it the examination table. I had to go through this. You were describing your customer churn thing, Austin. Very complex, specific, interesting problem. I don't think this belongs in this episode, but I just thought of that. This is a really important topic, Austin, of helping analysts think about... A new report is not the solution to everything that's going on, and that's one of the things I see a lot when we go do these assessments is you just see all these Swiss Army knives slaying everywhere, and you're really like, "Oh, what you really need is four or five precision instruments." I thought of that, and it's been a long time since I resurfaced that idea, so.

Austin Senseman (01:11:07): We need to get both of you on a jumpstart.

Justin Mannhardt (01:11:10): Yeah. Let's do it.

Austin Senseman (01:11:11): You should do it.

Rob Collie (01:11:13): Anyways, this was great.

Austin Senseman (01:11:15): Yeah. It was very exciting for me, and thanks for having me on the show.

Rob Collie (01:11:19): Oh, thanks for coming.

Speaker 3 (01:11:20): 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.