The Plural of Alex is Alex, w/ Alex Dupler and Alex Powers

Raw Data By P3 Adaptive - Podcast autorstwa P3 Adaptive - Wtorki

Kategorie:

We welcome Power Platform expertise in the form of Two Alex! Alex Dupler and Alex Powers both work at Microsoft. The organization they work for and their first names aren't the only thing that these two share! They also both have a lot of experience with and passion for the Power Platform. Alex Powers is a member of the Power BI Customer Advisory Team (PBICAT), and Alex Dupler is a Program Manager focused on BI & Data Infrastructure. These guys know data! Follow Two Alex: Alex Dupler Twitter Alex Powers Twitter Two Alex Youtube Channel References in this Episode: Raw Data with Brad and Kai from Agree Media Episode Timeline: 7:00 - The woes of Stack Ranking, Data storage options, more fun with names! 22:00 - What draws you to data?, The value (and drawbacks) of Excel, and the path to Power BI 36:40 - Two Alex-similarities and differences, Rob tells a story of someone crossing him, and one of Rob's favorites-the art of using BI to drive action 59:00 - When BI and IT collide, the 2 Alex's non-traditional BI path, the value of being an expert even if you aren't THE expert 1:16:00 - Two Alex LOVE helping people, is there value to documentation?, knowing the Business portion of Business Intelligence 1:37:00 - Advertising performance discussion Episode Transcript: Rob Collie (00:00:00): Hello friends. Today's guests are Alex Powers and Alex Dupler, collectively known as Two Alex. They're both Microsoft employees in very different roles, but both have their feet rooted firmly in the power platform. You might be familiar with their YouTube show. I interact with them primarily on Twitter and a little bit on Reddit. And this is the first time I've had really any conversation of length with Alex Powers. And it's the first time I've had any conversation at all with Alex Dupler. And no surprise here, really, really cool people. We had a lot of fun, really dynamic and inspiring, interesting conversation that wound through a number of topics, including some show favorites, like non-traditional backgrounds, and closing the action loop, and imposter syndrome. We talk about how years ago Alex Powers wrote a review of my book that called out the intermission in the book and how, what a delight that was at the time to read. Rob Collie (00:00:57): And that leads to a conversation about how we're always essentially at our own little intermission in our expertise curve. You're always in the middle somewhere. And if we started doing metrics on this podcast, you'd probably find that this one ranked very highly in opinions expressed per minute. Ooh. What could he mean? Let's get into it. Announcer (00:01:21): Ladies and gentlemen, may I have your attention please? Announcer (00:01:25): This is the Raw Data By P3 Adaptive podcast. With your host, Rob Collie, and your cohost Thomas LaRock. 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 (00:01:49): Welcome to the show. Alex Powers and Alex Dupler. How are you today, gentlemen? Alex Dupler (00:01:54): I'm doing great. It's great to chat with you. Alex Powers (00:01:56): Rob, back-to-back meetings. I'm glad that Luke found us some time here. I was so hesitant about this podcast, just cause I love listening to it. I was like, "I don't know, should I do it? Should I do it? Should I do it?" Rob Collie (00:02:08): The answer is yes, you should do it. Alex Powers (00:02:10): I appreciate Alex D and Rob just pulling us all together. Yeah. Rob Collie (00:02:13): We've already backstage a little bit been laughing about this. So let's bring it out to the front stage. The two of you combined, what do we refer to you as? Are you the Two Alex's? Or something different? Alex Dupler (00:02:23): So we learned separately from our wives that the correct pluralization is two Alex. Rob Collie (00:02:30): See, I just don't buy this. I still think Alex's. I mean, we could get really funky and say, Alexi. Tom LaRock (00:02:36): I was going to say, that's what I think. Yeah, Alexa, Rob Collie (00:02:39): But I mean, think of it this way. There's fish, and that's plural. But even there, there's still fishes, which refers to different species of fish. Yes. I think. Is that what it is? Alex Powers (00:02:51): Yeah, that's right. Fishy. Yes. Rob Collie (00:02:53): I don't know. So the two Alex, are you guys seriously going to go by that now? Is that going to be the new thing, or? Alex Dupler (00:02:58): Well, the YouTube channel is called Two Alex. Rob Collie (00:03:01): How'd the two of you come to know one another? Is it just like, oh, we're both working in data and we're both named Alex. So you're like, you see each other from across the room and your eyes meet across the internet? Alex Powers (00:03:10): I would say across the internet, for sure there. Just because he's up in Redmond, I'm kind of located in St. Louis, Missouri. From there it was kind of this, I think natural, just both being active in the community. Alex D you can keep me honest there, I'm sure we were connecting on Twitter a little bit there before, definitely in the subreddits. One of my earliest memories of was, Hey, this thing isn't folding. And I was like, oh my gosh, it's Power Query. I've got to tackle this. I've got to answer this question. Reddit is where I hang out at. I would say from there that's when we really started coming chat more and more, but Alex D I'll let you kind of tell your side of the story. Alex Dupler (00:03:43): Yeah. Yeah. My recollection is that the first time we interacted with each other, where it wasn't just some random poster on Reddit, was side conversations on Microsoft teams within Microsoft. There's some internal discussions where salespeople can get their question answered and sometimes the questions are interesting. And so, yeah we had some side conversations. Plus back then, when Alex wasn't on the product team, he didn't always have full visibility into the roadmap. And so we would chat on the side about what we would do with the roadmap. Not that we would do a better job, just a different job. Rob Collie (00:04:19): Yeah, I get you. Yeah, I understand. I understand. What are the two of your roles at Microsoft today? Alex Dupler (00:04:25): I work for Microsoft advertising. We're the organization that sells the ads that go on Bing, as well as some partner websites like Yahoo search and AOL search and stuff like that. And I work in the business function of the sales org. So I do BI for a sales team. And it just happens to be at Microsoft, and that influences the technologies that we use. IPM are like data warehouse and big cube stuff. Rob Collie (00:04:50): Cool. We're going to have to circle back to that for sure. And Alex P what are you up to these days? Alex Powers (00:04:56): Yep. So senior program manager on the Power BI customer advisory team, so PBI CAT for maybe those out in the community. I'm called as kind of that last bastion of hope sometimes, where I'm not very close to the solution, not close to the architecture, just come in and fix it. Where Alex D, he owns the solution, he owns the finished product. That's a line of visibility that I completely lose in my day to day. But you get variety, you get to do different things. Some days it's maybe a DAX challenge, next day I'm writing C#. The next day, I'm writing kind of new report, kind of clicky, clicky, draggy droppy experiences. So a vast rich tapestry of Power BI. Rob Collie (00:05:32): So you're on the CAT team with a number of people that have already been on the show, right? Adam Saxton, Casper, Chris Webb. You're part of that crew? Alex Powers (00:05:41): Yep. Rob Collie (00:05:41): I hear that that crew continues to expand, it's like this great gravitational attractor. It's like just hoovering all of these people. Let's just have it on the record. Does the Power BI CAT team have ambitions of world takeover? Alex Powers (00:05:53): Every day. And I think what you're seeing right now is a lot of formality. Community contributors, experts, decades of experience. They're now turning into bosses, they're now turning into managers. So they're getting further away from the technology and kind of now being people managers. I'm enjoying our livestream here because Rob is laughing. He's like, oh, I know that exact feeling. Rob Collie (00:06:14): I do. I do, right. I got a request today from some media outlet to interview me for Power BI tips. And I'm like, gosh folks, I'm probably not that person. You want to talk about strategy, okay, that's different. But I have gotten further and further. I still build some Power BI stuff for sure, for my own purposes. But I don't have that day to day, like, this is my life. That's not how my day goes anymore. I'm back to the management game after years of being out of it. Yeah. Growing a company tends to keep you out of the actual hands-dirty data trenches that started the whole thing. Alex Dupler (00:06:52): Well, if you ever start stack ranking, that's when it's going to be time to sell it. Rob Collie (00:06:56): True story, stack ranking was the reason why I actually stopped being a manager at Microsoft. At one point, I just said, I'm done applying the system for you. I was sick of it. And I understand it's gone now. I found out the hard way that stepping back from a management position didn't just relieve me of that stack ranking thing that I found immoral and uncool. It also took me out of a lot of the important conversations. I just didn't have nearly the input or influence that I had before. And that was hard. If I was still at Microsoft today, my career at Microsoft would still have suffered like a multi-year setback because of this era where I just said, I'm done. I know that at this point, the whole stack rank thing has been gone for a long time, but it was still a number of years later after I left that it still persisted. No, we're never going to do that. We're never going to play lifeboat with human beings. I mean, it really sucked, right? Basically, if you built a really good team, either by recruiting or by development or both, you were punished for it. Alex Dupler (00:08:04): Yeah. Apply this to Alex's team. You want to stack rank Chris Webb and Casper and Adam? Tom LaRock (00:08:09): I will. I'll do it. Rob Collie (00:08:11): Which one of them gets told that they had a terrible year? Right? Tom LaRock (00:08:16): I'd be happy to do it. Rob Collie (00:08:20): Hey, listen. As long as we put that kind of phenomenal power in the hands of a benevolent tyrant, like Tom, it's perfectly safe. What could go wrong? Alex Dupler (00:08:29): That is what they said about solar winds. Tom LaRock (00:08:34): My first criteria, having known them for many years, is Jaeger consumption. So we'll just start with that and work our way down the stack. Rob Collie (00:08:44): Which way are we going to sort that list though? We sort it largest to smallest, or smallest to largest? I mean, I could see that list being sorted either way. Tom LaRock (00:08:50): We'll try it both ways and see how it shakes out. Rob Collie (00:08:53): Yeah. I mean, it could be like a honeypot, right? Put some Jaeger out there, see who goes for it? You're getting the 3.0. We won't be doing any of that, thankfully. Now, Alex P, you were previously in a different role, right? Alex Powers (00:09:10): Yes. So, here at Microsoft less than two years, came in through the premier field engineer side to support, really had a blast there kind of proactive engagements training, probably train like 4,000 Tableau users on the Power BI. So just like the grind of doing it day in, day out, talking about the product, I just absolutely loved that. Transitioned to kind of field sales roles. There it's competitor competes, a lot of disinformation where they're saying, well, Power BI can never do this. What do you mean it can't do that? Here's an article. Here's me, kind of the whizzbang demo. That's probably where I got my hyperlink chops for those that kind of know me on the community. Alex Powers (00:09:44): This is the good and bad of the pandemic is like, Hey, we're making some career advancements, we're working long hours, whatever else it may be. A lot of my goal whiteboard over here was, Hey, I want to be on the Power BI CAT team. Had that visibility, just kind of did those grinding over the fall and winter months when we're all stuck inside. But I'm sorry, Thomas. I don't know how good I would be at the Jaeger thing, just because I don't have that peer connection. I haven't met my coworkers. So that's tough for a lot of people that I think are just making career jumps during the pandemic right now. Rob Collie (00:10:16): Yeah. I mean, it's weird. I live in a completely altered reality where we've been a hundred percent remote, I've been a hundred percent remote for 11 years. Probably more closer to 12, actually. Our company was a hundred percent remote from the beginning, basically out of necessity. To me, it's shocking how many people who've been at this company for a long time have never met each other face to face. We did a gathering, a team gathering in 2019. We didn't do one in 2020. I don't remember why we didn't do that. We haven't done one this year, either. We're hoping to maybe do one in 2022. We've hired so many people in the last year that there's like half the company that I haven't ever been in person with. Alex Powers (00:11:02): It's tough. Rob Collie (00:11:03): It's different, isn't it? Alex Powers (00:11:04): Yeah. I think it was like the good meme the other day where it's like, Hey, here's your company culture, it's just like an empty cubicle. And it's like, well, people don't even have that anymore. It's just, here's your new job, here's your new email. Log in, welcome to the company. Great friend of mine, Mark Beedle, I know kind of joined T3 adaptive. I love that he's like, this is where I want to be. I think of the P3 of the past, where you take the group, I think, up to Seattle or some of the different areas. And then it was like, oh wow, they're all getting together and having fun. You know, I tried applying for the job, but unfortunately your Excel file was corrupt and I couldn't pass the test. Rob Collie (00:11:36): Oh, I see. I see how this [crosstalk 00:11:38]. Alex Powers (00:11:38): Yeah, what happened with that, Rob? Rob Collie (00:11:39): I don't know, man. Alex Powers (00:11:40): That's really what I wanted to corner you on today. Rob Collie (00:11:43): That might've been part of the test, Alex. Alex Powers (00:11:45): I literally thought it was, that responded that way. I was like, I don't know if they're testing me with a corrupted file. Alex Dupler (00:11:50): Yeah. You need to have mastered the Open XML format of the Excel file, and be able to track down the corruption in the Power Query. Rob Collie (00:12:00): I saw a joke or a meme on some social media a couple of years ago about cast iron, the hipsters with their cast iron and how you have to take care of it and everything like that. And then after you're done with that, you have to dry it in the sun for 24 hours. And someone goes, 24 hours? And they go, yeah, if you're not willing to go to the Arctic, you don't deserve cast iron. So it's like that kind of test. Yeah. Alex Dupler (00:12:23): We beat the crap out of our cast iron, it's just fine. Rob Collie (00:12:26): Okay. And now Alex Dupler. You're working in BI in the advertising wing, within Bing but also the affiliated networks like Yahoo and things like that. And so you mentioned that you're in charge of the data warehouse and you're in charge of, you said big cube. Alex Dupler (00:12:42): Yeah. Rob Collie (00:12:43): For a year I worked on Bing, and maybe this is a completely different dataset than what you actually end up caring about, but the state of the world back then was there was this giant distributed commodity hardware database system, data storage system called Cosmos. Alex Dupler (00:12:58): Yep. Rob Collie (00:12:58): One of the world's foremost write-only data stores. It was amazing at storing data. You could never get anything useful out of it. There was only one person in the entire organization, named Jamie Buckley, who was capable of actually running queries against this thing. And so if you wanted any information whatsoever about what searches were being run and things like that, yeah sure, you could try to write a query against this thing. And what would happen is you'd get syntax error after syntax error after syntax error, and then eventually you kick off a query and it wouldn't give you any errors. And you're like sweet. And it would run and run and run and you go away and you'd come back like a day and a half later and then you'd get a runtime error. Alex Dupler (00:13:38): Yeah. And when it works, you get a CSV. And so we still have that. I think when I was getting trained on it, which they said it had something like 5% of the world's data in it. Cause it's not just Bing, it's X-Box and a whole bunch of stuff. It's this really cool exabyte scale thing. But nobody knows how to use it, partially because it uses scope scripts, which the only commercial product they've ever been used in is the ATLS gen one analytics feature, which was not a successful product and is being deprecated. And so you can't hire people that know how to use it, there's just like a bunch of vendors that have learned it. And I can't write it either. Also, I don't know if this was your experience, but the engineers are allergic to writing documentation. It's got these petabyte sized tables with 400 columns and there'll be a data dictionary and it doesn't have any descriptions of any of the columns. Rob Collie (00:14:33): This does match my experience, yes. Alex Dupler (00:14:35): So we use that some, we also have other partners. I mean, it's a huge organization. We just missed getting touted in the quarterly earnings as having crossed $10 billion for the last fiscal year. I think the public number is like 9.95 or 9.5 billion. Yeah so it's a real business, even though the market share is pretty small. It turns out advertising is just a really, really good business. So we take a bunch of data out of there, and then also from partners that take data in there, and put it all in Databricks and make it available to folks that way. And we love Databricks because our analysts, they can come with whatever skills they have, and they can be successful on day one. Because they don't have to learn SCOPE or KQL or whatever. Alex Dupler (00:15:22): They can write Python, they can write R, they can write SQL, there's a cube so they can do Power BI, they can do Excel. They can do whatever they want, all in the same data. Now, if they want to do things that are super fancy, they may have a hard time using the cube. So they got to write something. Rob Collie (00:15:41): Yeah. Alex Dupler (00:15:42): But if you're a PM owning a project, you can drag and drop in that cube all day long and have a good time. And then the other thing that we like about the setup we have is, with the data in Data Lake, our partners that have their own generous Azure budgets, they're not running queries against our server. Whereas if we put it in Synapse or SQL, when they want to query our data, we're paying for this compute. But here they just mount it onto their own compute system, and they pay for it. And that's great. We like when other people pay to use our data. Rob Collie (00:16:15): So it's funny, I actually expected that the answer to the question was going to be, oh no, no, we fixed all that. That original system is completely straightened out, it's got a much more human friendly interface. But it turns out that you just have other systems that are human friendly. And those things have to... on the order of one-time investments to figure out how to populate those things from the great Oracle that is Cosmos. Alex Dupler (00:16:41): Yeah that's largely true. I mean, in Cosmos, they've implemented the ATLS APIs. So you can mount data in Cosmos directly to a Spark engine and do stuff that way, if you want. Yeah. Basically that's how they've done things. You will not be surprised to learn that Microsoft likes to reuse names. Maybe you've seen this phenomenon before in the word power, but yeah. Cosmos, the internal exabyte scale data platform is not the same as Cosmos DB, the Azure product, which is for, I couldn't even describe it. It's for, like, everything. Rob Collie (00:17:19): Yeah. I mean, there's only so many cool nouns. And furthermore, the set of cool nouns in the world is further refined by the ones that computer scientists gravitate to. So you end up with a really small population of words. And the chances... It's like the pigeonhole principle from math, right? You need 450 names, you only have 300 words. So you're screwed. And so you end up with things like the word dashboard being repurposed to mean something kind of niche in Power BI. That's one that I wish we could get a do-over on. And you know, I'm a sinner. I named some things poorly in my day. I'll give you an example. When PowerPivot V1, and actually several versions of PowerPivot, at least in 2010, there were those two drop zones, extra drop zones in the pivot table field list, for slicers. Rob Collie (00:18:11): Cause Amir insisted that we make slicer layout really easy as opposed to tedious. So we had these extra drop zones, and one drop zone put the slicers down the left-hand side of the pivot table and one put them across the top of the pivot table. What did I name those two zones? Horizontal and vertical slicers. For years after that, when I taught that product to classes, they go, oh, what does a horizontal slicer do that's different than a vertical slicer? And I just sit there with my head in my hands like, it should have been left and top, Rob. Why did you... Previous Rob, why were you so nerdy and stupid at the same time? Left and top. Alex Dupler (00:18:48): Well you see, in an indimensional cube, there are some things that are horizontal and some things that are vertical. Once you understand what the tubal is, it'll all make sense. Rob Collie (00:18:59): Yes. So let's go back to basics and... Yeah, no. It's just left and top. Yep. These are what you call own goals. Can't make these things up. It's even funnier, by that point in my career when I made that mistake, I was already kind of like this rabid high priest of naming. Like, we should be better. And here I was in the course of delivering those sermons, just committing tremendous sins out the back of the church. It's just like. Alex Dupler (00:19:31): Yeah, it turns out we should be better in, oh crap, I got an hour before this presentation, what am I going to call this thing? Those are two overlapping states of being. Rob Collie (00:19:41): You know, people's hearts are in the right place. So I still think that the two of you probably might've gravitated toward each other just a little bit, maybe like 1% more, because of the shared first name. Can I be allowed like an extra 1% gravity on this? Alex Powers (00:19:54): 99. I mean, a lot of Alex's within Microsoft that are doing Power BI, we've all kind of banded together. Rob Collie (00:19:59): There's like an Alex crew? Alex Powers (00:20:01): Hell yeah. Big time. There's multiple Two Alex's, too. Rob Collie (00:20:04): As we've established, once you get above like three or four Alex, it's suddenly Alex's. That's when it becomes plural. Alex Dupler (00:20:10): There are at least two Alex's working at Microsoft in the Power BI ecosystem that are smarter than either of us. Rob Collie (00:20:17): Well I mean, going back to something we were talking about earlier, every single person, every single consultant at P3 is a hell of a lot better at Power BI than I ever was. I can't even argue that it's like, oh, I'm off my peak. It's not that at all. They were always going to be much, much better. It's very humbling. Like in the real sense of the word, when you sort of get put in your place. Alex Powers (00:20:40): Is this like a time thing, Rob? Cause I feel it too. It's like the early days, Power Pivot and Power Query were something like, I'm digging, I'm learning all of these things. And then like everything else is kind of passing me by and it's like, yeah I'll catch up to that at some point. And I see the wild stuff that people are doing nowadays, like, I don't know what nights and weekends they're spending learning this product, but I'm working twice as hard and I'm still not catching up. Alex Dupler (00:21:00): Yeah. I was watching the other demo the other day. And he was talking about how you should have your report and your data model in two separate PBIX's. This was Mike Carlo. It was a great demo. But then he was like, and to make this really easy, what we're going to do is we're going to edit the PBIX. And I was like, hold on a second. You can't do that. That's not allowed. Rob Collie (00:21:22): [crosstalk 00:21:22] Like actually hacking the file? Like he got into the file structure? Alex Dupler (00:21:25): Yeah. Rob Collie (00:21:26): I do love me some file hacking. For me, I think it's not necessarily a question of time. It's actually that the universe has returned to its default state with respect to me. Which is, the whole time I worked at Microsoft, in all the years I was on the engineering teams, I worked with plenty of people who were super technical, but also enthusiastically technical. When VB.Net came out, and ASP.NET, I had some colleagues that just dove into that. They loved it, it was the most amazing thing. And I just could never... I was still at that point going like, okay, well I learned how to write my VBA, and I'm sticking with it. That's where the frontier of my coding, actual procedural coding, is still VBA six. Rob Collie (00:22:09): For some reason, DAX and data modeling, as technical tools go, DAX and data modeling really, really spoke to me. Like I freaking loved it and still do, still do to this day. And Excel formulas are kind of the same thing, right? This is the handful of exceptional technologies that really seem to appeal to my nervous system, and none of the others do. And by the way, M is another example that does not appeal to me at all. Alex Powers (00:22:38): I am the opposite. I love M. Rob Collie (00:22:39): Really? You love M? Alex Powers (00:22:40): I love, love, love. Hell yeah. Rob Collie (00:22:42): You're not my species then, you're something completely different. Alex Dupler (00:22:47): So I think one of the big things that drew me to data modeling, so there's a lot of constraints. And with programming, it's like, there's such an open world. Like the only programming I could ever really get my head around was VBA. That's where I started. You didn't have to have a big, complicated object model. There was just Excel. That was your object model. And it made everything so much easier. And you're like, okay, well, what I'm trying to do is move these cells to those cells. And with data modeling, especially in Power BI, it's like, well, I need one column for these relationships. And I need these relationships to flow in one direction. The constraints make it a much more manageable problem, but also opens up room for more creativity. Rob Collie (00:23:30): I agree. And also VBA comes with a macro recorder, the world's greatest set of training wheels. It's like, if I want to build an app from scratch, I can't like act out like pantomime what the app will do, and have something spit out code for it. Alex Dupler (00:23:48): Draw some stick fingers in Figma and just drag them around, and get some code from that. Rob Collie (00:23:51): Yeah. It's like, mock up the UI in Balsamiq or something, or Vizio, and then start mashing on the screen with your finger and say, okay. And then speaking out loud, what should happen at that... there's no macro recording for actual software developers. Alex Dupler (00:24:04): I think we got to tell Charles that, that's what he's got to do with his AI driven power apps development. Rob Collie (00:24:09): Yeah. It's we need to turn this into a LARPing thing, right? You just act out the application in the real world with these cameras... Holo lens. There it is. We've solved the world's problems. Take that for low code development. Alex Powers (00:24:26): Well, I like how Power Automate's now watching your points and clicks, and generating flows for you. Rob Collie (00:24:32): See, I didn't know that it did that. Alex Powers (00:24:33): Oh yeah. You're training the machine. You don't even have to write the code anymore. It's like, oh, automation is here. It's really here now. Rob Collie (00:24:41): It's always a feel good moment to meet a fellow VBA 6-er. The world used to be lousy with us. We were just everywhere. It's kind of a dying art. Office has got this new JavaScript API, Office Scripts. That's incredible. Again, in theory. I haven't touched it, because it's not reaching out and grabbing me by the eyeballs. I'm tempted though. It's sort of like, oh, a new VBA six and they have a macro recorder and I'm like, okay, maybe, maybe. This might be the way I learn JavaScript someday, is Office Scripts. Alex Dupler (00:25:09): Yeah, that sounds like how I'd learn it, except Excel is dead to me. I mean, I use Excel for note taking and PM stuff, but data work, I don't use it. Because first of all, Power Query is the way to go. And in Excel, when you have Power Query over, you can't save the Excel file. Rob Collie (00:25:28): Really? Alex Dupler (00:25:29): Yeah, Power Query takes a lock, like a lot of the old school windows. And you can't get back to the main- Rob Collie (00:25:34): Modal window. Alex Dupler (00:25:35): Yeah. So you can't save, you can't refer back to the data. You can't open stuff. And it's not like Excel ever crashes when you're working with lots of data. So saving, it's not that important. And if you want to say, first you have to evaluate your queries or set them to disable load. But if you've already loaded some, if you do something to disable load, it destroys the cells. I just said, I'll do it all in Power BI. No more Excel. Not because there's anything wrong with Excel. It's just that that user experience was just so unacceptable to me. I lost so many hours of work. Tom LaRock (00:26:10): Wait, what do you mean, not that there's something wrong... Clearly there's something wrong with Excel. Alex Dupler (00:26:14): Yeah. Rob Collie (00:26:15): Alex, you're cut from a cloth that I understand very well. Your sarcastic cynicism is, ooh, it speaks to me. Yeah, we've come to the right place. Even I, team Excel guy, I am really on team Excel. I haven't written any DAX in the Excel environment in several years. It's all Power BI, all the time now. Alex Dupler (00:26:38): The other big thing is why would you want to write DAX in an environment that you can't schedule to refresh? Unless you don't have pro licenses, like... Alex Powers (00:26:47): Hold on, let me challenge you now. Here we go, this is a little taste of Two Alex. So I love Ken Puls, where he's saying, Hey, I don't want the heavy weight of Power BI. If I can do as much as possible within Excel, be it Power Query or even Power Pivot. I would agree that. Alex Powers (00:27:03): Be it kind of power query or even power pivot. I would agree that the development experience is severely lacking. That's not to say that the power BI side is the best in the world, obviously Dax studio, et cetera. But I would much rather take a lightweight application over a heavy one every day and then just import that data model into power BI when I'm ready. Rob Collie (00:27:19): To me, the primary value of these technologies in Excel is as an on-ramp to the power BI universe for the authors. Tomorrow's power BI authors are today living in Excel. And the reason, I've said this multiple times on this podcast of multiple different people at Microsoft, but the reason why I'm, I don't want to use like the passive aggressive version of the word disappointed. Let's use the completely neutral version of the word disappointed. The reason why I'm disappointed that there isn't more investment there is because that is the gateway drug, and as a universe, as a community, like we really need to care about bringing those new people on. And that's where they're going to come to. To tell those same people, "No, put Excel down and start learning this in a completely new environment," their immune systems reject that because they've been sold a million times on the idea that something's going to replace Excel. They know better by now. Rob Collie (00:28:23): But no one in that category, like the V lookup and pivot route, none of them resist the idea of there being crazy, powerful new versions and features of the things that they're already doing. You get them 48 hours into that new world, and they're more than happy to switch to the power BI environment. They're excited about it. Those same people who would have rejected it 48 hours before. You got to take them on that path and this thing not getting the love that I think it deserves, I understand it's from the perspective of our real production environment is the power BI environment. I get that. But the on-ramp, they are doing some things about that, even things that I didn't know, because they're targeted at people who don't know about this stuff and I already do. Brian, when he was on the podcast, was talking about how they're using machine learning, advanced clippy generation seven, to detect the people who should be interested in this stuff and sort of pointing them to power BI. And there actually was really good uptake of that. That feature didn't fire for me because I don't use V Lookup or regular pivot tables anymore. Alex Dupler (00:29:23): That's almost exactly the journey that I went on. Like many of your guests, I did not go to school for power BI. I actually, I went to school for chemistry and I worked as a chemist for a couple of years. I was doing lab work and I was very bad at lab work. I mean, I understood the chemistry, but I would break glassware that was expensive and stuff like that. Which when you make $15 an hour, breaking expensive glassware is a good way to get in trouble. So I was like "Okay, well I grew up in a very computer centric family. Maybe I can do some of this Excel stuff." And so I was doing visual basic, and we were doing some dashboards, like operational reporting. And I had Excel in this company. I loved the people there, but it was not a successful business. We had maybe a hundred thousand dollars in revenue per employee with high CapEx, because we had these big, expensive instruments that we had to buy and chemicals and all sorts of stuff, lots of HVAC. So there was not enough money to pay people to live in Seattle, so every office license was a battle. Alex Powers (00:30:31): Wow. Alex Dupler (00:30:31): I was looking at, okay, what can I do with Excel 2007, because we had some of that I think we had enough licenses, but it didn't really check. So we didn't really pay too much attention. But then I was like wanting to use power query because I had sort of discovered it was easier, but I couldn't. So I was like, "Okay, how do I get this macro to run as a service so that I can refresh these dashboards on these dowels that we bought second hand?" Rob Collie (00:31:01): You know, if it weren't for the a hundred thousand dollars of revenue per employee, at a certain point, that story sounded like season two of Breaking Bad. The HVAC, the cap ex, oh you mean a hundred thousand dollars per employee per week? Okay. Alex Dupler (00:31:18): No, no, no, per year. Rob Collie (00:31:19): Then it's meth. Alex Dupler (00:31:20): Yeah, no. So this is the environmental testing industry. And the way it works is your tests have to be defensible to the EPA. So the EPA puts up a spec and says the test needs to be done this way. And when it's done, it has these parameters in terms of statistical reusability. And that means that one lab's product is a commodity compared to the other lab's product. And so you can't get outside profits. All you can do is compete on service and price. And if you take a high CapEx business and bolted to professional services, you're not going to get good margins. Rob Collie (00:31:59): Unintended consequences of everything, right? Alex Dupler (00:32:01): Yeah. I mean, Rob, can you imagine your business, if you are charging professional services business model, but you bolted on a whole, huge amount of consumable costs to every delivery? Rob Collie (00:32:14): Yeah. It sounds like we can safely not choose the wine in front of me. Alex Dupler (00:32:19): That's how I first encountered the 2017 standalone web, maybe it was 2016. The first time power BI was split out. I was doing office 365 admins and I got like a push notification. I was like, "This is cool." And I built some stuff and I showed it to my manager and he was like, "That's cool. How much is it?" "$10 a month." "Nope, can't afford it." And that's when I started looking for jobs anywhere where they had good Excel people. Rob Collie (00:32:46): Yeah, and to put that in perspective, this is the punchline to many jokes when people ask us how much it is. We go, "It's $10 a month per user." We all just start laughing. Like, "Oh my God, it's like stealing. It's so cheap." Alex Dupler (00:32:58): I didn't even have that many users. Rob Collie (00:33:02): I mean, this might be $30 a month. You know, like, nope. Alex Dupler (00:33:06): No. Rob Collie (00:33:07): It's like when we first moved to Cleveland back in the day, it was right in the middle of the financial crisis. We were looking at real estate and everything. And there were houses for sale in Cleveland for $10,000, like $10,000. And I started laughing. I'm like, imagine the deal going down. This house has been on the market for 180 days at $10,000. And you come in and say, "Look, I've got a cash offer. I'm willing to pay asking price. But the grill out back? You need to leave it." And the owner's like, "Mmmmm." Alex Dupler (00:33:43): My wife's best friend lives in Cleveland and they recently bought a house. And so we looked at a bunch of Zillow listings. I'm like, "Oh man, we could pay cash. Move next door." And they're sort of north of the Cleveland Clinic, that super nice neighborhood up in there. I was like, "Oh yeah, we could buy a very nice house, but our family is not there." Also, have you looked at the weather? It's not Seattle. Rob Collie (00:34:05): No, it's not Seattle, but I'll tell you what, here's an interesting description of statistics. When we moved to Cleveland, it wasn't because we wanted to, it was because basically my kids had been taken to Cleveland and so we're trying to console ourselves. We're like, "Okay, well, okay. It's going to be colder. There's going to be snow. Okay, okay. But at least it isn't going to be as overcast." And then we looked it up and Cleveland has more overcast days per year than Seattle. So we were like, "Damn, that sucks." However, it turns out that the definition of overcast days is very, very, very important. Because like an overcast day in Cleveland is like 75% of the sky is covered by clouds. That's an overcast day. At no point in time ever is Cleveland under a one mile thick, oppressive blanket that starves you, where you don't even have any idea where the sun is in the sky. So number of days is one of those misleading statistics. Total amount of oppressive cloud cover needs to be a different statistic. Trust me, there's more sun in Cleveland on an ongoing basis than there is in Seattle. Those winter and fall months, man, those are rough. Alex Dupler (00:35:16): That's not the part that bothers me. I was born and raised in Seattle. It's the shoveling your driveway in March, that part of it. Rob Collie (00:35:24): You could just be the delinquents that we were and just get a four wheel drive vehicle and say, "Screw it." Alex Powers (00:35:31): You don't shovel in March. After February, you don't shovel. It's in my contract. I don't shovel after March 1st. That's it. Because it's going to melt. Alex Dupler (00:35:40): Eventually. Alex Powers (00:35:40): It'll melt by the end of the month. Rob Collie (00:35:43): By the end of the month. Alex Powers (00:35:44): By the end of the month, it'll be gone. The rainstorm's coming. Sunshine's going to happen. I ain't shoveling. No, I put that in my contract years ago. Rob Collie (00:35:53): Cleveland's where I learned the rule, we do not adopt dogs that require walking. Alex Powers (00:35:58): Yes. Rob Collie (00:35:58): They need to be able to go out in the backyard and come back in. I fell so many times on ice. I eventually got, they're like crampons essentially. Alex Dupler (00:36:06): Yak tracks? Rob Collie (00:36:08): Yak tracks, that's what they are. Yak tracks or something else. You can't intentionally slip on yak tracks. It's crazy. But without them, just any day now broken hip. Alex Dupler (00:36:19): Our friends that lived there, they just got a golden retriever. We met the puppy when we were visiting with them this summer. Very cute, but I think they have some of those walks in their future. Rob Collie (00:36:29): So you start looking for a job, that's what led you into Mount Redmond? Alex Dupler (00:36:34): Yeah, I literally went looking for jobs good with Excel in Seattle. I found a contract position into Microsoft, making sure that the salespeople were assigned to the right customers and got paid on the right quota because advertising the agency model, it makes that much more complicated. Because we were in this model where we'd try and keep all the customers of an agency with the same salesperson which makes a lot of sense, especially when you're the underdog and you have relatively few sales resources, you get more leverage. But customer's change agencies all the time, have no respect for our compensation cycles, and so it was quite the nightmare. Rob Collie (00:37:15): Yeah, I love that. Like, so here's how we'll define the world for our benefit, "Oh world, you did not get the message. World, please don't change. Don't have your own things going on." Yep, that sounds like a software engineering problem from the nineties back before the industry kind of got wiser. So you start talking about the show. It seems like with a format like that, it's got to wander, which is what our show does too, by the way. So what are some of the most entertaining or valuable corners that you found yourself wandering into over time? Alex Powers (00:37:48): I'm still excited with our first episode where we talked about kind of beyond the desktop where it's no longer just development in [inaudible 00:37:56] desktop. It now almost takes like five different applications to build something at scale, which is like a good and bad thing. Well, you're getting more tools, seeing new things faster, more performing, et cetera, but why do I need 10 tools? Can we solve that within the desktop application? And we just had a really good conversation, a lot of attendees there, providing their own thoughts. And it kind of comes back to like this overwhelming feeling of learning power BI it's. Like I have to learn 20 new things all the time, learning, learning, learning. It's just never ending. That was my key episode. Alex Dupler (00:38:26): I agree. I mean, I think that's been the central theme of the whole show. I mean, we did that first episode and then we've talked, we've had the same conversation about these tools in so many different contexts. What are the different ways to do dev ops in power BI? What are different ways to measure how you're doing in terms of the effectiveness of your models? And so all of that is sort of external to the desktop application. Alex Powers (00:38:52): I think the best part, too, is that we're not from these traditional backgrounds of 20 years of BI or 20 years of kind of dev ops. We're learning this real time, sharing our experiences of, Rob, I think you call us power users or business users that find these tools, that get empowered by this technology. That is the seat in which we sit in. Hey, I found Excel 2010 power query add in 2013, 2016. I'm fighting with my IT admins. Can you please just upgrade to the next monthly release that will solve all my problems? Where Alex D is on the other series fighting tooth and nail for a 2007 license. It's kind of funny to hear that conversation. Rob Collie (00:39:32): $10 a month. We need those charity commercials like Sally Struthers used to do. This is Alex Dupler. For $10 a month, less than the price of a cup of coffee, you could get him an O 365 license. Alex Dupler (00:39:50): Yeah. So we ended up getting some E3 licenses and some E1 licenses, which meant I could work in power query, not using my personal license, but using the company's license. And then when I tried to share it with my coworkers, they only had E1. They couldn't use desktop. They had to use power query online or Excel online. And there was no power query online. And so even once we sort of modernized, we installed like a windows server 2012 and it was already 2015 and I was okay, this is our modernization. Alex Powers (00:40:26): So Rob, I'm going to steal one of your quotes here if you don't mind. Rob Collie (00:40:30): No, please do. We have an open source quote license at P3. Alex Powers (00:40:35): Well, I'll buy you a 2007 Excel license too, if I have to. Rob Collie (00:40:39): Fantastic. Alex Powers (00:40:39): But one of the items you had said a long time ago, I believe it was either in your book or maybe some of the video recordings you used to do in the studio with a nice button up shirt. You said, "There are two types of people when it comes to technology, those who can see the possibilities and bring about change and those who are about to be affected by it." And I always look at this and I look at things like had kind of talked about with the Excel users who haven't even gotten to this experience yet. There is still somebody out there today that is using Excel 2007 and their employer or their this, their that, or whatever their situation they're red is like, "I just see this little rectangle. It hasn't changed. Why do you want me to go invest in any of this stuff?" Like how often are you still seeing this? Rob Collie (00:41:17): Anecdotally in our public classes, which I haven't taught in a while, I've taught one as recently as let's say two years ago. And when I taught public classes for P3, I still stubbornly insisted on using the Excel version of this stuff to teach. Again, because of that onboarding effect. Alex Powers (00:41:34): Yep. Rob Collie (00:41:34): I think I was the only one left at our company that was still stubbornly doing that and I wasn't bothering to argue with others and whatever. So I did this for years, probably the first one of those classes I taught would have been like in 2011. So fighting with different versions of Excel, all the different students showed up with, for a long time, that was like a quarter of the class. The instructions for the class were very clear, show up with this version or don't bother. And they'd show up and no, they had a version, didn't even have power pivot and couldn't get power pivot. And so it was so bad for a while, we would bring spare laptops. If we traveled to another city, we'd be lugging spare laptops with us and they'd just be there ready to go like the hot swap with a student. That problem really went away though. I reached the point where I'd survey everybody at the beginning of class, "What version of Excel are you on," or whatever. And everyone, every single person in the class would be on the basically some version of recent 365. Rob Collie (00:42:32): I really do think that the person who's trapped on 2007 or hell even 2010 or 2013, they're out there, but they are really a tiny, tiny fraction of the world now. Whereas that used to be an overwhelming problem. So it's really testament to how successful O 365. Alex Powers (00:42:52): I would agree. Rob Collie (00:42:53): It's like, I was kind of like cynically betting against it forever, like the tortoise and the hare. Like I woke up one day and that's the world. The world is O 365. I think everyone's on the modern, not everyone, but it rounds to everyone, is on the modern wave of the tools. But they're still shocked when I show them, when we show them, "Did you know that this is in here?" And they're just like, "What?" Alex Powers (00:43:19): How is that in here? Rob Collie (00:43:20): They get angry because they start to realize how much of their life they have lost by not being told. Alex Powers (00:43:27): What I was always seeing was people had to live in two worlds. Like I went to some of the Excel boot camps, Michael Alexander, absolutely transformed on my personal laptop. I'm having the best time of my life in these three-day boot camps. I'm loving, loving, loving. At the very end though, I have to go back to work on Monday. I saw what could be, and I'm now back to what is. And it's just very difficult to kind of live in that middle space. For those that are still out there and listening to this, Hey, look at your surroundings. Hopefully Office 365 is coming within your organization. But if not, kind of like Alex D's story, I just went looked somewhere else. I saw the future that was coming, and I bet it all myself and I went for it. And I think that me and him both kind of share those stories, too. Alex Dupler (00:44:08): Inside of Microsoft, in my little corner of the Microsoft that most people in Microsoft don't even know about, I put together a class that I've given a couple of times, Modern Excel for Managers. And basically I would just show them power query and X Lookup. We didn't even talk about Dax. But just to like get them thinking like, "Hey, if you're doing some annoying thing in Excel, maybe there's a couple ways to make it a little bit better. Maybe you've never even seen the formula bar before." I had one person that I worked with who I was like, I didn't handle it very well at first. But she was like, "Can you add these numbers together for me?" And I was like, "Yeah." Alex Powers (00:44:52): Just a standard Excel formula bar? Is that what you're talking about? Alex Dupler (00:44:55): She was like, "Can you show me the difference between these two numbers?" "I can do that for you, but here, let me come over here and show you something." Rob Collie (00:45:04): So there was another program manager on being, because I was such an Excel, I'd come from the Excel team, I'm such an Excel zealot, that all someone had to do was say that they needed Excel help and I was there. Alex Powers (00:45:17): Oh yeah. Rob Collie (00:45:18): This person, they developed a habit of having me do all of their Excel work for them. This is one of my peers. And then of course passing off the work as their own. Fine, I wasn't that career minded, really. Six months after this is when I volunteered to no longer be a manager. So climbing the corporate ladder wasn't some voracious appetite of mine. So, okay, fine, fine. I knew what was happening, but I was still okay because the Excel problems were so fun. Keep them coming. Then one day this person asked me for Excel help. And there were these two columns of numbers. And this person had subtracted column two from column one to create column three and then added up column three to get the difference. Alex Powers (00:46:04): I'm waiting for the reveal, because there's a big story here and I'm loving it right. Rob Collie (00:46:09): I said, "Well, you know, you could have just summed column one and column two, and then taken the difference between the two sums." And they said, "But wouldn't the answer be different?" There was this moment of silence. I'm looking. I'm looking at them. They're looking at me. I'm looking at them. They're looking at me. At that moment, they realized that they couldn't use me anymore because I was now dangerous. I now knew that they didn't know math. They didn't just not know spreadsheets, they didn't know math. They were exposed. This person is now an executive at Google. Tom LaRock (00:46:48): This person being the executive at Google. I have no doubt probably doesn't know math. However, as somebody who uses technology and knows that data can be dirty and whatnot, I would actually, if it was me Rob, I would say do it both ways and make sure the answers match. Because you know what? We both seen it where it didn't work out. Rob Collie (00:47:10): That's true. But like when you see all the numbers in front of you, you physically see them all. You've got access. There's nothing hidden going on here. Oh, by the way, Tom, what's your degree in again? Tom LaRock (00:47:21): I have a master's in mathematics from Washington State University. Rob Collie (00:47:23): Masters, yep. Yeah, the masters in math is what allows Tom to say, "I'm not sure." Tom LaRock (00:47:29): Now hold on. Hold on. We've seen it. We've seen it. Rob Collie (00:47:35): There's a name for this. It's like the distributive property or associative property or something. There's some property that we learned in middle school. Tom LaRock (00:47:41): See, that's math with paper and pencil. Now we're talking about using Excel for math. So the tool, there could be something like, "Hey wait," and that's why we tell you, "well, just do it both ways." Even Wayne Winston would probably say, "Yeah, well have two columns. They should match. If they don't match..." Rob Collie (00:47:58): No, no he would not, not in this particular case. Tom LaRock (00:48:01): You're right. He wouldn't. Rob Collie (00:48:02): Every time I tell this story, someone always sort of like takes a sympathetic stance towards the antagonist and I end up feeling like a heel. Tom LaRock (00:48:09): You shouldn't. You shouldn't. Rob Collie (00:48:12): But come on. Tom LaRock (00:48:15): I'm with you. I have no doubt that they don't know math because I come across the same people. I do. Rob Collie (00:48:22): It's think it's the intersection of all of those things, That I was being used the whole time. Tom LaRock (00:48:27): Yes. Rob Collie (00:48:29): Which I had kind of made my peace with. But then on top of that, this incredibly aggressive ladder climber, the kind of person who really was kind of like willing to climb over the bodies of their colleagues. There's something delicious about, even though I was the rube in the whole story up until a certain point. I was being taken advantage of and I knew it. But even me in that situation, there was that moment of just like jaw dropping dumbstruck, like just looking at this person going, "Oh my God, you did not do that." Alex Powers (00:49:09): I'm going to lift us up from the depths here of career and everything else. I thought you were going to take us into that they didn't use cell references, which I've seen people type in column A plus column B's value in an equals. And it's like, "Well, why didn't you just do A1 plus B1?" Mind was blown. So I love that those moments still exist and you find them out in the wild every once in a while. And it's not massive warehouse MPP processing, et cetera, et cetera, that everyone's like, "Oh, this is the," I call it the BI bubble. Everyone's out here living in the BI bubble, writing C sharp, doing tabular and coding, blah, blah, blah. People are still excited about the very simple things that technology can achieve for them. Alex Dupler (00:49:56): My in-laws, they own a brewery in Rinton and they make great beer. I offered to help my mother-in-law with some of her bookkeeping that she does on inventory. And she was showing me how she was doing it. And she was like, "Okay, I get these numbers in Excel. And then I get out my calculator." And I was like, "Okay, let me show you how you can do this differently." And I showed her. She was like, "No, no, that's going to be too hard. I'm going to stick with the calculator." And I was like, "Okay, that's fine." Alex Powers (00:50:20): I still get the, "I don't trust Excel, so I double check with the calculator." Rob Collie (00:50:25): My first exposure to that, I was in college. I was working for a construction management firm that was building the new chemistry building on Vanderbilt campus and I was working in the management trailer. I was sort of all purpose ... we called me the lackey. I would just do whatever anybody needed. Sometimes I'd go out in the building and take measurements for things or whatever. But most of the time, I was just doing paperwork and stuff. They turned over the spreadsheet for this latest change order to the project to Vanderbilt management. And the price tag, it was an Excel spreadsheet and it had a column of values that were summed and there was a number at the bottom of it. And I remember the guy Tony who worked for Vanderbilt going, "Well, someone's going to have to double check these numbers. We can't just pay this contract." And my boss was just looking at him going, "Come on. That's what the spreadsheet is for is for doing that." And Tony's like, "Yeah, yeah, yeah, yeah, I know. But still, I mean, we can't just pay this number." I can understand that stance a little better, anyway, I just looked like a giant meany. But remember. This was someone who was taking advantage of me. Alex Powers (00:51:36): I agree. I agree. Alex Dupler (00:51:38): One of the things that I wanted to touch on in this conversation, something you've brought up a lot, which is going from BI to taking action within the report. And I got to tell you, this concept terrifies me. As a BIPM, I'm terrified of it. And I totally agree that the value is there, but in the BI space, we are really bad at testing. And if I think about how going from, Hey, I've got a report and these are the numbers to someone's going to click a button and it's going to change something in a system of record, the level of quality and testing goes up and I think really threatens the quick solution thing that you've also talked about is your bread and butter of like, Hey, we're going to do this really fast and it's going to blow your mind. But if I got to throw all that testing in there to make sure I don't blow up your source system instead, I don't know how those two things coexist. Rob Collie (00:52:39): Yeah. that's a fair point. I mean, for a moment there, when you were saying the taking action part and this terrifies you, before I understood the subtleties of your point, I was going to make the joke like, "Oh, you want this to be like the psychic hotline. It's for entertainment purposes only. Please don't use this report to take any action." Alex Dupler (00:52:57): It does make my job easier, I will admit, but it is a little bit more nuanced than that. Rob Collie (00:53:02): Okay, okay, fine. So anyway, I still managed to sneak the joke in there without ... it's not a joke at your expense because your point is different. Okay, there's escalating versions of this with escalating versions of responsibility and test implications and things like that. So you can just start with report design and working backwards from the types of action, your constituents, the users of your report. In classes what I would teach this concept on the end of the last day, as sort of like a religious sermon. I would encourage people to think of the users of their reports as each one of them sitting in front of like some gigantic cartoonish bat computer looking thing with these giant oversized 1960s, pop art colored buttons and they're labeled things like "open more stores" or "adjust hours of locations" or "increase head count, reduce head count" or "change product mix" or whatever. It's actually kind of interesting, when you imagine e ... Rob Collie (00:54:03): ... mixed or whatever, right? It's actually interesting when you imagine it that way to give it that physical manifestation, it actually becomes a little bit easier, for me anyway, to imagine what these people can do, because every role in a company really has a finite number of actions that they can take. Now, finite in terms of their categories of actions. It's certainly infinite when you get into the details of what are you going to do. And if you start to think of them from that perspective and you think, okay, what I should do is build reports that advise them or at least are helping inform them as to which control they should touch on their dashboard and directionally which way they should move it. Rob Collie (00:54:45): And it sounds like not that important of a trick, not that powerful of a trick, but if you actually apply that methodology faithfully, you end up with a vastly different portfolio of reports that you have built. Even I, very often, don't live up to my own principle in this regard. Because it's so easy. It's so seductively easy. It's the path of least resistance to grab all the data, load it up, make the model, that's fun, and then it's like flowing downhill. It's just like, oh, this is the easy and fun part, right? And then inevitably, you just start slapping together some reports. And those reports, in some ways, are just exposing the coolness of what you've built. Rob Collie (00:55:29): Now, that still leads to some very, very useful things. That's mind-blowingly better than what you ended up with in the old dark ages of Excel or even traditional BI. But I mean, oh my God, we were just talking about it on the last podcast. Some of the things that I have seen in the world that were supposed to be helping people make decisions were better described as their opponents in the process. This report was something that you had to fight to figure out what you should do at your dashboard. Rob Collie (00:55:59): So even before we start with any sort of actual software integration and taking action and things like that, that's a really, I think, important religion to develop. And again, when you're at your best, your absolute gold medal in the Olympics celebrated by the world best, maybe 30% of your output will live up to this. You just can't, you can't execute that way all the time. It's really, really hard. But it's software development. You are building software when you're building reports. You should have the same sort of mindset, if you can, as the Power BI team has when they sit down to design a new feature in their software. Alex Dupler (00:56:38): I totally agree. One of the questions I've been asking a lot, because I've been working on reports for the salespeople to take to the customers is, what is the conversation you're going to have with the customer? Not, what is the metric, but how does this fit into the conversation? And part of this is because my superpower and my career is going and building tools for the thing I used to do. And I think a lot of BI people come from that, where they were in the business and they were doing a thing, they just started making the reports for that thing. And somewhere along the line, they either work away from it for too long or they solve those problems. They had to learn how to make reports for something they haven't done for years. And I think that's a difficult transition and one I've been going through. Alex Dupler (00:57:26): But yeah, learning how to ask questions of the user, because they're not just going to tell you... what they tell you isn't what they need. You have to learn how to learn from what they say, what they actually want. Rob Collie (00:57:42): Yeah, it's a fine art. And by the way, when you've been in "BI", building the same reports for a long time, generally speaking, looking backwards anyway, those reports also sucked because they were constrained by what was possible at the time. And so they were never very ambitious. And most of those reports amounted to... A lot of times they just amounted to the data dump import that's used for something else. It's just, again, it's the opponent. It's better than nothing, but it's meager, meager help. And suddenly you're given this brand new tool set that's capable of so much more. Rob Collie (00:58:22): And unfortunately what I see a lot of times, when you give Power BI to an IT department, they go, "Oh hot damn, the new SSRS." This is the new reporting services. We're going to use it like reporting services. Load that big one flat wide table and pigeonhole it as visualization. It's just like, "Come on." Alex Powers (00:58:45): I'm telling you my favorite DAX is always written from the IT department. It's just written like a massive sequel statement, 400 lines. None of it makes any sense. It's like, "Can we just calculate, maybe another table here or there." Alex Dupler (00:58:59): The folks coming from the IT department, the one thing they do have going for them is that they did learn to format their code. Sometimes people coming from the Excel world, they learned that they can't format their code. And so I'm not sure that I would agree that the worst DAX comes from the IT department. Because you take a DAX statement and you take all the formatting out, and you've just made it 10 times worse. Alex Powers (00:59:21): From readability, yeah, I would agree. Rob Collie (00:59:24): I gave a talk one time where I asked the trick question of, what's the number one programming language in the world? And the punchline is it's Excel formulas. In terms of usage, it's overwhelmingly Excel formulas. And then I showed them, the audience, just to underline it, I take a chunk of Java or something off the internet. I put it on the screen. And I take an Excel formula, a really complicated Excel formula, and do the same. Put it next to it. But when I put that Excel formula on the screen, it's formatted, it's indented, it's got line breaks and everything. So that it looks, and it does, it looks ... suddenly, it looks a lot more like the Java above it. And then I go, but of course, the Excel programmers, they like it the hard way. And so then I clicked next on the slide deck and it all squishes down into one paragraph. It was randomly line broken. Alex Dupler (01:00:11): So at some point, Excel started supporting white space and formulas. And I don't know when it was, but it used to be, if you did that, you'd just break the formula. At some point in the last five years, you can add line breaks and indentations to your Excel formulas. Now, the code editor makes that very difficult, but at least the formula will still work. Alex Powers (01:00:34): I was doing this in 2013 so it's been around for a while. Rob Collie (01:00:36): Why don't these tools make this automatic? Alex Powers (01:00:38): I would agree with that. Why doesn't... Whenever you hit that checkbox and Power BI, why doesn't it just format it right then and there? Alex Dupler (01:00:44): I don't know a power BI product team person. Alex Powers (01:00:50): Well, I'll tell you what: ideas.powerBI.com. Rob Collie (01:00:56): So there was something you guys were talking about earlier that I wanted to come back to. Having both come from non-traditional backgrounds, by the way, that's the majority of backgrounds. Alex Dupler (01:01:04): That's the traditional background for BI. Rob Collie (01:01:06): That's exactly. Oh, that's right. Non-traditional is the new traditional. So do either of you ever find yourself wrestling with some version of imposter syndrome because of that lack of a credentialed background. Does that ever haunt you at all or "Nope, not bothered"? Alex Dupler (01:01:22): I am optimistic as a fault. So I got to Microsoft and I was like, "Okay, great. I'm going to find the secret Excel people in the internal stuff." Similar thing with Power BI. I'm like, "Okay, great. I'm at Microsoft, I'm learning Power BI. There's got to be people that are better than me." And it's been almost five years that I've been here, one color badge or another. And except for the folks on the CAT team, I really haven't met that many people that are better at it than me. Now, on the other hand, when you look at Alex's team and you're like, "Okay, well, I've been reading his blog for years. I've been reading his blog for years. I've subscribed to his YouTube channel" All of a sudden that job intimidates me. And similar to Alex, I wouldn't mind working there, but man, that's quite the hall of heroes. I definitely feel some imposter syndrome when I'm like, "Okay, well I think I have some of the experience necessary to do this, but I am not Chris Webb." Alex Powers (01:02:25): So I'll speak from that end words like, "Hey, I'm surrounded by the experts." Within the industry, I'll say that because they still have their own challenges. They still have their own day to day where there is something simple that blows their mind once again. It's like, well, I didn't even know that that would have been there for the past 14 months. I come from an Office background. This is part of my journey. And so I learned Excel. I learned all the modern Excel capabilities. Learned Power BI, I'm learning Azure. I'm within that learn at all journey and I am not the expert. My job is to find expert answers as best as I possibly can. So this is advice for everyone and I am the best at what I am the best at today. Alex Powers (01:03:06): Those around me, who I can help, I absolutely love helping. So Rob, when you were talking about, "Hey, someone reaches out with a small question like you do want to jump in and you do want to help. Obviously, it comes back to a point of scalability like, well, can I help everyone in the world? No. I just have to find new ways to spread my message or get my voice out there. There are so many things I still want to learn tomorrow. There are so many things where I'm just pacing myself. I'm not going to go become the expert at C# to write DAX. Power BI should just be simpler. It should have an advanced ribbon. It should have all these capabilities built in like the ... Alex Powers (01:03:36): I'm really excited like the DAX auto generator coming out in the future here. I want to say February next year. I don't think it'll be that great on day one. Because obviously it depends upon an incredible model to get incredible system written DAX. But, my God, everything should be made simpler with technology. If we require experts to use this stuff, I'm sorry, it's just not for me. It's not where I want to be. I love the low code side of things. Alex Dupler (01:04:01): It doesn't have to be good. It just has to be the macro recorder. Coming back to what we were talking about before. Rob Collie (01:04:07): Like LARPing, acting it out, pantomime, HoloLens. That's where the industry's headed. Alex Powers (01:04:11): I love, Rob, that you also said, "Hey, don't write any M." I would agree. I want people to do as much as humanly possible with the lowest barrier for entry. Go, Thomas. Tom LaRock (01:04:23): So Alex P, when you just said that about being the best person, that's something I've talked about before as well. No matter what you have me here for and if I know what I'm doing or not, whatever it is that you need help with, I'm willing to help. And the fact is I'm the best person for what we're doing right now until the best person shows up. And that's the attitude, I think, any data professional has to have. Even with Rob, and the guy who couldn't do the math, he's the best person at Google right now until the best person shows up. Rob Collie (01:05:03): They're the best at getting other people to secretly do their bidding. Tom LaRock (01:05:08): Until somebody else shows up that does it better. Rob Collie (01:05:13): I like to think that I educated that person that day though. Tom LaRock (01:05:17): Oh, you sure did. Rob Collie (01:05:18): They learned something. Their Grinch brain grew two sizes that day. Tom LaRock (01:05:23): They learned they needed to put some distance between you and them. Rob Collie (01:05:26): That's right. They knew that they had to give me a wide berth from that point forward. Anyway. And they probably then also proactively went about character assassinating me behind the scenes in case I leaked that knowledge. Knowing this person, that probably happened. Alex Dupler (01:05:41): They didn't go Google order of operations and learned how to do a little bit of math. Rob Collie (01:05:47): How would they even know to Google for that? Like order of operations? Rob Collie (01:05:54): You said Alex, you're not the expert. Oh my gosh. You are an expert. You're not the expert. Of course. Okay. Like anybody that wants to call themselves "the expert" earns every bit of ridicule that's about to come their way. Right? And every bit of comeuppance that's about to come their way. Right? Every bit of it. Oh, eat it up. Right? You asked for this. So I agree. You're not the expert, but oh my gosh. Are you an expert? When we had Hugh Millen on the podcast, the ex NFL quarterback, he pointed out sometimes... I really like... Let's pick any activity that humans engage in. And let's assume that you could know the rank order of how good people are, which of course is itself a fallacy. Right? But there's one person in the world who is the best. Everyone else is in the middle somewhere. Rob Collie (01:06:42): And there's also someone that's the worst, but like the overwhelming majority of humanity is in the middle somewhere. So like just the fact that you can look up the ladder and see people that you think... And oftentimes people that you look up the ladder, you think they're up the ladder from you. They're not, they're not, when you really get up close and personal or more realistically they're up the ladder in some ways, it's not one ladder, it's actually a million parallel ladders and you've all got a coordinate on each axis. This is why I wanted to talk about the imposter syndrome thing, is that everyone that's listening is somewhere in the middle. We're all also somewhere in the middle and relative to where we all were, including me, when all we had was Excel like V-Lookup and single table pivots, Alex Powers (01:07:27): I just want to go on the record, I'm an index match guy, just so we can kind of clear the air. Rob Collie (01:07:32): This is another joke that I used to tell in my public classes. I raised my hand and say "okay"... so if you look up, right, right, right. But there's a few of you who are index matchers, right? Like 1 out of 7 people put their hand up. Right. And I go, "and everybody you work with knows you're an index matcher... You're really vocal. You're in everybody's little face about you being an index matcher, aren't you?" And their colleagues are sitting next to them laughing like "Yes!". Alex Dupler (01:07:58): There's one person who thinks it's actually supposed to be index match match. Alex Dupler (01:08:04): And then there's me on the other side, which is like, why are you using V-lookups? Have you used [inaudible 01:08:08] ? Rob Collie (01:08:13): Yes. But I have yet to meet the index matcher that isn't proclaiming to the world. Tom LaRock (01:08:19): Is that like the CrossFit of Excel? Tom LaRock (01:08:22): Did I tell you I do CrossFit? I wonder what the Venn diagram, the overlap of CrossFit and index matchers. I wonder... Rob Collie (01:08:35): You'd have to control for the Excel audience first. I'm not sure that in general, the index match crowd overlaps a lot with the CrossFit crowd. And where does Camaro ownership fit on this Venn diagram? Alex Powers (01:08:56): I've seen some stuff like, these are some old war stories you guys get ready to tell Rob Collie (01:09:02): And you know that. It's not like this is the first time you've come to terms with the fact that you have powers. Oh my God, I didn't even mean to do that. Your last name is Powers. I just said you had powers! Alex Powers (01:09:14): I think a lot of it comes back to... Just those of us who have been around for a while, those of us who are reading the books, you can't see it on the stream here, but Alex D. has got all of his books on the shelf. I have the exact same stuff in the background. I've got books on my desk. It's just like, we're always constantly learning. We're not stopping. I think when you look around, you're just like, there's so many people out in the community doing weird wild things. I would rather read their article, listen to their podcasts, watch their video. I'm still in my little nerd corner here with the things that still excite me. I like how you alluded to at the beginning, Rob, where yeah, that looks cool. That looks fun. It's shiny. It's awesome. But this is what excites me. That's what I still just love day in, day out. Alex Dupler (01:09:55): I'll also add that when Powers moved over to the Power BI CAT team, I saw that job opening and I thought about applying. And when I heard he was going to apply, I didn't even bother. Rob Collie (01:10:08): Well, you just mentioned earlier that this is Thunderdome. 2 power BI professionals enter, 1 power BI professional leaves. Alex Powers (01:10:15): I had the commitment to change my last name. So, for me, it was kind of a slam dunk. Rob Collie (01:10:20): You were talking about books, I think you're the only book review on Amazon that mentioned the intermission. Alex Powers (01:10:26): Yeah. Loved it. Rob Collie (01:10:27): That's why I remembered that review. I had no recollection of who had written it. Tom LaRock (01:10:32): I forgot the intermission. Rob Collie (01:10:33): When I saw the text of the review, I'm like, oh, that one, that one. Yes. Like halfway through the book, I put a one-page intermission in that said, "Hey, you can just put this book down, go do amazing things. You're already amazing. If you understand what's here, just know that what's ahead of you is also awesome, but don't feel like you have to judge yourself by how well you recall and master the things that follow this page. It's like this explicit reminder of always being in the middle somewhere, I noticed this in classes as well. People would, if I teach a two day class or whatever, people would tend to judge how well they understood it by the list of things that they didn't understand as opposed to, by the list of things that they did. And so, I also put this concept of intermission into my teaching, where I would be like building up this list of cool new capabilities that we're learning relative to regular Excel. Rob Collie (01:11:31): And then eventually near the end of day one, I draw a line under this and say, okay, "Everything above this line. I want you to leave here having understood. So we'll drill this. You'll ask me questions about it, whatever, everything below this line is gravy and more of a tour of what is possible if you remember that there were things like this that were possible below this line. Even if you don't remember exactly how to do them, you're going to be in good shape. But this above line is the foundation". It's basically the same concepts that happen pre the intermission page in that book. Rob Collie (01:12:04): I was really grateful. A couple of things just broke the right way writing that book. One was that I didn't choose to go with a traditional publisher. I decided to publish through Bill Jelen's company. And he asked me if I wanted a traditional editor and I said, nah, let's not do that. And the other one was is that I found that I wasn't able to write in the technical voice. I tried, I tried to emulate, again the imposter syndrome thing. You tried to emulate the stars. You try to emulate like every O'Reilly book you've ever picked up. And I just couldn't, it was exhausting. And so I wrote it in that quirky, like, why not put an intermission page in the middle of the book, right? That was just so I could get through it. Just so I could survive it. All these years later, it stands as one of the things that people really like about the book. And I'm just really grateful that people accepted it. It wasn't some stroke of genius, but I do look back on all that with a smile. Alex Powers (01:13:01): So I love that it is your voice throughout the book. Question for you, because I'm kind of the weird quirky style too, where I'm going to write something that really excites me and it's going to be fun, and I'm going to have a good time when I get to the bottom of this. Start out with a two paragraph thing that turns into like a 17 page diatribe. You're just like, ah, that's not what I wanted to do, but here I am. The style of reading that I enjoy most though, is the most dry technical, boring, where it's a labor, it's a labor to understand it. It's a labor to get to that final page. John Walkenbach was that for me, from the Excel world where you would talk about O'Reilly, what is your personal style of learning? Do you still enjoy like a deep spec page and getting into the nuts and bolts? Talk to me about that. Rob Collie (01:13:44): I'm just kind of a terrible learner, it turns out. And this is why my adoption of something like Power BI, a DAX, a data modeling to me just like so extraordinary. This is something that, I couldn't help it. I also think benefited tremendously from the fact that there weren't a lot of people in the space when I started. There wasn't anyone to judge myself against, really. I was the only person at that point in time, like back in 2010, I was the only person who was living this every single day. That was their only thing they were doing. Even back then, like Chris Webb, like he was still doing a lot of MDX, the Italians, Alberto, Marco, they were still doing a lot of MDX. They had a lot of existing clients and I had a clean slate to do nothing but this, this isn't a strength of mine. This isn't something that I'm proud of. I want to think that I would, but I'm not positive that I would throw myself into this stuff with the same vigor today because I would be judging myself against those who went before me. Whereas before in 2010, I got to treat it like an adventure. Like I was out there discovering the frontier. I was like Lewis and Clark. And I don't consume very much. I'm not an avid consumer of YouTube. I don't really read books. I don't read blogs. I could paint a very negative picture of myself, for the listening audience, if I wanted, I think maybe the positive way to describe it is that I found that I'm at my best like in a phone booth, like very focused. Rob Collie (01:15:12): So I've got to be hands-on, I've got to be struggling against it myself. And then I will go and run the search that's looking for a particular technique, but I don't even like YouTube as how-to, I don't have the patience to watch a video. It seems like 99% of humanity does. I have such bad ADD, like I'm trying to scroll the page, get to the point. Where's the answer. You know, I don't have much success scrolling through YouTube fast-forwarding videos. I always missed the point that I needed. It's like trying to fast forward commercials. I'm always over running. Rob Collie (01:15:45): I don't think anyone should ever try to actively emulate my learning style because it's just not a very good learning style, but because it's such a poor learning style, it made this stuff seem even more special to me because even I, with my awful learning style, devoured this stuff. A true statement about this was that even as I was doing all of this, people were telling me I should write a book. And I was saying, "no, I shouldn't". If I understand this, everyone understands this. Alex Powers (01:16:12): Yeah. That's always tough. You always think that too. We're like, "well, everyone knows this". That is like 1% of the time that that ever is the actual case, is what I'm finding day in and day out. Alex Dupler (01:16:25): We were talking about digging in and helping users. And this has become a really big part of how I run our BI practice, which is that I'm a PM building tools to make analysts successful. The way that I know what tools to build is I spend most of my time helping them be successful, building stuff. I don't have a calendar full of meetings with various stakeholders. I keep my calendar open. And then any time someone on our broader team of analysts has problems, I try and help them. Sometimes it's like, well, you just need to learn this function. And sometimes it's like, if we just added a table to the [inaudible 01:17:08] , your life would be much easier. I'm going to put that in. And, sure, it's going to be 4 months before we get around to it. But the next time that problem comes up, I know exactly that we've covered that scenario. So I feel like that's been a really successful strategy for us. Alex Dupler (01:17:24): Now, the downside of this, and this is something that Seth and Mike were giving me a hard time on the other day on Twitter, is that I don't create a lot of documentation because every one of those conversations is unique. I could write it all down and tell them, "Hey, just go read the docs", but then I don't learn what they're trying to learn. And I really struggle with that trade-off. Rob Collie (01:17:45): At least, you know, there's always a job for you as a Cosmos engineer. Have you thought about getting into software development? I mean like really... Documentation is overrated too, right? For end users, it is. Cause again, the end-users are flawed, right? They're humans, we're not talking about an API here. We're talking about tasks. So if you really wanted to write documentation for your data models and stuff, it's basically a book. It needs to have a narrative. It's pick up a copy of my book and go, "okay, this book is written about a system and it's got a very deliberate reveal order. And it talks about the human problems that need to be solved and why you have to make your peace with this". And you have to help guide them down the path. To be successful documentation for something as complicated as a sophisticated data model, it's more like a book than docs. And then, you know what's going to happen? Two-thirds to three-quarters of people aren't even going to pick up the book. So I'm not giving you a license to never write documents, but I'm giving you a license to not write documents. Alex Dupler (01:18:57): Well, one of the things that I've struggled with, because everybody's like, "oh, do you have a data dictionary?". And I'm like, 90% of the questions that we get, it's about the glue. It doesn't fit in that column, name, description framework. It's like, Hey, how do these things fit together? Why is this missing? That doesn't fit in a data dictionary. The other thing is of course reading the documentation is like a superpower, Alex, you clearly have that in spades, but lots of folks, like, they're not going to go looking for it. Tom LaRock (01:19:27): They're not going to look for it because they're not expecting it that it exists. Alex Dupler (01:19:30): But even the power BI documentation, they're like, "Hey, how do I do this thing?". I was like, "well, did you try looking same period last year, did you try looking for a function to do prior year?" It's like, "no". Oh, well. Rob Collie (01:19:47): And then you searched the Microsoft documentation for the ALL function. You're a brand new user. You're the person who doesn't understand the ALL function. So by definition, you're the person who should not be told "this is the table function". That's a terrible way to introduce it to someone. Okay, it's technically true, and I know that on Reddit, that's the best kind of true, but it's also not helpful. Alex Powers (01:20:12): It's funny. Cause you're talking about docs. Thomas had brought up as well. I'm imagining someone going out to Microsoft docs and typing 'Power BI ALL', how successful is that search result going to be? And then it comes back to people like Rob coming up with these weird wacky names, a data flow, a this a that, how are people going to learn these? Is it with a space? Is it not a space? Terminology is hard, just in the product. And then like you had said, Alex, where would they even go with his basic knowledge to get a correct answer? Rob Collie (01:20:44): Let's also not focus on the downside. We're talking about the no docs, we're having a discussion about whether that's okay. Let's just make our peace with it. It's kind of okay. Kind of not okay. But, look at the upside of what you've been doing. You've been having all of these dynamic interactive conversations with all of these users. Rob Collie (01:21:02): Interactive conversations with all of these users and that advances the art. That advances things so every one of these people is capable of engaging with and wants to engage with a conversation. And when they're not understanding it, most people will admit it. Those who don't, you can usually read it on their face. You get that two way conversation. The opposite approach would be like, oh, write the docs and I never need to talk to anybody, right? Okay. If you had to pick one extreme or the other for something like this, and we're not talking about an API, right? It'd be different if it were an API, right? Rob Collie (01:21:33): And even APIs with their documentation still require books and YouTube videos and training and all that kind of stuff. The amount of sort of human good that you create with this interactive model, this open door policy and helping people directly. I don't think we should get like target fixation on the downsides of this. The upsides of this are enormous and it's got to be super, super gratifying for you. Alex Dupler (01:22:00): Yeah. I have a lot of fun. Rob Collie (01:22:00): It makes you want to do your job as opposed to hate your job. On another podcast that we haven't released yet, it's coming up, we talk a little bit about how it's a better strategy to lean into one's own strengths than to try to mitigate one's own weaknesses. You don't have an excuse. It's not like you get to ignore your weaknesses, right? But you can't get through life mitigating weakness. Alex Dupler (01:22:27): Yeah. Yeah. You got to strive for being T-shaped. It's sort of what you were saying before, which is like, "Hey, there's a list of things you got to know, and then you need to know what's possible." If you come into a circumstance where you need to write some custom M, go look it up. But you got to know that, "Hey, there's a script back there. And if I get into real trouble, maybe there's a solution." Rob Collie (01:22:50): Conversation beats broadcast every single time and docs are broadcast. So ideally you have both. If you have to choose though, conversation wins seven days a week, twice on Sunday. Alex Dupler (01:23:02): The other story that I wanted to tell, which ties right into this, which is that when you're doing BI, it's really important to know the business. So I grew up in the Pacific Northwest. My dad was Boeing IT for many, many years. He wasn't in the BI space. He was in networking and communications. So in the eighties he was working on a DOD contract. The admiral he was working with told them he was the first admiral in the whole Navy that could email with his project manager back and forth because they were the first ones to get email that worked across the networks. And then later he was working in the Bellevue campus, so right next to where Advanta is, where part of the Power BI team is. Alex Dupler (01:23:45): His office was in the Darth Vader building right across from there. And I remember going to his office, this must've been in '93, '94, and under his desk he had a server. And he said, "Oh, that's got our pilot for Exchange." It was the first beta version of Exchange. And he was running, self-hosting it in his office for Boeing. So they were one of the first customers getting in there. But he used to rail against that Bellevue campus, partially because the traffic sucked but mostly because they had IT in Bellevue and they don't make any planes in Bellevue. And he always thought that that was one of their biggest mistakes was not having the IT folks where the business was. And I think that directly translates to BI. Rob Collie (01:24:34): I want to go back to '93 or '94 and tell your dad, "Oh, you think traffic is bad?" Alex Powers (01:24:44): It was. Rob Collie (01:24:44): Just wait. Alex Dupler (01:24:46): We lifted in View Ridge and sort of north Seattle, and so to get to Bellevue, he'd have to do one of the two bridges. And yeah, obviously it's so much worse now. Rob Collie (01:24:58): I mean Microsoft still had like 10,000 employees at that point in time. When I got there in '96, they had 17,000 in the Puget Sound, right? It's like, we're just getting started. This virus is really just come to shore. Alex Dupler (01:25:11): Jeff Bezos hadn't even moved to town yet. Expedia was still a twinkle in someone's eye. Rob Collie (01:25:17): This is a very powerful concept. It's super important, right? This goes hand in hand with that other question about non-traditional backgrounds. Alex Dupler (01:25:24): Yeah. Rob Collie (01:25:24): Being the tweener, having the business domain in your head and the technical skills to execute is everything. My old joke, and it's not even a joke, I think it's the truth, that greater than 99% of a traditional BI project was just lost on communication of requirements and clarification and mistransmission and misunderstanding and iterating needlessly because human beings don't merge brains. If we had Vulcans with mind-meld, we'd be different, right, but we don't have that. We have a very narrow bandwidth between us, unfortunately. Alex Dupler (01:25:58): If anything, it's worse now because the tools have gotten so much more powerful, but we still have the same communication challenges that we always did. Rob Collie (01:26:07): Yeah, but I think you're closer though. You're closer, right? You understand the business. Alex Dupler (01:26:11): Yes. Some of the self-serve revolution has also made more business people do the work, but in general it used to be that you'd spend all this time getting the requirements right, and then you'd have to go write code for a month. And now you spend all this time getting requirements right, and then you would drag and drop a few fields. And so it takes so much less time to build the thing that the communication is an even bigger percentage of the job. Rob Collie (01:26:37): Right. Alex Dupler (01:26:37): And yes, you can get a nice feedback cycle going, but that still means that it's the communication that's the job. Rob Collie (01:26:45): Yeah. But you're not doing it that way, right? You've already described how you work. You're not going out, sitting down with them and saying, "Hey, write me a doc. I'll be back next week." Alex Dupler (01:26:54): No. Rob Collie (01:26:55): You're building in real time in front of them and saying, "Oh, do you mean like this?" Right? Picture worth a thousand words and all of that. You're not writing documentation on the data model. You're also definitely not writing requirements documents. Alex Dupler (01:27:07): No. Rob Collie (01:27:09): So the proof is in the pudding. Tom LaRock (01:27:10): So what you're missing is a layer, an organizational layer, of what I would call is a functional analyst. And it doesn't exist for the reasons you've stated. The tools are easier to use, it's self service, people can do these things. So you've lost that bit of communication because traditionally, I would say you would have people there who were the functional analyst that would bridge the gap and with them gone, this is the pain that you feel. You are essentially the functional analyst. You just don't have that title. Alex Dupler (01:27:40): No, we do have analysts. That was when I first went from Fender to Blue Badge, I was an analyst focused on a particular sales group. And I did quota setting and some performance reporting, and now I'm a PM for the tools for that group of analysts. And so I don't actually build any reports, zero visual content, only database layers. Tom LaRock (01:28:06): No, I was talking more about the business knowledge that you had touched upon. The person that knows the business and knows the technology enough and can speak to both groups and say, "I know what's possible. I know what you need, and I know what's possible and I know how to get it done. And I can even write something for you so that you'll understand what we're doing and why." And I think that layer has disappeared over a couple of decades as tools make it easier for somebody to just sit down and build shit for themselves. Rob Collie (01:28:35): Let's paint that in, I think, a more positive light. That role, this sort of middle layer, was always inefficient, and that role has just migrated into someone that's embedded in the business. Tom LaRock (01:28:49): They have a different title. They do something else. Rob Collie (01:28:53): Yeah. They're the ambassador, right? Tom LaRock (01:28:56): But there's still a communication gap because there's nobody above them saying, "Hey, these are the three skills that we need to bridge the gap." We need to have people doing that. I don't care what they're called and that doesn't exist, and therefore we still have this gap and the gap is widening. Rob Collie (01:29:15): I disagree. I think the communication gap is dramatically smaller and it's shrinking every day as we get more and more people skilled up on this stuff within the business. I don't think we've lost anything. And when was the last time that someone up the chain from you actually knew more about it than... That doesn't usually happen, right? Tom LaRock (01:29:34): I didn't mean to imply that somebody up the chain knew more. It was somebody up the chain understood the value of the role and said, "I got to make sure I have a team of these people." Rob Collie (01:29:42): Oh, right, right, right. We talked about a little bit earlier, right? In essence, I wasn't properly valued where I was for a number of different reasons, and so I went looking for a place where I was properly valued. Well, this is the story of where P3 employees come from. Tom LaRock (01:29:59): The downtrodden. Rob Collie (01:30:01): I mean, so it's kind of good news, bad news, right? It's good news for P3. The world keeps manufacturing underappreciated, awesome people, right? Bad news, it's like, oh, that's really kind of human sad. I mean, it's important to our business, but it's human sad. Tom LaRock (01:30:18): It's important for [inaudible 01:30:20]. Alex Dupler (01:30:24): Well also I think it's amazing how much, when you learn these tools, how quickly you can go from contributing at one level to contributing at a whole nother level. I'm not really a big fan of the 10X engineer thing, but I definitely think from before I learned Power Query to after I learned Power BI, I can do the same work in a 10th of the time. Now my ability to focus over the course of a year, I'm not 10 times more productive, but in short bursts, definitely that was a 10X. The people that learn those skills, the value that they can create in a single business in that single role, I think can quickly outstrip the sort of scope and business model associated with that role. That was part of why I left the lab. It was like, well, listen, this business model only supports a certain salary and that's not going to cut it so. Rob Collie (01:31:22): There's a quote, it's probably falsely attributed to but it's attributed to Stalin. The Germans were building these incredibly high quality machines, it was like the Mercedes of tanks. And the Soviets were just cranking out just tons and tons and tons of these T-34s that were meant to run for a week before or whatever. And he said, apparently, again who knows, he apparently said, "Quantity has a quality all its own." And they won by the way. You're talking about the 10X thing, right? Alex Dupler (01:31:58): Yeah. Rob Collie (01:31:58): I want to come back to that and say that it might be that 10X is real, even when you account for friction losses and attention span and things like that. Okay. There's a point at which something is so much faster that it's not that you get to the same place 10 times faster, you get to a place you never, ever would have had the ambition to go. Alex Dupler (01:32:21): Yeah. Rob Collie (01:32:22): And in terms of your impact on the business, I believe this stuff is oftentimes more than 10 times faster, honestly. Let's accept it as 10 for the moment. That's ambitious enough. I think the impact on the business, the value of a Power BI empowered human being that's embedded in the business and understands the business domain of the department or whatever is greater than 10 times what the original value of an Excel powered version of that same person. Even if they're not, again, due to friction losses, they're not 10 times as productive in terms of whatever, right? Rob Collie (01:32:56): The quality that comes out of that quantity is different and is far, far, far superior. It reflects a question formulation muscle that is rewarded by being exercised. You ask far better questions than you ever would have dared to formulate before because it would just been discouraging and depressing to formulate those questions that you knew you'd never be able to answer. Now it's practically a lay-up. It's an addictive positive cycle. So I'm going to take your less than 10X and go I disagree. I think it's more than 10X. Alex Dupler (01:33:36): Well, I definitely agree that there's a quantity piece, which is, I think about before scheduled refresh, how many reports could one analyst support? It's like, well, okay, at some point all they do is refresh the same reports. Rob Collie (01:33:49): That's right. Alex Dupler (01:33:50): And sure, there's a point with automatic refresh of reports where all you do is serve as tickets on those reports, but that number is way higher. Rob Collie (01:33:59): Yeah. And let's be clear. It's not just scheduled refresh. It's the fact that everything in a Power BI model and report is built from the beginning to be one click refresh all, right? That's not how Excel reports work. There's no one-click anywhere right? But then you can schedule it to click that button itself. Oh wow, that's cool. There's a certain point at which an Excel analyst is now sunk by the weight of their own spreadsheet portfolio. Just carrying it around with them and turning the crank on it, they're out of time in the week. And so the joke there is, if you're good at Excel, if you're an Excel analyst and you're good at it, what's the worst thing you can ever do for yourself. And everyone always says, "You can tell people you're good at Excel." And I say, "Yeah, but that ship sailed." The worst thing you can do is to build a spreadsheet that people find really valuable. Your success is going to become your luggage. It's become your ball and chain, right? And so you're actually disincentivized from ever creating another spreadsheet because you know better. Alex Powers (01:34:56): I feel like you're speaking to my soul right now, Rob. You're just, I've been here before, but I work in a spec factory too where I wasn't part of the business and it was, "Here are docs, here are specs, just go build it. We don't care." You have no context, and then make sure that you hit that home run with whatever report you're building. You get very good being in those high pressure situations, but I didn't have that connection to the business. I understand that it is valuable, but I don't always think that it is the most important thing in the world if you have someone with the expertise to come in for those last mile efforts because at some point those business analysts, they'll kind of hit that brick wall of capabilities and like, "Oh, we want to scale this. We want to get this to the next level." I think you still need those other people too, accurate or not. And that gap will be closed over time, but that's what I'm still seeing, at least from where we're at today. Rob Collie (01:35:48): We still need people. Alex Dupler (01:35:50): Yeah. Tom LaRock (01:35:50): For sure. Rob Collie (01:35:51): I like that. I wake up every day and check has the general AI come yet? Nope. Okay. So we're still needed. Alex Dupler (01:36:00): So Rob, if you've got some more time, there's one other area I wanted to ask you about, and this is a callback to one of your first episodes, which I listened to not that long ago, but I think it was Brad and Kai, I think were their names. Rob Collie (01:36:12): Yeah. Alex Dupler (01:36:12): But one of the things you talked about was working on building a report package around looking at advertising performance, and one of the project I'm working on right now, traditionally, I focused on internal business performance. What was the change in revenue? But now I'm starting to work on how can we use a template app type experience for our salespeople so they can build on demand, a Power BI app for one customer so they can have conversations with the customer? And I am really curious one, how that venture has gone, two, what are the things that are most important in that session or that project? Rob Collie (01:36:52): The problem with that ambitious project is that we said as a first step, we're going to get really, really, really solidified for our own stuff. We're going to get our own P3 advertising optimized. That kind of put an impossible ever-moving goalpost between us and that project. Some other things changed too, but we're still working with those folks. So we haven't really gotten to that. I'll be perfectly honest. It hasn't even really started. I think the biggest challenge with any effort like that, were we to go down that route, is that I don't think you can templatize it because so much of the important data is going to be on that customer's side of the fence, right? The actual performance of, for example, for us, the actual performance of this advertising is dependent on what happens to the lead or the customer after they start to work with us. Alex Dupler (01:37:46): Yeah. Rob Collie (01:37:46): And only we know that. Our advertising data model unsurprisingly spices together advertiser data, like from AdWords, with our own CRM and accounting data to calculate how profitable our... We're not running web commerce, right? Alex Dupler (01:38:05): Right. Rob Collie (01:38:06): We're not running e-commerce. It's not like we're selling widgets that the customer buys it and then we say, "Oh, well they bought a $500 product and we paid $300 for..." or whatever, right? It's not like that. It's a long running thing. But even if we were doing in e-commerce, even then, right, there's a long-term value of this customer. Maybe it's profitable to only break even on their first purchase. It's not a last mile, it's like a last marathon. And of course, Power BI is really good at this. Splicing silos, that's a Power BI strength for the ages, but you have to do it. I mean, even in our stuff, there's some very complicated SQL that sort of calculates in real time, our best guess as to what these leads will be worth because their long-term value is never known for sure. Alex Dupler (01:38:52): Right. Well, and if you want to optimize a campaign, you can't wait three months to get a signal. Rob Collie (01:38:56): That's right. Yeah. That's exactly right. That's the old knock on BI as a rear view mirror, right? If you only get the information when it's too late to act on it, you're never able to steer. So yeah, every time that that data model refreshes, it's making a statistically historically informed set of estimates as to how those campaigns are performing. Alex Dupler (01:39:17): One of the ideas that we've been tossing around a few years back, there was this Azure solution templates thing where you could click a few buttons and spin up an ADF and a SQL DB and a Power BI report, all sort of in one thing. And so one thing that we could do, because we're designing this initial version based on our internal DBs, and we can't expose that. Also lets us expose some data that we can't put in a public API. A lot of customers want to know what their market share is or their share of voice. And we're not going to tell you the denominator for the total marketplace. That's just not going to happen. We're happy to tell you, "Oh yeah, you're at 10%. And if you bid a little more, you could pass your rival there at 12." We love that sales pitch. Alex Dupler (01:40:04): But one of the things we've played with is we've got this API that you can get data out and it's not a REST API. It's a SOAP API because it's Microsoft, which is one reason why I don't think we can build a custom connector to put on it. I talked to Matt Masson about it and he was like, "I remember that. That's a SOAP API. You don't want to do that." And if we do leave, they said Matt Masson is one of the engineers on Power Query, one of the most senior Power Query engineers. Alex Dupler (01:40:31): So what we were thinking about was, okay, we've got these Python scripts that our support folks wrote to iterate through the API so that they could get data out in bulk. Could we make a version of this that runs in an Azure function or ADF and dumps into a SQL DB that the customer owns and then give them some out-of-the-box reports? Hell, maybe we can integrate Google and Facebook data too because we won't forget to put our data on the first page whereas if you were designing a rational product, you wouldn't necessarily put somebody who's got... Well, we've got 30% share on Desktop, but Desktop isn't the whole ballgame. Rob Collie (01:41:10): It's purely alphabetical. Alex Dupler (01:41:11): Yeah. That's right. Rob Collie (01:41:13): Bing, Facebook, Google. It's defensible. Alex Dupler (01:41:15): Yeah. Yeah. So in any of that, we have thought about maybe we should build some reporting, not to be biased just so we can make sure that we're in the conversation. But automating these APIs and putting them in Azure partially so that we can say, "Oh yeah, 50 bucks a month of Azure, and you'll get all your data in this beautiful format. It's not just API calls. It's a tabular model and here's Power BI. And for another $10 a month, you can have this thing going and share it with everybody. And do you need to build some custom reports and integrate some stuff? Sure. But you're halfway there." Rob Collie (01:41:47): That architecture stuff, I'm the wrong person to ask about. No, but if you would like to hire P3 for a strategic... So we have people for that, there's an app for that. There's definitely people for that. For our stuff we actually use a third party service called Stitch that extracts data from the AdWords API and dumps it into, guess what, Azure data warehouse. For a while there we were advertising on Bing, but we stopped because the Stitch integration with Bing didn't give us the click IDs. Alex Dupler (01:42:17): Yeah. I've heard that before. Rob Collie (01:42:18): So we couldn't track anything with any sort of... And again, this is an instance where, because we couldn't evaluate its effectiveness, we actually stopped doing something. It's not a Bing problem I don't think, it's a Stitch problem. Alex Dupler (01:42:30): No, it's definitely a Bing problem. Rob Collie (01:42:32): Oh, fantastic. So Microsoft should fix that and then Stitch will do it and we'll go back to advertising. But there's an argument to make, right? There are people who are not advertising on that platform simply because they can't report on it. Alex Dupler (01:42:43): Yeah. And we've got some products that would be very attractive to you. We have some LinkedIn-based targeting that for B2B sales make a lot more sense than anything you can get from Facebook. Rob Collie (01:42:54): Oh, we are about to get into some LinkedIn. Don't you worry. We're coming to that. Alex Dupler (01:42:58): So yeah, totally makes sense. And also the other thing that's really big for us, if we go back to when we were talking about the laps, where the business model was commoditized. Well, I don't know if you read Stratechery, but Bing where simultaneously as super aggregator and we have been modularized by Google. You can import your ads from Google into Bing because, to get Mindshare, we need it to be as easy as possible. And so having the reporting be compatible is another part of that. Yeah, people aren't going to just like, " Oh yes. I could get 10% more impressions if I went on to Bing, but if it costs me 30% more time, that's not a good trade, even if the unit economics work." And usually the unit economics are good. Rob Collie (01:43:42): We actually found Bing to be very profitable for us to the extent that we could measure it, but we couldn't get granular with it because of the lack of click IDs. And so we just backed off. Well, fellas, I am so glad that we got this together. I know that we talked about it a little bit in the past and we kind of danced around the idea for a little while and then we did the scheduling dance and there's multiple series of dances, but then we finally danced. Tom LaRock (01:44:05): We danced. Rob Collie (01:44:06): In Thunderdome. Alex Powers (01:44:11): Alexi, it was wonderful to meet you both. Alex Dupler (01:44:14): Thank you. Rob Collie (01:44:14): Two Alex. Tom LaRock (01:44:15): Two Alex, one report. Announcer (01:44:17): 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!

Visit the podcast's native language site