#244: Data Is Everywhere. Why Do We Limit Ourselves by Default?

In order to produce a stellar analysis, have you ever requested a team to teardown a Tesla and count every last washer and battery cell? No? Well our guest this week, Jason DeRise, joined Tim, Julie, and Val to share that story and others on how alternative data can be used to enrich analyses. Luckily you don’t have to have a Wall Street-sized budget in order to tap into the power of alternative data. Looking just outside your tried and true data sets and methodologies to see how you might be able to add to your mosaic of understanding a business question can be powerful! In this episode we talk about some of the considerations and approaches when you put down that hammer and see the world around you is more than just a bunch of nails.

Links to Resources Mentioned in the Show

Photo by Elena Mozhvilo on Unsplash

Episode Transcript


0:00:05.8 Announcer: Welcome to the Analytics Power Hour. Analytics topics covered conversationally and sometimes with explicit language.

0:00:13.0 Tim Wilson: Hi everyone. Welcome to the Analytics Power Hour. This is episode number 244 and I’m Tim Wilson from facts & feelings. You know the old saying, if the only tool you have is a hammer, it’s tempting to treat everything as if it were a nail. It was actually likely coined or at least written down by none other than Abraham Maslow back in the early 1960s. This actually makes me wonder about the physical construction of Maslow’s hierarchy of needs and whether it was maybe a pyramid assembled with nails, but that’s actually neither here nor there.

0:00:46.7 TW: The adage absolutely holds up in the world of data and analytics. When we live and breathe our digital analytics platform or our voice of customer tool or our AB testing system or even our BI platform that brings together data from a half dozen or a dozen or 50 internal data sources. Well, that familiarity can breed if not contempt, then tunnel vision. And that’s the topic of today’s episode. Why and how and when should we be considering data from sources that are maybe off the beaten path? For that discussion, I’m joined by a couple of co-hosts, Julie Hoyer you’re with further. Have you ever used a hammer to conduct an analysis.

0:01:25.7 Val Kroll: Hypothetically or like an actual hammer? I hope not hypothetically. Unless it was on a…

0:01:31.8 TW: On a stakeholder, I don’t know. On your machine, on your computer.


0:01:36.2 TW: On a crash star session.

0:01:38.2 VK: Actual hammer? No. Never used them.

0:01:39.5 TW: Okay.


0:01:40.6 VK: For an analysis.

0:01:41.6 TW: And Val Kroll. You’re my, once again, colleague at Facts & Feelings.

0:01:45.6 Jason DeRise: Woo-Hoo.

0:01:46.1 TW: Have you ever felt like all the world is a nail?

0:01:49.2 JD: I definitely have seen that out there. I would like to think though that I haven’t fallen into that trap, but hopefully we can dig into that here today.

0:01:57.6 TW: Oh, we’ll talk later. [laughter] All right, maybe you just didn’t recognize it. Well, for this discussion we thought it would be good to bring in a guest who’s been tasked with projects throughout his career that required thinking a bit differently about what data he and his team should bring to bear on the work. Jason DeRise is currently the Head of Data and Analytics products at Liberty Mutual Investments. Before that, he did a pretty long haul and a variety of research and analysis roles at UBS, including as the Head of Data Strategy for the UBS Evidence lab. And today he is our guest. Welcome to the show Jason. Thanks.

0:02:31.0 JD: Thanks so much for having me. Thrilled to be here.

0:02:33.5 TW: Awesome. So, I’d love to kick things off by maybe getting your thoughts on this kind of overall framing that we have data that is like standard and familiar within an organization regardless of what the role or the task is. But then there are scenarios that in a lot of roles we run into where completely different data is worth tracking down to sort of best address whatever the business need is. Is that even a fair way to divide up data that makes sense?

0:03:03.8 JD: Yeah, I think it’s a good starting point because everybody just gets sort of comfortable with the data that they were trained on or the same approach that the business has always done. But you start to get limited in the types of questions you can answer, and I think being able to expand out to other types of data sets, maybe to fill in the gaps that the current toolkit doesn’t really enable. Hopefully, we’re beyond hammers and nails, but everybody seems to have their favorite tool that they’ll go to.

0:03:30.4 TW: So, this may be more of a UBS question, just generally that role, the type of analysis you were doing that was kind of analysis to actually provide differentiated services. Is that fair? Can you give the quick non-proprietary get no one in trouble description of your role?

0:03:49.1 JD: Well, I started at UBS in 2007. I was a senior sell-side analyst and it meant that I was writing research reports with recommendations on what to buy and sell. And so very publicly visibly right or wrong on these recommendations for a very specific sector. So, I was based in London covering European beverages and for a time based in New York covering retailers. And to do the job well, you needed to be different than what everybody else thought to be right. ‘Cause it doesn’t really help to say, “Oh yeah, everybody agrees this is gonna happen. And I agree too.” Sure if that’s what I believe, that’s fine. But I would focus on where I was different to try to add value and help institutional investors make a better decision with whether that was our hedge fund clients or clients like where I am now at Liberty Mutual Investments, which is more of a long-term 100 billion dollar fund that’s investing across multiple strategies.

0:04:42.4 JD: Ultimately us investment professionals have to make a decision to try to do better than the market and allocate the capital, right? And that’s a competitive market. What most people do, is they’re looking at the same data that everybody else is. And that’s a little bit more challenging ’cause you have to then be smarter than everybody else. Some people are, but most people are into, there’s the Wisdom of Crowds. And so the way that I’ve always approached this as an analyst and especially as an outsider, ’cause I would never have by law, rightfully so, never have insider information. I’m looking at the same financial statements as everybody else. Looking and talking with management. Maybe I have a little bit of access in public forums to be able to ask a question on a conference call instead of reading about another analyst asking it.

0:05:28.1 JD: But to really be able to make a difference, I would need to get other data to fill in the gaps versus what’s already available to everybody else. And so it really started with having the right question, having the right framework of how the world works for the sector or for the economy. And really just trying to make sure when I bring in extra data, it’s actually adding value. So, as an analyst, I was constantly looking for proprietary data or proprietary techniques. And as technology continued to advance and more data became available, it started to get really interesting of what could be done. So, we started to explore lots of different types of data sets beyond what the core hammer and nail for a financial analyst would be. It had your financial statements and the company commentary, the news that everybody else is looking at. And you could do a lot with that. You can do a lot with that prices and forecasting where share prices are going. But again, everybody is doing the same thing. There needs to be a way to get an edge.

0:06:30.6 JD: So, much to unpack there.

0:06:31.8 JD: Yeah.

0:06:31.9 TW: Well, but so telling that description, I feel like I remember maybe this is like a famous story that people in the circles, there was somebody who wound up looking at, I think must have been a similar type of role that was like, “Oh, with satellite imagery we can now tell the depth how full, because you got overhead views of tanks of oil or gas or something.” And it was like, “Oh, that’s data we can get that tells us kind of where demand is gonna be ’cause literally how much of stuff is in a tank on the other side of the world?” It’s that sort of, it’s not insider, it’s available whether you pay for it or not. But somebody had to have the idea that this might be relevant and then presumably went through a process of… Is that the right sort of thing?

0:07:12.0 JD: Yeah.

0:07:12.7 TW: Okay.

0:07:13.8 JD: It’s a great example because it’s exhaust data from another purpose. So, all the vessels have tracking devices on them and it’s reading out the speed, it’s reading out their geo coordinates, it’s telling how deep the vessel’s sitting in the water, and this is really to make sure one, the vessel doesn’t go missing. Two, it doesn’t crash into things and it’s really more about safety and insurance actually. But if you take that data set and you turn it on its side and you aggregate it for specific questions around what’s happening with global trade, what’s happening with maybe specific commodities because you can combine that data with a data about the different ports and you can know certain ports are oil ports or iron ore. Now, you can start to get more of a real time feed of what’s happening in the supply chain.

0:08:06.9 JD: So, we actually did that. That was one of the many things that we tried to build and successfully build was one of the more popular products that we created using these alternative data sets. But to do it properly, it’s billions of combinations of data points ’cause you care about all the combinations of points and you care about 20,000-ish vessels. You don’t need every vessel, just the larger ones. Looking at how deep the vessel sits, tells you about its capacity utilization. So, we figured out a formula that would work well with what’s actually happening in real life. And then you’re taking some ground truth data and you’re building a model to make sure that it’s actually making sense. So, there is some data available at ports and there are company results that maybe a certain set of vessels that can be mapped to a port, mapped to a company, and you would know that it’s that company’s coal shipments. And then ultimately you’re able to tie it out with the trends that the company is reporting.

0:09:04.3 JD: We were able to see… If you rewind the tape back to 2019, the debate pre-pandemic was around trade wars between the US and China. And maybe grain wasn’t gonna be imported into China. You could actually see the… Is this rhetoric or is this actually reality happening? And we could see that the shipments out of the US for grain actually to China actually dropped. And you could see that before government statistics were available, which would use other methods. And then through the pandemic and all the supply chain issues, we were able to see congestion at the ports and start to piece together the mosaic around what was happening with supply chains. And that’s just one data point. Now, there’s limitations to the data points too ’cause you don’t know certain commodities, you would know what it is, but if it was just a container vessel you wouldn’t and it was sitting deep in the water, you wouldn’t know if that was filled with iPhones or filled with marbles.

0:09:58.9 JD: So, you have to know the constraints of what the data can tell you and what it can’t tell you. So, that’s an important part of using this type of data. Some people can get themselves into analytical trouble by over extrapolating what you can and can’t learn from a data set. But that’s where you start bringing in different types of data to piece together the mosaic.


0:10:19.3 Michael Helbling: It’s time to step away from the show for a quick word about Piwik PRO. Tim, tell us about it.

0:10:25.6 TW: Well, Piwik PRO has really exploded in popularity and keeps adding new functionality.

0:10:29.9 MH: They sure have. They’ve got an easy to use interface, a full set of features with capabilities like custom reports, enhanced e-commerce tracking and a customer data platform.

0:10:41.3 TW: We love running Piwik PRO’s free plan on the podcast website, but they also have a paid plan that adds scale and some additional features.

0:10:49.7 MH: Yeah, head over to piwik.pro and check them out for yourself. You can get started with their free plan. That’s piwik.pro. And now let’s get back to the show.

0:11:01.3 TW: Wow. So, how did you get to the point of being able to think that way? Were you trained in economics going into this job where you could think big picture, bring those things in or what was the gateway alternative data set that kind of was the snowball effect to get you to that scenario? Because that was mind blowing and so impressive.


0:11:19.8 JD: Well, my training was Finance and Accounting at Syracuse University and a minor in Economics and I knew I wanted to be a sell-side analyst coming out at school and…

0:11:30.3 S?: Ah.

0:11:30.6 JD: So, everything to me was like, “How do I do financial analysis?” And I remember there’s one finance movie and it was making jokes about, “Oh, it’s boring to do cash flow analysis.” I’m like “Wait, what’re you talk about? This is amazing. I wanna do cash flow analysis.” [laughter] Discounted cash flow models. And what I really like about the financial markets is it’s a puzzle that you don’t get all the pieces and actually there’s probably some other puzzles thrown into the mix that don’t matter. So, you have to actually figure out which puzzle you’re trying to solve and just as you’re getting close to seeing that picture, probably the rest of the market is too. And at that point no one cares what that answer is because now it’s in the share price.

0:12:08.2 S?: Oh my gosh.

0:12:08.6 JD: Like once it’s known, the company reports the results and they were better or worse than expected or whatever catalyst happened. Now, everybody knows, but the market doesn’t stop moving, they’re moving on to the next debate. So, just imagine another puzzle dropping from the sky with half the pieces and you got to sort through it. So, I always enjoyed that part of it.

0:12:26.6 TW: I just got an ulcer just listening to you tell that story.


0:12:29.8 JD: You have to be comfortable with being wrong sometimes, and if you’re wrong, being able to cut your losses quickly, and then if you’re right…

0:12:37.9 S?: Oh my gosh.

0:12:38.3 JD: Being able to stick with it ’cause there’s always gonna be an uncertainty and being able to just sanity check your views always. And that’s where data was important to me. So, where did I learn to take apart these problems? It was my first sell-side job at a small boutique called, it’s not that small, but Sanford Bernstein. Very fundamentally driven research. And I saw firsthand the senior analyst, I was a junior associate. The senior analyst, if they ever got up in front of the Salesforce with their recommendation and they didn’t back it up with some data point, that Salesforce would just like tear them to shreds ’cause there was no way that Salesforce was gonna go out to our clients and say, “Here’s our call.” If they couldn’t back it up. It was a very data-driven approach to research. And so that always stuck with me. Plus I was working with brilliant people who were exceptionally good at first principle thinking and were able to take apart problems.

0:13:25.2 JD: And so I just tried to absorb as much as possible from really bright people. And I think in Evidence Lab we’d had an amazing group of people. So, we had this idea about looking at the vessels, as I was talking about. I took a shot at it and I kind of got stuck in the details. We actually had somebody who was an expert on it. We hired them to really come in with that expertise on how commodities traded and what different types of data sets would align to that. But we had all sorts of different backgrounds. There was multiple people with psychology backgrounds. There were climatologists, there were geospatial experts, there were people experts at social media data. And so we brought this unique collection together and it was very much a design thinking model. We just wanted to get as many ideas on the table as possible and then kind of filter down to what we actually thought could be done. Do a small test, see if it worked. If it worked, try, try again, try again. And eventually if we found product market fit, we really tried to scale it up.

0:14:27.8 VK: So, one of the things that I do have the privilege of, I’ve worked on the smart platform team alongside Jason for UBS at a couple years. And so learning some of these concepts was incredible to see happen in real time. So, one of the things that I could not wrap my brain around and probably still can’t to this day is, so you do this huge analysis, you take all that data, you turn it on outside, and you ask those questions and use it for a different purpose. But how do you… I’m using air quotes here, how do you price it into the model to understand how it actually affects the share price, and how do you not get entertained by what’s amusing or interesting versus what’s actually gonna move those debates forward? Because I’d be like mind blown so many levels. How do you not get distracted, but I’m interested in your thoughts.

0:15:11.6 JD: Yeah. Well, that is a challenge ’cause you can easily get excited about a technique or a data set and not actually be focused on the question that we’re trying to answer. And we weren’t perfect at that. This is one of just the learnings along the way about how do we make sure that we’re generating value with the work that we’re doing. And what we found the most important piece of it was not starting with the data, it was starting with the questions. And in particular it’s questions where you imagine you have the answer, would that answer actually change your opinion? And so we didn’t get the analyst doing a full scientific method. Val, I think your AB testing was way more scientific than anything that the research analysts would do. But we would try to get them into that mindset where, “Okay, I think x, y, z is gonna happen for this company and if this data came through and disproved it, that would actually be a huge value because now we know that the data is saying to go in a different direction.” And if it failed to disprove it, great, I can keep my view.

0:16:11.9 JD: But all it takes is one data point to really break an investment thesis. And the best analysts that I worked with were really good about just following the data and not falling into any biases about what their views are. So, when we were very question-focused, a lot of these questions were big questions. So, like imagine it’s January 2020 and China announces that they’re doing a quarantine for a few cities. Like at that point there’s just complete uncertainty and huge questions. And we spent time figuring out, “Okay, well, what are the smaller questions we can answer?” And follow that through. And some things worked, some things didn’t. And if it started to work and it was showing us what was really happening, we put more weight in those data points. So, that works pretty well. I think the things that don’t work very well is like a senior person is like, “I have a great idea, let’s do it”


0:17:02.4 JD: And then no one feels comfortable saying no. [laughter] So, you end up doing the work and then no one uses it. So, I think it’s being very outcome-driven about what you’re trying to do. But in an investing world, I really think in any business intelligence, you have a question, you don’t know what the answer is, you start to bring in information. It could be anecdotes, it could be data. But if it’s not answering that question, it’s just time to put it off to the side. And sometimes a simple, A divided by B is a good enough answer. Other times we needed to bring in machine learning techniques and natural language processing ’cause that actually was the thing that unlocked the insight. So, it really depended on the situation.

0:17:43.7 S?: I love that you called out too that it was asking questions that if you got the answer it would change something you were doing or something you thought. Like you’re not going for questions that it’s like, “Oh, okay, cool.” When you got the answer. [laughter] I love that caveat and distinction that you made because I think that’s really powerful, especially in a lot of scenarios I find myself in… With a lot of questions that are like. [0:18:02.8] ____.

0:18:03.5 S?: Interesting.

0:18:06.2 JD: Yeah.

0:18:06.3 S?: Good. Glad we know that. [laughter]

0:18:06.4 S?: Yes.

0:18:06.7 S?: Yeah.

0:18:06.8 JD: As we started, we were just happy to take any work and try to move forward. But as we proved the concept out and we had more buy-in. And that was also… I felt like we were leading the way in the financial industry. There was a few others that were growing at the time that we were… We learned along the way if we heard interesting, that was like, “That’s like a hard no, we’re not doing it.” And we would tell you, “It’s like interesting, it’s not good enough.” It’s gotta be have to have, otherwise we’re gonna deliver something to you and you’re gonna be like, “Okay, what do I do with that?” [laughter]

0:18:37.0 S?: All right, because these stories are all so good, I would love to hear what’s one of your favorites, especially if it was like, oh, this is never gonna work, or it’s gonna take too long for us to put our hands on this or wrangle it. How is it ever gonna actually inform the research? But then your mind was totally blown by it. ‘Cause if our mind was blown by your first example, I can only imagine what you have in your back pocket.

0:19:01.4 TW: This could be good. It’s got to be sufficiently obfuscated that it doesn’t get you in trouble, but sufficiently specific that you blow our minds again. Go.

0:19:10.5 S?: Walk in a tightrope.

0:19:11.7 S?: Are you up for the challenge, Jason?

0:19:13.4 JD: Yeah, I think so. So, there’s a few that come to mind. I think I was extremely proud of all the work that we did as a company around the pandemic, because there was just so many uncertainties. And it really required us to look at all sorts of data. And it also was at a time where the market knew that they couldn’t wait for company results. They couldn’t wait for GDP to come out. At that point, everybody would have known the answer. And even if you waited for company results, what management team knew what was gonna happen next? And so we put together a whole combination of data sets where I think it was, sorry, so I’ll pick one and then we can always talk about a few more. Going into the early stages of the pandemic, I think we’re gonna remember like March 2020, things were selling out from grocery stores and other things were just plentiful and no one wanted it. The behaviors were totally changing. We were using web robots to go across Amazon and other retailer sites every week collecting as much or if all of the assortments that were available, the prices and a whole bunch of attributes about the products, the brand name, the product, the category.

0:20:21.2 TW: So, can we pause for a minute to let everybody who, all the analysts who are listening, who are like shaking their fists at their headphones right now, you’re like fucking bots yeah driving traffic that I had to okay.

0:20:32.7 S?: Jason’s like that was me.

0:20:34.7 TW: You know bot exclusion the bane the bane of the digital analyst but there you go.

0:20:38.3 JD: Oh, yeah so no web robots I mean it’s important to talk about the compliance around that and making sure that we’re good corporate citizens as we do it that we’re not taking down a website by accident.

0:20:49.2 TW: They were easily excluded bots okay.

0:20:55.4 JD: Yeah I mean we you… At UBS, and I would say the same for every financial firm, if they’ve got a good risk department, they don’t wanna do anything with alternative data that could put the firm at risk reputationally or otherwise. So, our standards for web scraping, even though we could anonymize our IP address, we actually would put in the header how to contact us to tell us to turn off the robot.

0:21:13.5 TW: Oh, wow.

0:21:14.7 JD: And we would make sure that we were less than 5% of daily traffic, which… For some of these consumer sites, there was no chance that we would even be close to it. Maybe some B2B sites, we actually could have gone past it and we just wouldn’t scrape the site ’cause why risk it? But yeah, we added some traffic to the site.

0:21:31.4 TW: Yeah, sorry. Distracted you.

0:21:33.3 JD: Yeah, no worries.

0:21:34.6 TW: I was a little triggered.

0:21:37.7 JD: By collecting every week and going through the full assortment and seeing the prices and understanding how we were basically able to reverse engineer price and back into what was happening with inventory. So as the pandemic hit, we could see certain categories selling out just from the scraping. We could see if the prices were rising as a trade-off. We could see where there was products maybe like in a fashion space where they have to get rid of the inventory for the next season, and it would just sit there and sit there and sit there, and the price would just keep dropping because they can’t move the inventory. Meanwhile, other categories you know, they don’t have the inventory because they can’t get it in. And so the availability of new products is dropping. So, we were able to reverse engineer what was going on. And there were certain times where, you know, companies would run into an issue and the market, it was hiding there in plain sight. It’s on a public website.

0:22:30.2 JD: It’s available to be used, but without the technology and without the thought process of setting up the price analysis and the inventory analysis, the way somebody with actual data, you know, from the inside of the company would look at it. You just wouldn’t be able to know until the company revealed on the results or pre-announced. And so there were times where we would see, in particular in fashion companies, they would get into trouble where it would get to the clearance month and they would promote heavily and the items wouldn’t clear. And they would start spilling into the next quarter and still there. And that was just like flashing red signs like something’s wrong. And then there were other times where you would see the inventory just clear and they didn’t promote that much. So, you had like really positive signals.

0:23:15.3 JD: So, that’s just like one example of what we did. But we were combining that approach with where are the cell phones anonymized, transaction data anonymized at an individual level. App usage anonymized. And you start to triangulate all these data points and you can look at it from a geographic point of view and start to see the consumer behavior as lockdowns were lifted. Did they actually come back or not? Some markets came back more aggressively, some didn’t. There were all sorts of debates happening about what the new normal would be. And so de-urbanization became a hot topic. We were scraping the prices from you know, leading property portals in multiple countries and analyzing it the way a residential real estate analyst would and seeing that shift in inventory levels. There’s also government data. If you do a freedom of information request, one of them became very popular and multiple people asked for it and then they standardized it, but you can get mail forwarding anonymized and aggregated. So, you could see people were moving to.

0:24:18.4 TW: Can we like for a second, it seems a lot of your examples, it seems like you were having to balance. You’ve kind of blended together, I kind of wanna tease apart, there’s pandemics hitting or whenever in many, many scenarios, there’s a need for speed. I have an idea, but if it’s going to take me a year to go and build a web scraper and stand it up and process that. So, there’s a need for speed. There’s a need for the cost of actually getting the data, whether it’s scraping or freedom of information. I mean, that can take a while. Then there’s kind of the quality of the data itself. And then there’s like the relevance of the data. Like it seems like all of those are playing.

0:25:00.1 TW: I mean, I’m flashing back to the start of the pandemic where one big client and they’re like, what’s gonna happen to me? And I’m like, I don’t know. You want me to compare it to the 1918 pandemic and web traffic back then? I certainly did not have the ability to, it was kind of the speed thing I think is what killed me. It’s like even if I wanted to go get this other thing, I gotta go uncover a whole data source. So, did you, when it, the speed and the relevance like well, could we look at this? Well, could we get it fast enough for it to matter? And then does it actually matter? Were there dead ends you were pursuing where you’re like we could look at this? Well, it turns out we can’t get that data or we can’t get it soon enough. Or we can’t wait for the by the time the GDP is reported, that’s, you know, too much of a lag. So, were there dead ends as well? And what caused the dead ends?

0:25:51.3 JD: Yeah, there were definitely dead ends. We tried a lot of things that never saw the light of day. And maybe it was because by the time we started collecting the data. That question was no longer relevant. And so we just sort of paused it. We reached out to companies that never sold data before and would just have a conversation about, would they be willing to sell their exhaust data? It was interesting at that time, lots of companies wanted to be helpful. Even Google put a free data set, a number of big companies that were tracking locations put free data sets up with no history, right? They just turned it on. And so you didn’t have even a baseline to compare it to like forget it. What foot traffic looked like a hundred years ago. You didn’t even know what it looked like in like January. Like it started in February of that year or something like that. So, you have to come up with ways to use the data within its limits, have the right questions or thresholds or some sort of benchmark of, okay, if we’re gonna reopen, this is what it could look like. If it’s a new normal, this is what it could look like, and just sort of think it through and then try to test versus that.

0:26:57.1 JD: But yeah, it’s starting from scratch, obviously, in a moment like that meant that there wasn’t much history. But at the same time, even I don’t think that history would have been very helpful. Like everything was just throw all the, it’s actually harder now use, I would say using some of these techniques to understand the nuances of what’s happening in a normal time where the market cares about a few percentage points, acceleration or deceleration there. It was like no one’s moving. And now there’s a strong signal that people are starting to move again or not, depending on where we were talking about. So, that was a little bit easier to get a signal from it. Also being early is a huge advantage.

0:27:38.9 JD: Because especially in the financial markets, I mean, I’d be curious to get your take on this in the corporate world if there’s a parallel, but because of the way the financial markets work, where everybody’s doing work, if you’re working on the right question and everybody in the market’s working on that question and there’s really uncertainty, the first thing you do is gonna add a huge amount of value. Then the next bit of analysis somebody does has to be better than that because otherwise people already get it. And if you’re the last one to come out with this perfect analysis, it’s probably too late. So, being early and not precise, but in directional proximity to reality, then you’re gonna be in a better place. So, I’m really actually curious about from a corporate side, is there like a parallel with that? Maybe with new product launches or.

0:28:22.7 TW: I mean, yeah, I feel like it’s just, we don’t even think about it. I mean, I have like this one simple example, which was a former coworker of all three of ours was. Sam Burge, when she was working on with the home builder, and it was just one where trying to figure out if stuff was having an effect, stuff that was being done, details don’t really matter. And she was like, well, what if we looked at housing starts data? We can get housing starts data with relatively low latency, pretty simple. I mean, it sounds like playing with the oversized Lego building blocks compared to some of the stuff you were trying to pull in. But just saying that would be a normalizing factor because we’re really trying to figure out if we’re outperforming the market is a fundamental question. And it was like, well, that’s amazing. And she was like, “Can I go get it? Can I get it relatively quickly? Does it actually seem to help?” As I recall, there was a little bit of hesitation of, from some party saying, why are you bringing in this other data set? And I’m like, this seems so obvious, like this seems so helpful. And there was a degree of pushback. So, I feel like the corporate side just gets. Really stuck in, let’s just go deeper and deeper and deeper in the data that we have.

0:29:40.7 JD: Sure.

0:29:41.4 TW: And doesn’t like… Doesn’t look out. They don’t have that kind of competitive pressure. They may think they do, but they don’t. And therefore there’s not an incentive.

0:29:51.1 JD: I see. But it’s interesting because from an outsider looking in, there’s moment, I mean, they don’t come along all the time, but there’s moments in industries where there’s something new, it’s a new product or two companies merged, or there’s a regulation change. And you kind of have to say, okay, this could be a new normal, or it could be a temporary thing. So, so many different debates could be like that. Rewind the clock five or six years ago on electric vehicles, and I would say electric vehicle is still a debate. Where is the penetration gonna be? But if you go back five or six years ago, it was still a question of, is this a risk to the big companies? Is this an opportunity? If you were just sitting around waiting and collecting the same data that everybody else had, how would you even try to solve that problem? Because there’s no history to really look at in electric vehicles. So, if you were going out and getting different types of data to try to understand what was happening, reverse engineer the buyer’s decision process, how companies make decisions, what the economics of the batteries are. Now, you’re starting to piece together something, and maybe that leads to an early decision versus competitors. But again, you have your hammer and your nails.

0:31:05.5 S?: So, this is the perfect tee up. I’m hoping we can talk about it, which is why I got excited to the test. It’s like Val’s ready to fly out of your chair.

0:31:16.6 TW: Val needs a potty break. I don’t know. I can’t tell.

0:31:19.0 S?: But can you talk about the Tesla teardown?

0:31:22.3 JD: Yeah.

0:31:23.6 S?: That’s a super fun one. Rewind the clock again.

0:31:26.1 JD: So, we decided that the biggest part of the debate, and this is really coming from our analysts and just understanding when would the cost of an electric vehicle to produce it be comparable to a combustion engine vehicle. And the battery is the biggest part of the technology. So, first we bought a Chevy Bolt and we had paid for a company to take it apart piece by piece.

0:31:51.3 S?: Like someone counted all the washers.

0:31:53.6 JD: Yes.

0:31:53.7 S?: All the different like I’ve put it in little piles like wearing white gloves in like in this huge, it was in Detroit, right? It was like this huge like video production, all these little pieces and counted them all up.

0:32:04.6 S?: How does it feel to hold that kind of power? I wanna say. Somebody go take apart a car and count the pieces for me so I can do it.

0:32:14.9 JD: And the only reason that this made sense is there were so many questions about it. So, actually probably the biggest surprise at the time was how many semiconductors were in that vehicle. And then I think you roll forward and I may be messing it up, but I’m pretty sure it was a Model 3, so the cheaper Tesla we purchased that one, had another company tear it apart piece by piece. We did it again with, I think it was a Volkswagen a few years later. And other times we would just buy the battery and we would take apart the battery. And over time, you were seeing the cost to produce come down, the number of semiconductors and who the semiconductor companies were increasing.

0:32:47.5 S?: Like the whole supply chain they figured out from this.

0:32:50.4 JD: Yep. And then that, you know, good research leads to more questions. And then it’s like, all right, how do we track, you know, the supply chain of this and who’s gonna win there? And then we would do lots of market research type surveys. There’s, I don’t remember the exact number, but let’s say something like 10,000 consumers globally asked every year about their, you know, perceptions and what they think they might do. Obviously you need behavioral data to validate that because just ’cause you say you want to buy an EV doesn’t mean you’re going to, but we started to combine all of this together with collecting charging locations and thinking about, okay, how many charging locations do you need to, because the research showed range anxiety is a big deal.

0:33:35.9 JD: So, how do you solve that problem? The Tesla and other cars, the apps are such a big part of the driving experience now. So, you can look at app downloads and app usage to get a sense of what’s happening. So, yeah, we just pieced together a whole mosaic around it. But again, it was a debate where any answer would have been valuable. Like if the answer was EVs total, it’s a whole nothing, it’s never going to amount to anything. Like if the data showed that’s where our analysts would have gone. If it showed that the S curves of all these metrics were accelerating, they would have gone that direction. And, you know, and then as the reality shows up in the results. That starts showing up in the share price and the questions start to change. So, then the work starts to shift again.

0:34:19.5 S?: It’s amazing to just hear you kind of follow the thread of you started with the big question and you started evaluating it from all these different angles and it led you to ask this question and think about it, you know, this part of it. And I mean, do you have, have you ever been able to distill down like that you are working through somewhat of a consistent process when you get a brand new question like that? Like is there any tips, tricks, or like steps you could outline for us here? You get the big crazy question. Yeah, like.

0:34:50.9 JD: Yeah.

0:34:51.0 S?: How do you even start to tackle that mental model?

0:34:54.4 JD: Yeah. I think in the beginning, it was a bit harder to get people to think this way. So, when it was working well, we would try to get our stakeholders to open up about what they actually were trying to solve. You know, get them to talk more in outcomes. You know, I think the adage of, oh, I just need the data, just get me the data. For us, it wasn’t that. It was actually, I wanna do a survey on X. And it’s like, okay, but why do you wanna do a survey? Which to me, you know, there’s a lot of value in surveys, but it can’t answer everything. So, trying to tease out of them what they actually were after. And to me, it was more about what they didn’t know. More so than what they thought the answer was. Like I wanted to know both. And once you could find what they didn’t know or what the uncertainty was, and we had a way of answering it that was different, I think you had their attention as long as they were open-minded about it.

0:35:45.4 JD: And so over time, we found early adopters showed value, would create a bit of social proofing or FoMO and others would be like, oh, wait, that question is similar to my question. Can you help on this one? And, you know, we took on questions that we ultimately couldn’t answer sometimes. I mean, we were given a lot of room to try things and have it, you know, fail is not a big, you know. Bad. I mean, obviously if the teardown somehow failed, like that would have been pretty bad, but because so many questions were important about that, that, you know, we, it turned out it worked out well, but there were other things we thought we could try to answer something. We played with the data for a while and eventually had to give up and it wasn’t, we kind of celebrated that. Okay. That’s one more way to get value out of data and just move on to the next one. So, I think it’s really about that, a process around the questions is really the main driver.

0:36:38.6 TW: Would you go to, I’m thinking through more of the background I’ve come from that if it’s, oh, there is some other data we could go get that may be useful and may be cost effective and may be timely enough, but I would wind up going through and kind of going to my business partner and saying, Hey, that question, I think the best way to answer it is we need to go get this other like there’s a cost like there is a cost to going. The data we have in hand is much, much cheaper. So, I guess were you fighting for budget? Did the people who were coming to you with the questions, did they have skin in the game or, you know, and I guess for Julie and Val on the more corporate side like I don’t know how we. How do you handle that? There is a cost to it, right?

0:37:24.4 JD: Yeah, we didn’t have a huge budget to start, but as there was more demand and positive feedback about what we were doing and that we were showing value.

0:37:34.3 JD: We were able to grow the budget. Eventually we got to a point where you couldn’t grow the budget. I created a, so I had covered European beverage space and I covered Anheuser-Busch InBev, and they were part of their model was the zero based budgeting approach. So, I spent a lot of years studying, and so I brought that to Evidence Lab for data. And the idea is if it’s done properly, it’s not really a cost cutting tool, it’s a way to optimize your spend. So, if something’s not generating value, we set up metrics on our own business to know if something was generating value. If it wasn’t, we were happy to stop doing it and then take that money and put it behind something that was generating value. So, that’s more of like an optimization. But yeah, if you know that something’s important in the investing world and you can have an edge, and I think more from, even from the buy side point of view, if buy side meaning the institutional investors, we were on the sell side selling advice to them.

0:38:25.4 JD: I mean, some of the asset managers like ours, huge amount of money and small basis points of our performance could be a huge difference. And so as long as you can show that connection, either directly or through our stakeholders ascribing value to the work that we’re doing, we would be able to keep doing it. But not every, again, not everything works. So, we have to have that flexibility of trying something. And that doesn’t kill the budget forever if you have a couple failures in there, it’s like that’s just part of what we’re doing.

0:38:54.6 TW: So, you didn’t start by saying, we’re gonna solve the entire electric vehicle forecasting supply chain. You probably could start a little smaller than we need to buy a new EV every six months and have it dismantled.

0:39:09.2 S?: That’s where you use the hammer.

0:39:10.4 JD: Yeah.

0:39:11.1 TW: You said starting small, it probably starts with the real clarity on the question, questions that are coming in. Are there ways to start small, relatively accessible, relatively timely, relatively straightforward, and it is something that would have to grow to the point that it was like, oh, this is a cost of doing businesses. We need to be able to explore or have that arrow in our quiver.

0:39:38.3 JD: Yeah. And the tear down also had a benefit for us that it was, there was a bit of marketing as well.

0:39:43.2 S?: Totally.

0:39:43.6 JD: I mean, it actually did answer some questions, but I think there were some higher ROI data sets that we either partnered with a data vendor or created ourselves that were also answering important questions. But that one in particular helped a lot on the cost of the battery. We didn’t have really better ideas than that. And so it was an expensive one, but thankfully I had budget for it.

0:40:05.6 S?: And it sounds like part of the case for getting that data was saying it was crucial to answer questions that were really impactful. Like their answer was going to be very impactful to what you guys did based on that information. And maybe for people that aren’t getting to tear down cars and working in that space, but like in our space, maybe that is something you have to think about when you want to request.

0:40:31.9 JD: Yes.

0:40:33.0 S?: The time cost or the actual cost to pull in a different data set.

0:40:36.3 JD: And that we needed a lot of buy-in from a lot of stakeholders, senior management support. It wasn’t just like, oh, we could just do whatever we want. Like we, that took some convincing, and but what’s interesting about it too is that sparked a whole bunch of other ideas of like teardowns or like physical tests of things. And one of our analysts covered companies that had ball bearings and there was a debate about if manufacturers in China could undercut the European companies with a product that’s cheaper. And the question was, well, okay, maybe they’re cheaper, but they don’t perform as well. And we literally paid for these ball bearings to just be like, put like the pistons just firing or whatever. And the thought was, oh, in 30 days these things like just 30 days straight, this thing will fail. I can’t remember how long we let it run before we stopped. Before we said, these are equally good. Like, I mean, months. It just did some labs somewhere just running. One of my favorite things that we did, I mean, it was surprising like everybody would’ve assumed that Oh yeah, eventually the low cost providers one would fail and it, it didn’t. So, but the research reports weren’t that exciting. We didn’t have anything to show other Hey, it’s still running.

0:41:43.2 S?: That’s interesting.

0:41:44.3 S?: At the end of the research paper. And it’s still running to this day.

0:41:49.7 S?: Check in next time for updates on.

0:41:52.3 TW: Yes. Well research paper volume second edition. Yeah.

0:42:01.5 S?: I have a slightly different question for you, Jason. So, obviously you did a stint as a sell side research analyst before you were on the other side of the house in these evidence lab or your role today at LMI. How much do you credit some of your ability to like cut through and have some of those successful conversations like you were just describing because you were in the role of your stakeholder before they became your stakeholder or your partner? Like how much easier was it for you to either empathize with them or just really craft your approach to the problems? I’d be curious how you feel like that played a role.

0:42:33.4 JD: I mean, I guess you would expect me to say I thought it was really, valuable because that was my background, but I think it helps being able speak the language of your stakeholders and use the terms that they use in the right way. And also, yeah, the empathy about it, understanding what they’re going through. I think one of the, on the sell side, there’s this morning meeting where the best, I mean UBS would publish the sell side in general, just publish thousands of research reports and only the best ideas get on the morning meeting for the Salesforce to amplify it. And they would ask questions and I would hear analysts struggle to answer a straightforward question. I’m like, ah, that’s an opportunity. They just struggled through this. They sort of said, well, management said this, or I think it might happen. Just going back to the Bernstein days. As soon as there wasn’t a data point behind it, the UBS Salesforce was much nicer than the Bernstein Salesforce. But that would be the opportunity to reach out.

0:43:25.4 JD: I’m like, Hey, I heard what you were talking about and I actually, I think I can help you answer that question. And if you catch them at that moment, they’re like, oh yeah, anything you could do it would be helpful because they just, you know, they went through that. But then there were other times where we had an idea of what we could do and it’s understanding the calibration of the answers in the financial markets a little bit different. So, I covered retail and I covered like grocery stores and like it was when Whole Foods was a public company before it was bought by, Amazon, and at the time they were growing 6% same store sales.

0:43:58.8 JD: So, that means not counting open and closed stores, how fast is revenues growing. And so if the market expected it to be 6% and they did 6.1%, the shares would go up like three or 4%. And if they put up 5.9 versus that expectation of 6, the shares would be down like 5-10%. Like those are the same numbers. Like this is kind of crazy but not understanding that. And you build a model that, oh, it’s gonna be right plus or minus 300 basis points, it’s just adding noise. So, I think if you’re a tech, you know, we have brilliant people who are brilliant data scientists or technologists data engineers, not having that understanding of the actual problem to solve, I think we were able to do well because we had enough people who understood the investment process and the technology and the data, and we would train people to be able to do all of that.

0:44:50.2 JD: It’s very hard to find people that can do each part of that, but I think that’s important in the hiring we’re doing, so, in LMI, we’re building a data practice here. We’ve always had data, we’ve always had technology, but what we want our ambitions for what we can make the data and technology do for us is higher than it’s ever been. And we’re building a data practice from the bottom up and having people who understand the investment process is a key part of what we’re looking for in addition to understanding the data and the technology.

0:45:18.1 TW: Wow. And after this, there will be a spike in applications to LMI, libertymutualinvestments.com/jobs? No, I’m not sure exactly what the, what it is we’ll.

0:45:32.4 JD: We will get if we get you something for the show notes. And the insurance part of the business. Like we support the insurance business. We have one client. I mean, actually that was one of the things that attracted me to the role. Not only was it getting to take all the things that I had, we made a whole bunch of mistakes in Evidence lab too. Thankfully we had really strong support and along the way we figured out how to scale it up. What’s exciting about starting fresh here is I kind of feel like I know the playbook. I’m sure we’ll uncover different ways to mess it up along the way, but you know, we’re starting from a way to get to the good part quicker. But the other part of what came out of, you know, my due diligence before joining LMI Liberty on the insurance side has a great data science practice really well regarded in the industry from the insurance side. So, I’ll definitely get you the way to see what positions are open at both parts of the firm.

0:46:21.3 TW: Well, the alternative data I’m consulting is my timer that says that it’s time for us to head towards a wrap. That was a terrible segue. And I was, I well I was also ready to say a great place to leave it is that Jason was saying, you’ve really gotta understand the business. And then we kept going for another couple of minutes. So, that was also gonna be a probably a clunky awkward segue. Michael, we miss you. Your silver tongue, normal standard bearer for our primary hosting. But this has been a great discussion. We’ve got like, I don’t know, 75 other things that I would love to ask you as well, but alas, time run short and we are the analytics power hour, not the analytics Power Day. So, we’re gonna have to move to wrap. So, one little thing we’d like to do before we close out is head around the virtual table and everybody share a last call. Something that might be interesting of note that somebody can take and use with themselves is if it’s we can skip you Jason, if you think it’s just go and disassemble a car But if you’ve got something else.

0:47:30.0 S?: Count all the washers.

0:47:33.7 TW: If you have anything else to share, is count all the washers as our guest. Do you have a last call for us?

0:47:38.4 JD: Yeah, I think one of the things that I’ve really always enjoyed having been a sell side analyst and putting ideas out there and being able to meet with clients, that wasn’t a key part of the evidence lab experience. Like we let our analysts be the outward facing parts of the business and we were sort of like the, if you take a Formula one analogy, we had our star drivers. We were like the thousand engineers behind making them successful. But being able to have a community to go to and share ideas and especially in, it is a niche space alternative data in financial markets, but there’s a couple low key happy hours that a few people in the industry have set up. And in New York and you know, if you Google it, there’s so many of these other lowkey happy hours or meetups, but like, to name a few for anybody that’s in New York, the Jordan Hauer from Amass Insights has this lowkey happy hour where you just pay for your own drinks, he’s just securing the place. There’s no cost to anybody just come and hang out, meet like-minded people trying to solve financial market challenges with wacky data sets.

0:48:42.9 Julie Hoyer: But aren’t they all aggressively told that they can’t say anything? I mean, is it like after the, do you worry about like narcs being there that you’re gonna like.

0:48:53.4 TW: Well. Share something?

0:48:54.1 JD: I think people are pretty well trained in the financial markets what they can and can’t say. And it’s, I think you have a good mix of investors and data companies. It’s a good conversation for sure. But then there’s like other data ones that are more tech like Ethan, Aaron from Portable does one in New York City. There’s even one where I live in Westchester, which is, you know, north of the city. Jason Taylor from Automated Data ADI has has one. So, they’re just like everywhere. And I just think it’s great having a community and being able to bounce ideas or what other most, we’re all dealing with like the same challenges of getting raw data and turning it into something useful. And so even just hearing that, we’re not the only crazy ones that are spending all this time cleansing data that’s pretty valuable.

0:49:36.5 TW: So, when you go to those, are there always people, because I run a meetup in Columbus and like you it’s that first time showing up. Like is it, I mean do they run them on meetup.com? Is it like, can anybody show up, if somebody’s like, this seems wild. I’ve been trying to figure out what I wanna do next with my career. Is one of those meetups a good place to show up and chat with people?

0:50:01.4 JD: Yeah, I’ve actually recommended that to a few people that were interested in getting into the space. It’s really, anybody can come to these. The way, I mean as an example, I think actually both Ethan and Jordan do it this way. There’s just like, a Google form that you, you put your email in and now you’re on the invite list. It’s not a high hurdle. But yeah, I mean like you get, I mean granted it’s New York City but there were times I would go and there’d be like 90 people there and a whole basement of a pub. Just data, people chatting about.

0:50:35.7 S?: Just nerding out.

0:50:35.8 JD: Whatever they want.

0:50:35.9 S?: Having a beer.

0:50:36.0 JD: Nerding out. Yeah, That’s right.

0:50:36.8 S?: That’s my people.

0:50:40.6 JD: Yep.

0:50:40.7 TW: You know what Excel shortcut I learned this week?

0:50:46.5 S?: Love it.

0:50:47.3 JD: Yeah.

0:50:47.7 TW: Wow. Awesome. Julie, what about you? Do you have a last call?

0:50:51.5 JH: I do. Mine is an article that has actually been shared with me by multiple people. It’s been very top of mind and I still have mixed feelings about it, but it was such a thought starter. I wanted to share it with the group. It is by Keith McNulty on Medium, it’s decision makers need more math. Rigorous mathematical thinking is missing in most decision making environments. Half of me in my heart’s singing like, yes, I totally agree. And then you read it and you’re like, uh, maybe I felt then suddenly conflicted. But it’s really good. It’s not a super long read. He makes some really great points of what he thinks decision makers are missing. I think he has three key things and then he even includes like a nice visual Tim, I think you might think it’s a decent one even. But it was kind of cool. It was talking about for jobs, all these different jobs, how do they fall on the scale of requiring social skills and requiring math skills? And some of them I was definitely surprised like where they fell, what quadrant they were in. So, it was just chock full of good thought starters and hopefully someone else out there will find it helpful.

0:52:00.9 TW: Ah, that looks amazing. Val, what’s your last call?

0:52:03.8 VK: So, my connection to this one, Tim, you’re gonna laugh. So data. So, I’m gonna talk about alternate browsers to Chrome. So, huge shout out to Kyle [0:52:15.3] ____ from Further and also Gabe Webb and Allison Murphy for getting me totally hooked on using Arc and as a Windows user insert laugh track, I had been dying for Arc to come out of beta for or to come out with the beta for the Windows version and fell in love with it. And then they came out with some new features called the easel where you can make this whole section of your browser like live and you can kind of share, it’s like very image based. You can do like web crops. So, if you wanted to grab your weather and put it on your easel, so that’s like what you open it up to. You always have that like live updating for you. But it’s super powerful, super flexible and I can like, the more I play with it, the more ideas I think of ways that I can make my productivity work harder for me. So yeah, huge fan. So, it’s the collect your internet with easel feature for the ARC browser. So, I’ll have both of those links in the show notes. But Tim made fun of me earlier because.

0:53:12.4 TW: Well, I feel like we jumped back and forth. I was like, I don’t know that you opened up by saying I’m recommending a web browser and then you got pretty deep into all the things that you were like arc, arc, arc. So, okay, so we’re talking about the ARC web browser.

0:53:25.5 S?: Yes. And the easel feature with it.

0:53:28.7 S?: Easel’s a feature within the ARC browser. Yes.

0:53:30.7 TW: Within the Arc browser. And does it suck down memory? Chrome just kills me with how it eats memory.

0:53:38.5 JD: You’ll have to test it out. So that sounds very cool.

0:53:41.4 S?: It is very cool. All right. Tim, how about you? What’s your last call?

0:53:43.6 TW: So, I’m gonna straight up Log Roll as I think listeners probably know I’ve got a little new company that Val and I are two of the co-founders of. But one of the things I’m really excited about with it is that I think we’ve got some great content. So, going back to the medium world, so we actually have a publication, Val’s the one who did all the research on this, but in our little newsletter or publication is called the Focus. So if you go to thefocus.factsandfeelings.io one, it’s kind of gotten me back into medium and actually reading more medium stuff consistently. But I kind of actually think we’re putting some good thoughtful content out there, being sort of aggressively keeping it to a 5-minute reader or less. So I would say check that out.

0:54:31.5 JD: I gotta say I like the stuff I’m seeing. I like the one that you put out about is there too much data, can we get too much data? Which is a good counterpoint to like what we were talking about. So.

0:54:41.8 JH: Yeah, definitely. We working on that while thinking about this one. Yeah, I’m like go get more data. I’m like, but you did start with a business problem. Yeah.

0:54:52.6 JD: Yeah. But I really liked it. ’cause that is a problem. There’s too much data if you, it’s too much noise if you’re not focused. So, I thought It was a good read.

0:55:00.1 TW: Awesome. Look at that. Even a plug from the guest. And now I’m gonna request the bonus last call ’cause Val kept saying that. Jason, you actually have a band.

0:55:09.2 TW: Yeah. Oh that’s so funny. Yeah, I play bass in a indie pop band that’s Brooklyn based and it’s all originals. You gotta have a hobby beyond data and family and work.

0:55:22.5 TW: You got stuff on Spotify.

0:55:24.5 JD: There is. It’s okay. So this is funny. So, the band name is Kill Devvils but two Vs. And you would think it’s like a metal band by that name, but it’s not. It’s sort of, somebody said that we sound like the Strokes. I thought maybe they thought we were having a stroke, but they said we sound like the Strokes. And somebody else said, we sound like Modest Mouse, which I thought was pretty, pretty good. Awesome.

0:55:42.0 JD: I thought you were gonna somehow do Val bringing up the office Olympics and that we were once on a group at UBS called The Fun Police. And I would recommend not doing an office Olympics with competitive finance people ’cause they will cheat.

0:55:56.2 S?: Oh my God.

0:55:57.2 JD: Even the most senior people.

0:56:01.9 S?: There were tears. There were tears. But Julie, just think about it, synchronized, swivel chair dancing was one of the events. And there was choreography. There were props. There were prizes. It was a half day event.

0:56:14.1 S?: I can only wonder which one of you two that thought of that event.

0:56:17.8 S?: Val.

0:56:18.5 TW: It’s funny ’cause I was like, ’cause Val was like, I want to do this fun thing with the newcomer. She’s like, but we’re gonna be the fun Gestapo ’cause I’m gonna make people be, have fun. It’s killing me.

0:56:28.1 JD: Val’s choreography though, for the swivel chair, rhythmic gymnastics thing. I think that won.

0:56:33.8 S?: Oh my god.

0:56:35.3 S?: It did.

0:56:35.4 S?: I would kill to see that.

0:56:35.5 JD: Think about this. So, we had paper airplane throwing and we never specified. This is how insane that we didn’t never specified that it had to be a paper airplane. So, somebody decided to go last and crumple up it as a ball and just throw it across the room. And then there was just arguments about whether that counted or not.

0:56:52.0 S?: Oh my God.

0:56:53.0 S?: The cheating was just, oh my God, I was unprepared.

0:56:57.2 JD: I’m sorry.

0:57:00.6 TW: Wow.

0:57:00.6 S?: They’re like, we know our way around, you know, following the letter of the law and you didn’t say, there’s a big loophole here.

0:57:06.0 TW: I’m gonna drive her right through it.

0:57:07.0 JD: Get a bunch of creative competitive people together. And that’s what happens.

0:57:10.7 S?: Oh my gosh, hilarious. Never forget.

0:57:14.0 JH: Well, on that note, I think that’s like a bonus, bonus last call. So, perfect. So, flipping back to the getting together with people in-person and chatting about stuff, I am gonna throw a reminder out there that Julie and Val and myself and Michael Helbling will be at the Marketing Analytics Summit on June 4th through 7th in Phoenix. So, we’re gonna be recording a episode in front of a live audience. So, if you want to hang out with us or hang out with other, analytics types, check that out. Marketinganalyticssummit.com. And if you use the discount code APH20 you’ll get yourself a nice little discount on the, registration. So, if you can’t make it to that, we’re available virtually digitally as well on the measure Slack on LinkedIn. We’re easy to find, love to hear from listeners. So, reach on out to us.

0:58:07.7 TW: We’ll also wanna, no show would be complete without thanking our very, very soon to be no longer single producer Josh Crowhurst. So Josh, congratulations. Just a smidge early and I’ll throw in a plug for once again, keep our fulfillment warehouse busy. If you would like an analytics Power Hour sticker. You can go to bitly/aph_stickers. They have not threatened to unionize yet, so we throw enough work at them, maybe they will. And then we’ll have a labor relations episode to talk about. So, with that, Jason, thanks so much for coming on. This was a super fascinating discussion. So, glad you were able to jump through the various hoops required to pull it off.

0:58:52.1 S?: We appreciate you.

0:58:53.1 JD: Yeah, thrilled, thrilled.

0:58:53.8 S?: Yes. It was so great.

0:58:55.2 JD: I’m really happy to be here. And yeah, thanks for inviting me.

0:58:58.4 TW: And you dear listeners, whether you are disassembling vehicles or whether you are installing alternative web browsers, however you’re crunching the data, trying to answer business questions, just never forget to keep analyzing.


0:59:16.2 TW: Thanks for listening. Let’s keep the conversation going with your comments, suggestions, and questions on Twitter at @analyticshour, on the web @analyticshour.io, our LinkedIn group and the Measure Chat Slack group. Music for the podcast by Josh Crowhurst.

0:59:35.9 Charles Barkley: So, smart guys want to fit in. So, they made up a term called analytics. Analytics don’t work.

0:59:41.6 Kamala Harris: I love Venn diagrams. It’s just something about those three circles and the analysis about where there is the intersection, right?

0:59:51.6 TW: Now, you’re ready.

0:59:52.7 S?: That’s good to hear.

0:59:53.4 S?: Did you notice, Julie, that you’re our rock flagger?

0:59:56.8 JH: No.

0:59:57.3 S?: Okay. Just putting in the back of your mind. So you can think about that.

1:00:01.0 S?: I’m so used to when Tim’s on that, it’s automatically Tim…

[overlapping conversation]

1:00:05.7 TW: Well, since I always did it, I was like, can we please tell him? Like, no, no, no, it’s great to not. And then as soon as somebody else had to do it, they’re like, oh my God, that is super awkward. I’m like, you’ve made me do it for 150 fucking episodes.

1:00:24.8 TW: You know the old saying, the only tool you have is a hammer. It’s tempting to treat yourself. I’m gonna start over.

1:00:32.9 S?: Treat yourself.

1:00:35.6 TW: Treat yourself like a nail.

1:00:42.2 S?: Tim’s reading it now. Sorry. Should we pause for you to read Tim?

1:00:44.9 TW: Sorry. Well see. Yeah, hold on, hold on, we’ll cut this out later.

1:00:47.5 S?: It does say a four-minute read, but come on.

1:00:49.5 TW: You’re not gonna inspect that chart in four minutes. You gotta zoom into that sucker. Requires social skills. Oh crap. That’s one of the mentions. I’m doomed. Okay. Find everything on the low choir.

1:01:03.6 TW: Rock flag and interesting.

1:01:08.8 JH: Doesn’t cut it.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Have an Idea for an Upcoming Episode?

Recent Episodes

Three pay phones mounted on a wall next to each other

#245: Dear APH-y – An Analytics Advice Call-In Show

https://media.blubrry.com/the_digital_analytics_power/traffic.libsyn.com/analyticshour/APH_-_Episode_245_-_Dear_APH-y_-_An_Analytics_Advice_Call-In_Show.mp3Podcast: Download | EmbedSubscribe: RSSTweetShareShareEmail0 Shares