Cream City Calculation
Three friends talking about data and how it impacts our lives and the lives of others.
Cream City Calculation
"Tap" Into Your Data
Podcast Description:
In this episode of Cream City Calculations, Colleen, Frankie, and Sal explore the fascinating intersection of beer and data. They discuss how breweries are using data-driven insights to improve business operations, capture market share, and enhance customer satisfaction. The episode also touches on Tesla’s hidden autopilot data, privacy concerns in AI, and industry trends like the rise of non-alcoholic beers. Whether you’re a beer enthusiast or a data aficionado, this episode offers a refreshing take on how analytics are shaping the brewing industry. Sponsored by Continuus Technologies.
Welcome to the Cream City Calculations Podcast. We're three colleagues and friends that love data and to talk about how data is impacting our lives. I'm Colleen. I'm Frankie. And I'm Sal.
Sal:Welcome to our fifth podcast. I think it's a great time to crack a beer. Today's podcast is on beer and analytics,
Frankie:Welcome to the Data Pulse, your quick hit source for this month's most impactful data news. The first topic that we wanted to bring up is an article about Tesla from the Wall Street Journal, and it's the hidden autopilot data that reveals why Tesla's crash. And so this is really interesting, because, you know, autopilot's kind of a controversial thing, and there's a lot a lot of, data that's coming up recently. And so there's this story about a man who is using his autopilot on May 5th, 2021. And driving to work when he came across an overturned truck and trailer and his car did not know how to process that he hit it and he died. and so this information was all submitted to federal regulators and there's been more than 1000 of these crashes submitted since 2021 with all the details of these crashes hidden. Because it's proprietary information for the company. So it's just, it's very tragic, but Tesla argues that the driver was warned to put his hands on the wheel 19 times before the crash, and the car did initiate braking before impact, whereas the family obviously is like, You know, that shouldn't be the case. because Tesla is always so confident in their autopilot and they speak to it as if nothing happens that nothing could ever go wrong. And so they're kind of initiating a false sense of security. Yeah. so it's super tricky.
Colleen:It's surprising, but maybe shouldn't be that there's been more than a thousand of these. What's the threshold at which we say it's no longer proprietary information and you have to provide this information to whatever government agency monitors that?
Frankie:Right. Yeah. And it's kind of hard because the family, they want to have that information too, and they won't give the family any of that information. The family's suing Tesla and the case is going to go to trial in 2025. So I could see why they wouldn't want the family to have that information. But I also, as like, it was my family member. I would want to know exactly what happened and what went wrong.
Colleen:But also, what are you going to do to fix this? I think the article goes on to say that other autonomous vehicles use other technology that Tesla is just saying is too expensive and isn't needed. But it'd be interesting to hear how many of these crashes happen in these other vehicles that have this other technology. I think it was called LIDAR.
Frankie:Yep. Yeah. And, how do they compare to the companies that are using LIDAR? Yeah, but yeah, so the other companies use radar, computer vision and LIDAR, whereas Tesla is just using computer vision. But yeah, it's interesting. And, I've ridden in one of the automated vehicles in the Bay Area. And so that was a very interesting experience. But what I would say is that. These, um, I think they're they only drive in, the city, they don't leave the city. And so, they're probably maxing out at like, 30 miles per hour. What I would say is that in that kind of situation. I felt safer getting in that vehicle than getting in an Uber. Yeah. Because the Uber drivers go so fast and they're like weaving through traffic. Whereas the Waymo is just doing its thing and it, you know, it's going 25 and following all the rules. And you might get somewhere faster in an Uber. But I honestly felt better, like, getting in the Waymo.
Colleen:Yeah, I think we've talked about this on the podcast before. If not, it was a side conversation we had outside the podcast, but we've talked before about maybe that there are certain situations where we would feel more comfortable getting in an autonomous vehicle. And When you're out in the country where it sounds like may, I mean, I don't know where this accident happened, where this family is suing Tesla, but there's sort of more things that you can't account for. Whereas in the city, you may be able to control more of that, especially if your vehicles only travel certain paths. And then
Frankie:the other thing that I thought was really interesting is they brought on a machine learning expert, and he was talking about how the model is only trained every so often, so the company doesn't train the model on all of your driving patterns. It's when the company wants to retrain the model that they retrain it, but it's a super expensive process, right? To feed in all that data and train it. And then because of that, the model isn't trained on every possible scenario. It's only what it's trained on. This particular situation, the model had never been trained on an overturned double trailer.
Colleen:Yeah,
Frankie:so it just didn't know what to do. And so when it got to that point, it basically like the model is panicking and just crashing. Yeah,
Colleen:it's that's crazy too, because it seems like that is not outside the realm of what you might expect to see in traffic. if it was some crazy, you know. I don't, you know, combination of animals and vehicles or something in the roadway. You could see how the vehicle could get confused, but anything having to do with the tractor trailer seems it should be relatively common, right?
Frankie:Right, especially if you're driving in Wisconsin on a snowy day or something
Colleen:yeah, or just not
Frankie:frequently.
Colleen:Freeway system at all. Yeah. Our second article we wanted to talk about today was it's titled TalkDesk customer Patagonia sued over its contact center AI. This one is not the first case of its kind, but essentially there's been a lawsuit filed in California for against Patagonia for breaking privacy law. Patagonia uses a piece of software called TalkDesk and this TalkDesk what it was doing was using recordings and videos. made when people called their customer service center to sort of train models and, detect how people felt either about the company or about the topic that they called in about. The lawsuit argues that while customers who call Patagonia are told that the conversation may be recorded for quality and training purposes, they're not informed that the information will be accessed by a third party company, in this case, TalkDesk, or how said third party company will use that data. And the person who filed the lawsuit, it's interesting, that they brought up the fact that these communications are being recorded. Essentially can convert to dollars for this company and that people should at least be warned of the fact that more specifically, these recordings are going to be shared with this third party company because TalkDesk is going to use this not just to make their experience at Patagonia better, but make the experience better for all TalkDesk employees. Customers.
Frankie:Sure. And I have a quote that I pulled that was actually very much so related. Personal information is an important currency in the new millennium. The monetary value of personal data is large and still growing and corporate America is moving quickly to profit from the trend. Yeah, and yeah, I think that's so interesting. And we've talked about this kind of. A lot, actually, because it's so important and I had no idea that Patagonia was using a third party and providing data around customers and then they're using that customer information in whatever way that they want.
Colleen:Yeah,
Frankie:yeah.
Colleen:Have you ever called Patagonia? I have not. I was surprised to see that, you know, that there was some sort of notification made to customers like these. These calls may be recorded and used for training purposes. I would have thought that would have been more of a blanket statement that might cover something like this, but they're, they're calling out specifically that they want people to be notified that their data is being shared with a third party company.
Frankie:I
Colleen:don't know.
Frankie:Yeah, and I can see why they would just say, oh yeah, we're just using this for training. Yeah, I mean,
Colleen:in a sense you are, right? You're training the model.
Frankie:Yeah, that's
Colleen:true. Yeah, and again, it's not the only time that this has come up. Some telecommunications provider called Ayer was accused of instructing its employees to violate legal regulations on handling customer complaints. So, I think this is something that companies are going to have to be more. Aware of so that they word their, disclaimers in such a way that covers potential use of an AI.
Frankie:Yeah, yeah, it's going to get interesting. And who knows, once they throw regulation in the mix, like, I mean, this is just California's regulations and how it's impacting them. But I think it'll get really interesting once they do. There's federal regulation.
Colleen:Yeah, for sure.
Frankie:Or if there's, if, if, right,
Colleen:right, if, when.
Frankie:Yeah. And to stay kind of on this same, similar topic, the Wall Street Journal also had an article around the$600 billion digital ad business is hanging on a few words from Google. Yeah. And so Google is abandoning their efforts to eliminate tracking cookies in Chrome. but they're instead going to. offer the customer if they want to allow cookies to be tracked, then they will track them, and if not, they will not. So this is kind of similar to Apple's Ask App Not to Track language that was rolled out, and they had this really eye opening statistic in here that said that in the U. S., users opt out of tracking about 74 percent of the times they encounter the language. And I know, myself, I opt out. Every time I do, yeah, it's really interesting because the companies that are using ads and everything are very concerned about what the language will be. when Google does release this, because they think that the language will impact how customers respond.
Colleen:Yeah. It was surprising to me, though, that not just that Google abandoned this project of getting rid of cookies, but that they're basically the last one, the last browser left to include them. That was surprising to me. I guess I hadn't realized that had made so much progress. I think that's a very popular. Stance a lot of people are taking these days. Like you said, we opt out of everything. So that seems very popular.
Frankie:Yeah, that's an interesting point. And I have been seeing the commercials lately for safari. I don't know if you've seen them too. They're really strange with like the camera flying and I don't know. It really caught my eye because the first time I saw it, I was like, what the heck is this for? And it was for safari and how they're no longer tracking cookies and, things like that. But, I did, I did think that was interesting too, that Google Celeste one and so much of their revenue streams from their ad business. I can't even imagine what kind of impact that would have. Companies aren't going to stop having ads, but they might not be able to charge as much if they're not, you know, as specific to the customer.
Colleen:Yeah, they're not targeted. Yeah, I think the article says that it's expected to bring in 677 billion in annual spending this year. Wow, that's incredible. Yeah.
Sal:Our first article today is actually going to be talking about five ways breweries are using data to improve business and gain market share.
Frankie:So this one is really interesting because I've never worked with beer data before, and so I didn't even know like what types of metrics that Beer companies were looking at or how they were utilizing them to improve their companies. They were talking about how a president of a large brewery, said that they're leveraging data driven insights and breweries are improving their business operations and gaining a significant edge in capturing the market share. They're using consumer insights. They're looking at supply chain. They're improving their quality control. And many other things.
Sal:Yeah, I think it's really interesting. Like, again, I'm in the same boat as you. I've never really worked with any beer information or bear data. I've done a couple of models on like, nearest neighbor of like, hops, like what beer what you would like, but I would never dove into this. And this is actually really interesting. I'm like, How much it's really used in all parts of the business. That's I found that really interesting.
Colleen:Yeah, for sure. It's definitely not something I've thought of before, but then reading this, this article, I was like, Oh yeah, that's, that makes sense. I think one of the other things they brought up too, is just even improving their engagement with their, their customers or their consumers. It's kind of like an, uh, no duh type of situation, but, interesting stuff.
Sal:And honestly, that's a huge thing. Like you just look at how many beers are out there, uh, and all the different flavors, all the different things, but roughly it's the same formula in it. Right. It's hops, barley, you know, water. Water like fermentation, right? Uh, and honestly, that is crazy that you can figure out exactly the detail of like, Oh, this person likes this much alcohol content because it has a bite, or this person likes this much hop. And now we're going to call it a hazy IPA or whatever. It's unfiltered, all those things. I think it's really interesting that they can go down to those levels. And there's those when you think of just beer is beer, right? Yeah.
Frankie:Right. I think I saw, there was a stat somewhere and I can't find where it was, but, I'm pretty sure there was like over 9, 000 beer companies just in the United States, which is crazy. Kind of outrageous when you think about it. How many of those are in Wisconsin? That's a good question too.
Colleen:Yeah, we should have looked that up before today,
Frankie:right? But yeah, I think, you know, that just kind of goes along with your point, Sal, there's so many different companies. And then each of those companies, I mean, how many flavors of beer do they have? there's many options. And so, you know, figuring out What does the customer want? And, you know, like, is it different in different regions? Like, if what we drink in Wisconsin.
Colleen:Yeah.
Frankie:Versus what they might drink in California.
Sal:I would say that I surround myself with a lot of beer drinkers. So some of the things that I've like been kind of privy to is as they both drink beer and they also love the industry of brewing. So we've gone to Germany to figure or to go and go into the Hofbrauhaus to see the different types of beer and the rules that they had in their beer making is like, you cannot change the formulas. at all. It's actually like legally not able to change the formulas where like in the U. S. I think that's why it blows up nationally because like other countries are not producing all different flavor complexes and looking at the insights because they can't, they can't produce multiple different things that are different alcohol contexts because against the law in a lot of cases.
Colleen:And I just had a friend that got back from Switzerland the other day and she was actually disappointed. She was looking forward to trying some beers in that region and come to find out that most of the restaurants and hotels that they ended up at were under contract with one particular brewer. So every single one of them served the same beers at every single place they went to, and she was really sort of disappointed. So I think those kinds of things could affect your ability to get these different beers to your point, Sal.
Sal:Yeah, absolutely.
Colleen:And I think, you know, if you take it one step further and then in the U. S. where we do have more freedom over kind of what, what we put in our beer, you can use that, analytics to take a look at like which ones sell better and where, like to your, your point, Frankie, like California probably sells much different beers than Michigan, for example.
Frankie:Yeah. I think there's a lot of IPAs in California.
Sal:And Sours.
Frankie:And Sours, yes. Yeah.
Sal:One other thing I found kind of interesting was like the quality control. I once toured a, uh, I think they're a Miller supplier or they're one of the brewer supplier, like the large, like manufacturing suppliers. And when you think of hops, you think of the flower or you think of that, like, growing on a vine, you see the picture all the time, right? Every brewery has an image of it. Where, when you're going to quality, you have to actually understand the data that goes into it and making sure that you're not deviating from that product, especially when you're talking about like a Miller light where it has a flavor. Really like that flavor. And if anything is slightly off, they don't like it. And so this company makes them into like, little pellets, so that they can get a really consistent, hop ratio and help with the quality control. A lot of the data that they're collecting, a lot of these breweries are collecting is that production data and now analyzing it in real time and understanding. All right. What are the temperature differences? What are the fermentation rates? Then what are the pH levels? Are we maintaining those so that our customers have a consistent brand? I thought that was really interesting.
Colleen:Yeah
Frankie:then I was thinking too, like, you know, if they're not meeting those expectations and they have to dump out their beer or they have to go through a recall. I wonder like how much product is actually dumped every year. But I couldn't find any data on that.
Sal:They just give it to their employees. They're like, all right, that's the bad beers. You guys take them. They're free.
Frankie:I guess we're in the wrong industry. We should be getting in there. The other thing that they talk about, too, is leveraging predictive analytics. That's another way that breweries are kind of using their data to improve. Gain market share and they talked about like how they're planning their production. They're optimizing their inventory and managing their resources so that they can swiftly move with the change in the market. And I think back to covid, for example, and for a while there, I think. People were running out to the stores to get all their things, including beer, Before the world was gonna shut down and so I think they were able to maneuver and Work with that change in the market Very quickly because of that
Colleen:I will say too, there were some little breweries that did really interesting things to try and reach their, their customers too. I participated in a beer drive through, and a little micro brew brewery downtown, where they set up in their alley. And, they had signs further down the alley and you could decide, I want a four pack of this or, you know, whatever of that. And by the time you got up to the end, they, you know, they'd ask you for your order. You'd hand them cash, they'd hand you beer. And. You'd be on your way. It was pretty, cool thing for them to adjust like that.
Sal:Absolutely. So Colleen, you've worked in manufacturing and in data. Like, I know there's a lot of things that you've probably had to do, like reducing waste or like making sure that things are in stock. And again, I might hopefully not talking for you here, but, Do you see any major correlations between maybe what you did kind of working in manufacturing to this?
Colleen:Yeah, for sure. I mean, a lot of it is sort of the same, right? It's manufacturing of a product, and so all the themes that pertain to making widgets and sprockets are going to pertain to making another product like beer. But I think you have to also take into consider the shelf life of beer as well, right? Like it's a product that you consume, so it's only going to be good on that shelf for so long. So I'm sure that plays into it. as well. So you don't want to overproduce to the point where you've got too much of like, let's say a summer shandy sitting on the shelf that it's going to be there the next year because right, that's usually like a kind of a seasonal product. So you want to make sure that you kind of leverage those predictive analytics to make sure that you're producing beer at quantities that sort of makes sense for the market.
Sal:It is crazy how much it's growing. I think there was a stat that was at home growth is 2%, which is a massive amount of money in the industry. And then, it's going to go up to around, 654 billion in revenue. That is crazy. There's a lot of money to be made in the beer industry, for sure. And there always was. Uh, that's kind of like where we grew up, where this area, like Pabst, right? You, you have Pabst Farms out and you have the Schlitz House, like you, there's so many breweries and bar, or brewery owners that have so much money in our area, old money, which is like really kind of crazy to think that it's continued along this, this path.
Frankie:Yeah, I think it's interesting to like the share of non alcoholic beer. That's, you know, starting to grow as well. So not just at home, traditional alcoholic beer, but we're starting to branch into some different products too, that are also helping them grow in their revenues.
Colleen:And you've got the combination too, right? You don't just have your, Kind of standard stereotypical brewers where they would make the beer and bottle it in a brewery and ship it to like a retail location But you've got at least around here. There's tons of them and in my travels I found many more where you know You've got a brewery that then has a beer garden or some like customer facing retail facility where they can come in and order the beer at the bar as well so that you can go sit in the microbrewery and have one of their beers on top and then take home with you a four pack of whatever else to enjoy later on. So I think that's another kind of aspect of the industry too that's really taken off. We didn't have that traditionally when I was growing up anyway.
Sal:Yeah, absolutely. And like getting back to the data side of it, like I even think these like micro breweries or these even small mom and pop, breweries that are coming out is their social media analytics. They're probably looking at all that information of like, all right, what if I post about a specific type of beer or sour beer or a hazy IPA or whatever, right. I'm curious on like how much they are taking in. Of that information and making actual business decisions based on that data. I can only imagine some of these locations. There's one, you know, in the city where I live, which is pretty great. They're called Ope Brewing. They've got a fantastic lineup of you know, musicians and things that come through there and food trucks that come through there. So I'm sure that's part of it too. Analyzing what their audience also likes in those other realms as well. But there's all sorts of data to analyze, you know, like which of our food trucks sell better, which of our, you know, brews sell better in house versus you know, in our pre packaged sales.
Frankie:Yeah, I would imagine, you know, like they're thinking about all those events, like volleyball too. Yeah. How many pitchers are they selling? Pretty much every team gets a pitcher of beer.
Colleen:Yep, and they just added on, you know, a bunch of fire pits, and like a bocce ball court, and bean bag toss, you know, field? I don't know what you'd call that. There's a little area for bean bag toss games, or cornhole as we call it, but all those things bring different consumers in that may have different taste preferences for their beer.
Frankie:I think, this next one was around brewery industry trends and statistics in 2024. And I thought they had some really good information too, just very general information, like a brew pub. Is a restaurant brewery hybrid and they have to have at least 25 percent of their onsite brewed beer sold and consumed on site and a microbrewery has to sell at least 75 percent of their brewed beer to other retailers. But they often have a small area at their large facilities for on site consumption. And then for those, they can offer food or not offer food, but. I thought that was just interesting because I didn't know the difference between the brew pub, brewery, and microbrewery.
Colleen:Yeah, I probably use those terms interchangeably incorrectly.
Sal:And I've been to all three and I don't know the difference.
Frankie:I think it's very much so around their sales. And so I don't think that the general public would really understand the difference unless you had their sales data.
Colleen:Right. Or care, right? Tasty beer.
Frankie:Yeah, this article, they talked a little bit about how the pandemic also changed the beer industry and how some of the craft breweries haven't been able to come back from it. But once the pandemic started, there is actually a decline by 9 percent in sales for 2020 and then by 2021, they saw an 8 percent recovery and then. So on and so forth. They've been able to recover fully and I think they're growing now But I thought that was interesting to how long it took for them to bounce back and get things back to normal. It was probably partly due to events being canceled and I think about like all the revenue in their, their tub, but like a lot of revenue has to come from those festivals and events like local events that have local breweries and, even for the bigger companies, like any sort of festival that has beer or a beer festival. Baseball game or, you know, we have the Milwaukee Brewers
Colleen:just in the restaurants, right? Like if you were like Molson Coors and I'm sure they took a big hit during the pandemic because they're, you know, I'm sure a certain percentage of their sales comes from somebody sitting down in a restaurant or inside a bar that orders their beer and that was no longer happening.
Sal:Yeah, and that kind of leads to one of the things like they had the staffing shortage and labor shortage for a lot of these breweries and brew pubs. One of the really cool things that within this article that I guess I didn't realize it so much or internalize it until I read this article but around the QR guides for customers and their ordering. Yeah. Now I can't go to a single restaurant or anywhere without having to scan a QR code. And now they're tracking, they're probably tracking all that information and saying, all right, this person typically buys this or whatever. And that helps them with their supply chain, right? Like, all right. Our biggest beers, obviously they can see it by consumption, but they, they probably can look at what types of, additional things that they can add to it.
Colleen:And you know, I'm willing to bet they can track all of your activity on that site as well after you clicked on that QR code, like you can see, okay, all these people are looking at this beer, but they're not ordering it. Like, and then you could then do further analysis to try and figure out like what you might need to change about that beer to get people to order it more often.
Sal:Yeah, and you're only one click away from merchandise, right? Like, oh, all right, I'll just buy this shirt or the brewery, which I always buy shirts at breweries. It's kind of same, same, what my father in law taught me to do.
Colleen:One of my favorite shirts is from the Moab Brewery in Moab, Utah. It's very cool. That's great.
Frankie:What is your favorite brewery that you've ever been to? Oh,
Sal:I would say that you like one of the coolest experiences is the hot brow. So if you ever have a chance to do it, the hot process is really, really cool. Cause we did it during October fest. So that also made it very fun and Great, but made us slightly hungover for Oktoberfest, which is probably a good thing. But I, I definitely would recommend it. And then there's a couple of bars right across, or a couple of pubs right across the street. So like all these major German breweries are right there. And so you can just hop. between them, but it, I will say it's like thick beer. And so like you have one liter and you're like, okay, I'm done. I'm full. It is a leader though.
Colleen:Yeah. I've been to a lot. It's kind of like my rule of thumb when I travel and I love to do road trips that I, I only, I'm good for about a three or four hour stint of driving. And then I, I get bored and I want to get out of the car. So it's a normal thing for me to choose like a little brewery to stop at. When I'm traveling somewhere and recently I've traveled a lot this spring and summer. And I've visited quite a few some that stick out to me. There was some pretty cool ones in Kansas city. When I was there recently curious animal probably had one of the best, like ambiance that I've ever seen. It just was like a really cool, chill hangout space where people could bring their dogs. And then Brewery Imperial they just had some really good beers that I really enjoyed. And then there's my hometown favorite, I've already mentioned them, which is Ope. They just always have a really good rotating list of, of performance, performers that come through there and also a really good list of beers that all have like Wisconsin funny names. One's called like, excuse me, I'd have to look up their other ones to be sure, but yeah, those are probably some of my favorites.
Frankie:Okay, I was just googling trying to figure out the one that I really like. So, Funny enough, I don't drink beer so I'm not the best person to talk to about breweries, but my husband does and we went to a brewery in Spokane, Washington, and I loved it so much because they had coffee and beer. And I thought this is amazing because I was able to get a coffee and he was able to get a couple different beers and try them out and whatever. And it was called Four Eyed Guys Brewing Company. So, if you're ever in Spokane, that was a really good one. And it's just got a really nice vibe. Cool. I think it was a couple of brothers that run it. Sure. So, yeah, it was, it was really good. They've got like the garage door that opens.
Colleen:Four Eyed Guys?
Frankie:Yep. Four Eyed Guys Brewing Company.
Colleen:That's funny because I was at a brewery once in like the Broomfield, Colorado area called Four Noses. And I think it was like a similar thing, like four guys ran it or something.
Sal:I actually, I think I've been to four noses. It's my brother lives right there. So I think it's a pretty good bar.
Colleen:Yeah. I like that one.
Sal:Did you guys read the article when it was, it tells you all the different types of beer? I didn't know there were like rosé beers. Uh, and. Uh, all these like just the market, like you have eight different types of beers, hard seltzer, rosé beer, sour, lagers, hazy IPA, stout, pale ale, and non alcoholic beers. The massive amount of different ranges within here, I thought was incredibly interesting as well. Yeah.
Frankie:Yeah. I don't know what rosé beer is, but I'm very curious. I think they described it as like, they add like a. a very fruity taste to it if I'm remembering right. And so that's what separates it from some of the other beers, but I'd like to try that. I think that sounds interesting. Yeah.
Colleen:I bought a cool gift for one of my brothers one year for Christmas that listed all the, it was like a pictorial graphic thing, really big poster that was like all the different types of beers and sort of how they're related to one another.
Frankie:Oh, that's cool.
Colleen:Not doing a great job explaining it, but I think most of these were in there and like, some of them are very different, right? Like a sour is going to be very different than a stout, but,
Frankie:what's your favorite? My favorite? I do like to have Weissen beers. I also like amber ales. I kind of, I'm a bit eclectic, I guess. But those are probably my top two favorites. What about you, Sal?
Sal:So if I'm at a German place, I'll do Heppenweizen. Uh, or any beer garden in our area. I'll typically do that. But I'm actually, I like a blonde. I don't know if you, yeah, so I like they're lighter, but they're a flavorful for me at least. So I really do like those.
Colleen:Yeah. I will say too, I've had some fruity beers. I don't like beers when they're too sweet, but I've had some that were just the perfect blend of just a little bit of fruitiness in there that I really enjoyed.
Frankie:I'm feeling really bold. I'll probably get a lemonade if I'm at a beer garden. Okay. Otherwise I drink water. I pretty much just drink water, but I do know that I'm never supposed to get IPAs for my husband because he does not like them.
Colleen:Oh, gotcha. Too hoppy. Too bitter. Yeah. What do you guys think about this article about steps to brew better beer? with analytics. I think this was more on the like the brewer side. If I'm not mistaken, this was from a publication called beer and brew or craft beer and brewing. So I think it was meant more for the folks that brew the beer and kind of helping them understand why doing, you know, analytics is so important for them.
Frankie:Yeah, this was interesting because it was a very different perspective and it was more so talking about like some of the software that's available. And so there is the software name. It's really fun. Beer 30 software by the fifth ingredient. And so it's like a grain to glass data management platform. Designed for helping breweries brew more and brew better is what their little slogan is, which is very cutesy but I really love the name beer 30 because sometimes like we'll get together and have like a little cookout. We call it beer 30 when we're just randomly getting together and having beers and whatever. But. I think that they had some really good points in here and they talked about like the four P's for successful data management which were proactive validation, predictive analysis, process improvement, and performance indicators.
Sal:Yeah I liked it because it it hit upon those but it also kind of dove into a little bit more detail like hey that you're use data to understand and make sure that you're producing a consistent Brew, right? And every, so like when you're doing validation tests and stuff like that on, on the different beers, whether or not you like it or not just building that in and having a system that you can say, okay, this time let's change our temperature by X amount and see if that changes our flavor profile. I was like, wow, that's like really in depth and kind of a really cool way to. Start thinking through analytics and data and that aspect of like, all right, just the micro changes and how much that really affects. a beer really now. It's made. Yeah. Yeah. I've never brewed beer before, Sal. Have you? I feel like if there was one, it wasn't. You have. Yeah. It was, a lot of work. It's not a crazy amount. It's just, so you brew beer or you make the mash, I guess they say. And so put in the hops, all the flavors, oranges, if you want for summer beer, and then you boil it. And so it's like making tea. And then you add in the yeast and that yeast eats the sugar and the sugar then turns into or that yeast produces alcohol. And then, but you bottle it in this container that it pretty much farts out. Yeah, it does. That's exactly what it's doing. And what it does is it, so it's constantly, you just constantly smell beer all day because beer. Cause they like, you put it in a garage or somewhere. Cause if you put it in your house, your house is going to smell like beer nonstop.
Colleen:You smell like beer farts.
Sal:And then you got to bottle it. And when you don't realize how much, I just did a five gallons. Yeah. That's. A lot of bottles. And I like each one manually. It was like, no, no, thank you. Where's the automation here.
Frankie:Can't just put it into like gallon jugs.
Sal:Are you going to drink it out of a gallon?
Colleen:Right.
Frankie:Maybe.
Colleen:I like the section here in here on process improvement because they basically say, you know, basically process improvement is increasing brewery efficiency and they talk about using your data to try and figure out how much per, I don't know if it's batch or per, yeah, it's just per batch or per gallon your beer is costing you to see if you're keeping your cost point down to the point that your product will be profitable.
Sal:Yeah, some of the performance indicators. I thought that was cool because that's where my brain goes. Like, what are the things that you can get out of it? So, minutes per brew, total brewed time. So I don't know what BBL means total pack. So how much per beer do you get out of it? And I thought, yeast health, I don't even know how you measure that, but I guess you can measure that. PH consistency, I thought it was really interesting. The ABVs, uh, and the IBUs and color. So matching that, and probably that's where a lot of it's manually put in, I would imagine, uh, in some of those aspects and then, the different costs per barrel was pretty interesting.
Colleen:Cause if you're a guy in your garage and you just want to make a cool beer, it doesn't matter if it costs you hundreds of dollars to make, cause you've made this cool, like one of a kind beer. But if you're a microbrewery and you're trying to, you know, turn over product at a point that makes you profitable, it's probably not going to sell well at whatever price point that makes a six pack. You know what I mean?
Sal:Especially at scale. Like as you circulate. You can't, you probably have very thin margins in general. Probably everything that you can get out of all these performance indicators, if you can tune it just perfectly to what you like, I would imagine that it makes it a lot easier to kind of make a profit in the beer industry.
Colleen:Yeah, but I imagine this is the point that makes a difference to those, those brewers that are relatively small, right? Yes.
Frankie:Was just going to add in BBL is brewer barrel and HL is hectoliter.
Colleen:Okay. Very common measurements. Yeah.
Frankie:You guys don't know that? No, I just Googled it. Good. Yeah. Thank you. Thank you.
Sal:Now I'm curious what the difference between a hectare leader and a leader is, but I'll Google that later.
Frankie:And when I Google BBL, you're going to want to put BBL in brewing because the first thing that came up was Brazilian butt lift. I was like, Oh, that's not it.
Colleen:Lizards, anybody? Oh, funny.
Frankie:The other thing that I did want to touch on too is the proactive validation I want to dive into that just a little bit, but I thought it was interesting because, they gave a really good explanation of why you would do that and how you would do it. And so the why is. If you're proactively validating and checking over to make sure That things are looking good. You're recognizing issues very quickly compared to, you know, maybe you don't catch something. And so the way that they were recommending doing that was by, like, setting target tolerances for critical brew specs. So some of those performance indicators, you might be measuring those. And keeping them in a dashboard or something so that you're able to quickly see them and maybe setting notifications on them so that you're getting a notification when it reaches a certain threshold. But I thought that was really interesting just to bring up for anybody that's listening that actually works in beer data and thinking about how to improve it.
Sal:So this is just screaming. I have to bring it up because my brother is a food scientist and actually a class at Madison is beer making for his food scientist degree. Like, that's kind of a cool class, but he's been screaming at the top of his lungs for food industries and different to use analytics for this process because he's in R and D. And so when they're testing, it's these really micro movements of maybe temperature changes and stuff. Okay. Again, he's not in beer, but he's in other things. But I think it's really interesting how, like, even the food scientists, and people are asking for more data so that they can have a better understanding of their beer or of their foods, his case.
Colleen:I was just thinking as we're talking through some of these issues and things that you just said, Sal, I'm thinking a lot of these same issues probably pertain to other uh, like food manufacturing concerns, right? So if you have, I don't even know if you're making tortilla chips or something, you probably are concerned about a lot of the same things that we've been discussing here. I don't know why I chose tortilla chips. That's probably not a great example, but hopefully the point there is that when you're manufacturing things, I think a lot of these data elements or these data concerns are the same, in that you want to control the quality and you want to control the price overhead and you want to make sure that you're producing enough to stock on the shelves, but not so much that you have more product than what will sell.
Sal:Yeah, and again, you're working with Agriculture, right? As well. And so like you're trying to dissipate. All right. Is it a drought year? Because like, yeah, am I going to have the supply of hops that I need? I think that's really a scary thought for these companies and these breweries. So the more insight that they can gain, and understand drought issues understand what kind of issues. Yeah. Yeah.
Frankie:Yeah, I always think it's interesting the industries that rely on the weather. Like insurance, that was always interesting. Just, I mean, one of the actuaries jobs was literally to watch the radars and come up with predictive analytics around how much that storm was going to affect the insurance claims.
Colleen:Which is crazy when you think about how imperfect the science of predicting the weather is. Yeah. You can get it right. Yeah, yeah.
Frankie:So I just want to give a quick overview of what this next article was. It's worldwide brewing company turns to data science to tap into customer satisfaction and then I'll give a quick description and then we can kind of dive into it a little bit. So the use case is that a large brewing company with a market share of more than 44 percent of the U. S. beer industry Had a problem because they were trying to understand the pain point of their customers across the world But they didn't have any data on it so what their solution was was to send out a customer survey and then have questions around their logistics division to see how customers felt their delivery drivers were doing. And then if they had unsatisfactory results, they would go and visit that customer and then, understand and address the issue. And so, in doing this, I mean, there's a lot of challenges and 1 being that it's incredibly manual. You have to design the survey. You have to send out the survey. You have to get responses and. They're not, they weren't getting enough responses, but they ended up using this information to, to create a machine learning model. And, they only had 15, 000 rows of data, which was less than 10 percent of the, companies responding. And then they overall ended up taking, that information and figuring out that bad scores were comprised of off day deliveries, missed time windows and warehouse breakage. So that's the summary, I have thoughts and I have my feelings around the amount of data they used was only 10, was less than 10%. I also think about survey responses too. And when I would respond to survey, it's usually either really good or really bad. So you're not capturing probably that full mindset.
Sal:I always heard that if, if you're building a survey and you're going to run any kind of machine learning on it, know that your models it on the top end of it is 30 percent right. Yeah. Like it, there's just too much human error there. And that's, that was kind of my thing is like, but again, the question here is how different. Okay. How much of that 15, 000, surveys or rows of data, how much different is that from the universe or the population? Right. And so like, I hope that they went that far to kind of look at that, and, and help them out, but it does seem like it worked for them.
Frankie:Yeah, right. I do. I know, like, in college, the classes they teach that you really don't need that much data to have an accurate analysis, as long as you have a full understanding of the population, but it is really hard to know if you have that full understanding of the population or full representation, I should say. Right. Because it would almost be like if you're picking, if you're picking 10 percent of the United States, do you have enough people in every state to fully represent the United States? That's kind of like the thought process that I was kind of going through.
Colleen:But also if you consider like the open and reply rates to emails, for example, a response rate of 10 percent is really good.
Sal:And by the way, we, we pick a president on that.
Frankie:Yeah, we do.
Colleen:Or we predict, you know, we, we come decide who's winning in the polls. So, you know, quote, unquote, based on that very small amount of data. Yeah. Yep. Yep. So yeah, that's definitely a downfall. Can you maybe reiterate which points they found were the biggest issues? I remember something with warehousing and delivery windows.
Frankie:Yeah, off day deliveries, missed time windows, and warehouse breakage.
Colleen:So, you know, it just begs the question, how much of that is under the brewery's control and how much isn't? Yeah. Any tiny little thing they can do, even those like delivery windows, if they can even improve that by probably a small percentage, they probably see not only an uptick in sales, but also an uptick in kind of positive messaging on social media.
Sal:Yeah. I wonder if they could learn from how UPS or Amazon delivers, I believe I heard like. UPS they optimize their routes to take only right turns cause you can go on a right turn. And stuff like that. I wonder if like brewing, brewing companies can, can start optimizing a route. And then, yeah, I don't know how much control they have on the distributor, right? Like, yeah, it's usually a separate company.
Colleen:Right, right. But warehouse breakage, I mean, does that come down to employee training? Yeah. And that probably, yeah, or, or, you know, think about where that breakage occurs, you know, it's been zero days since our last incident, like, it's probably not great for worker morale either if that's such a big concern.
Frankie:Sure. And I think too, it really hits home on that point of like, there is a lot that goes into manufacturing beer and brewing beer. it's not just, you know, putting it together. A lot of it's distributing it and getting it out there. And I just thought, like, this is such a small thing. Yeah. Just getting it to the distributor. Yeah. It must have had a big enough impact if a company that large wanted to analyze it and go and visit those customers and make it. Yeah. Right.
Colleen:Yeah. But I guess too, like if it's like, if it's one of the larger brewers, something like warehouse breakage is probably like a bigger percentage of their income than for a smaller one might be. Yeah. Well, good stuff guys.
Frankie:Yeah, so that's a wrap on today's episode around beer and data. The purpose of this discussion was just to have a little bit of fun. It's summer and we're on our fifth episode and wanted to talk about something Wisconsin ish, so we picked beer. Thank you again to our sponsor, Continuous Technologies, for providing us use of their space and technology. And if you loved today's episode, be sure to subscribe to stay up to date on other topics related to data. Next time, we'll be talking about something really fun, politics and data. With the election coming up, we thought it'd be a fun topic.
Colleen:A pertinent topic. Yes, yeah. So thanks for listening to Cream City Calculations, and until next time, keep calculating.