005 – Jason Krantz (Dir. of Biz Analytics/Insights, Weil-McClain) on centering analytics around internal customers


005 – Jason Krantz (Dir. of Biz Analytics/Insights, Weil-McClain) on centering analytics around internal customers

 
 
00:00 / 00:39:57
 
1X

Jason Krantz is the Director of Business Analytics & Insights for the 135-year old company, Weil McLain and Marley Engineered Products. While the company is responsible for helping keeping homes and businesses warm, Jason is responsible for the creation and growth of analytical capabilities at Weil McLain, and was recognized in 2017 as a “Top 40 Under 40” in the HVAC industry. I'm not surprised given his posts on LinkedIn; Jason seems very focused on satisfying his internal customers and ensuring that there is practical  business value anchoring their analytics initiatives. We talked about:

  • How Jason’s team keeps their data accessible and relevant to the issue they need to solve for their customer.
  • How Jason strives to keep the information simple and clean for the customer.
  • How does Jason help drive analytics in a company culture with a lot of legacy (from its people to its parts)
  • The importance of focusing on context
  • How Jason drives his team to be business partners, and not report generators

Resources and Links: 

Jason Krantz on LinkedIn

Quotes from Jason Krantz: 

"You realize that small quick wins are very effective because, at its core, it’s really important to get executive buy-in."

"I’m a huge fan of simplicity. As analytics pros, we could very easily make very complex, very intricate models, and just, 'Oh, look at how smart we are.' It doesn’t help our customers. …we only use about two or three different visual types and we use mostly the exact same visual set-up. I can train a sales rep for probably five minutes on all of our reporting because if you understand one, you’re going to understand everything. That gets to the theme again of just simplicity. Don’t over complicate, keep it simple, keep it clean.”

"…To get buy-in, you really got to have your business case, even to your internal customers, really dialed in. If you just bring them a bunch of crap, that’s how you’re going to lose credibility. They’re going to be like, “I don’t have the time to waste with you,” even though we’re trying to be helpful.”

"What my team and I do is we really help companies weaponize their data assets."

Episode Transcript

Brian: Jason, are you there?

Jason: I’m here my friend.

Brian: Sweet. How’s it going?

Jason: It’s going very well today. How’s your Friday going?

Brian: I’m doing awesome. We’re going to talk a little bit about analytics. Is it Wile McLain or Weil McLain?

Jason: I say Weil McLain. If I’ve been saying it wrong, I’ve been saying it wrong for a while.

Brian: As I recall from my musical training, I think in German, the second syllable is the one that says its name. I guess it would be Wile McLain, like if it was W-I-E-L it would be ‘Weil.’ But I don’t know. Its anglicized as they come over the pond.

Jason: I’m going to go with you on that when you sound like an expert.

Brian: Nice. Well, you sound like an expert in analytics at Weil McLain. Tell us about what you’re doing over there. We met on LinkedIn, I’ve been enjoying your postings on the social feed about your approach. You seem really passionate about what you’re doing and I’m like, “I don’t know who this guy is, but that was really interesting.” I just have. Tell us about the company, what they do. I know they’re in heating, right?

Jason: Yes, absolutely. The company I work for, and I work in the HVAC space, we’re a 135-year-old boiler manufacturer. Whether you realize it or not, you probably have one of our products in your house or building or very close to where you live. What my team and I do is we really help companies weaponize their data assets. As you know, a lot of companies are very skilled at acquiring data since the Big Data Movement. But the reality is that a lot of these companies don’t know what to do with all this data. That’s where we really come in.

What I always tell my team and our business partners that we work with internally and externally is that our focus is on solving business problems. In order to do that, you have to identify what is the business problem that you’re trying to solve or strategic agenda that you’re trying to address. In order to do that, you really have to be anchored in the biz. Again, that’s just my perspective, but if you’re in the business day in, day out, you develop this very keen stand of what the business would need to accomplish its objectives.

Just like right now, we are based in the marketing group and it’s a great spot to be. I’m a firm believer that every analytics team should be based in the business for a reason that I just talked about. But what that does being business-first is that gives us a great lens to look at data from. Sometimes analytics people would be IT-centric and they can do a lot of academic work against the data set or different data sets. But the business might look at the output and be like, “Yeah, that doesn’t help us.” We always, always, always start with, “What is the business problem we’re trying to solve or strategy we’re looking to address?” It also helps us when it comes to curating data also. That’s one of our primary response [00:03:21] this too, is to look for different data sets both internal and external that can help us identify strategic opportunities.

It sounds really unsexy, I’m not going to lie. I think some of my LinkedIn post just say that data is boring. It really is. It’s mind-numbing, too, about 85% of my customers. But that’s the important part is understanding what do our customers need and that’s really the lens that we look at this through. We are a service provider, our customers are internal and external, we have customers just like any other business. We have to take this really boring, but really potent product in data and make it accessible to them. That’s really where we use design to really try to make that magic happen.

Brian: I love that you said, “Trying to understand what the problem is.” This is something we talk about on the mailing list quite a bit. In fact, falling in love with the problem is a good basis for doing good work instead of kind of jumping to solutions or feeling […]. As I tell my clients sometimes like, “Our job is not to go and visualize the data. It’s not […] available for someone to put into another tool or whatever the heck it is. The job is to find an insight that already is used. Probably they’re already in your job and you’re there to make […] if you’re doing internal analytics. Help them do a better job at what they’re doing, offer more value. You need to figure out how to work that into their life.” For example, for you guys then, your customer, I assume is it primarily sales people that you’re working with? Who are your customers and your […]?

Jason: Yes. Great question. One of our biggest customers is sales. Sales has been one of my biggest customers for the past 10 years of my career. I’m very intimately involved with the sales team, sales operation, sales optimization, insight gathering, pricing, things like that, but also marketing. We do a lot in terms of competitive intelligence gathering, market research. We also do a lot of operations in finance obviously related to the prices, that sort of thing. We really touch all areas of the business, but without question, our biggest customers are going to be sales and marketing.

Brian: If you were to bring a new initiative like, “Hey, we have access to…” I don’t know what it might be but for you maybe your point, [might be a line 00:05:49] of data that could actually give them more leverage. We know what the negotiation brings, better than […], we know we kind of have an idea now from what the industry is doing for their sales such that we can now tell the CRM like, “This is your […] or something.” When do you get that sales person involved? Do you deliver a solution and get feedback? Do you bring [...] early and say, “Hey, we think we can tell you more about how to do better pricing on the spot with this thing.” Do you bring them in or when do they fit into your process?

Jason: Great question. A lot of times because we spend so much time actually in the trenches, that’s one of things I think is unique about the way that I design my teams to do analytics. It’s not like hand off product and we’re like, “Godspeed. Good luck.” Once we deliver a solution, we’re actually in the trenches with the business trying to implement what we’re talking about because it just works better. The team work is just more effective and they know that they’ve got back up, they know they’ve got air support.

Really, a lot of times when we come up with something new, a lot of times we will frame it from the lens like, “Hey, we know that we’ve got opportunity A or issue B, or whatever it is. This has been an issue or an opportunity for months or years or whatever.” We think that we’ve identified something that could help us in solving that issue or realizing the potential of that opportunity and then it becomes, “Okay, let’s sit down and talk about, do you agree that this might actually help us in this process?” Because the one thing that I’ve learned is, in order to get buy-in, you really, really got to have your business case, even to your internal customers, really dialed in.

If you just bring them a bunch of crap, that’s how you’re going to lose credibility. They’re going to be like, “I don’t have the time to waste with you,” even though we’re trying to be helpful. What we found out is if you really dial in what are we trying to address with this, just as you would with any business case, and you bring that to them, I have found that they tend to be much more receptive. It’s not to say there’s not going to be resistance—resistance comes with any change—but we found that typically framing it from that lens and saying, “We’re trying to solve a problem that you have, we think that this data will help,” that’s a great starting point.

Brian: Do you have an example of a before/after with that? I don’t want you to get into proprietary stuff you can’t talk about but is there like a, “Before they did it this way,” and then we brought them in and said, “Hey, we think we can get […].” and how you went [00:08:19].

Jason: Yeah. What I can talk about is just the manner in which we distribute sales information, specifically insights. I think that, for your listeners, this is going to ring true to a lot of sales forces. I know for all them that I’ve been in or worked with, this case was true 100% of the time. But one of the things that, again, keeping the customer-centric focus, that if you look at your sales reps, a lot of time is you’re going to be what I call casual data consumers. By that, I mean that these are guys and gals that aren’t really into data day in and day out like guys like you and I or some of the listeners maybe. What we have to do is, as I always encourage my team to take empathetic lens and look at, “Okay, if we give them what our first […] is going to look like, how are they going to interpret this?” A lot of times, to be honest, it’s not very good. Now that’s where we have to look at internally and kind of rationalize and say, “Okay, let’s find this. One of us will find [00:09:14].”

But one example of that is traditionally, sales reps and sales teams will get the information in a flat Excel table. Just lots of rows and columns and just gibberish everywhere. That’s a very financial-centric view of sales data. But the reality is—I don’t know about the rest of mankind but I know for myself—I can’t remember much more than 10 numbers. The mental computational cost of extracting insights is just gargantuan. What happens is, I just don’t even bother to do it. I’m just like, “Yeah, whatever.”

An equivalent of that is, you know if you get a big block of text in email? Even though if you took that same block of text and broke it up into two paragraphs or two sentence segments which is very easy to read when you put the effort in, but for me, if I get a big block of text, I’m not even going to read that. It’s kind of one of the same things that we see on the sales side. What we do is just say, “You know what? There’s a lot of really good information here and we need to make it digestible for our customers.” That’s where we found traditionally, visualization can be an incredibly effective tool to communicate insights to this casual data audience, to this casual data consumer.

Brian: Do you have to work through the visuals with them? Do they tend to get it the first time? Is it a process of you share, “Here’s a report or here’s some new view on X.” How do you know if the visualization is actually allowing them to pull the insight out of what other [00:10:46] broad data? How do you know they’re actually “getting it”?

Jason: That’s a great question. I’m a huge fan of simplicity. As analytics pros, we could very easily make very complex, very intricate models, and just, “Oh, look at how smart we are.” It doesn’t help our customers. It doesn’t help anything. Really what we do—this is going to get to the theme of simplicity—is we only use about two or three different visual types and we use mostly the exact same visual set-up. Just to kind of frame it, what I’m a big fan of is a simple bar chart. There’s more details attached to it but to the right of the bar chart, we’ll typically put a tabular data set. What we do is, as you think in US at least, we start in that left-hand side of the page or we […]. What we do is we look at the visual real estate. We say, “Our customers are going to start in the left-hand side. We want them to look at the bar chart because it allows them to very rapidly assimilate it at a high-level what’s going on.”

It’s great at communicating at top-level churn very quickly but the trade-off is, is this horrible imprecision. You have no precision at all. What we like to do is then we address that issue by putting a simple table, very clean, very simple table over to the right. What that does is that then provides the precision that the customers are seeing in most financial-centric tables. What we found that does is that we have to train our sales team on one set-up and then that set-up is used virtually universally on all of our solutions.

As an example, I can train a sales rep for probably five minutes on all of our reporting because if you understand one, you’re going to understand everything. That gets to the theme again of just simplicity. Don’t over complicate, keep it simple, keep it clean.

Brian: I think those are good. A lot of times, when I work with engineering clients, they fall in love with consistency. I guess one point to maybe just the contrary of this is that, I think consistency is generally a good rule with design. We want to minimize unnecessary change but at the same time, I would recommend to listeners is to always look at context first, and context should always come in.

Let’s say Jason comes up with report number 12 and they have 11 now or whatever, and it doesn’t feel right for number 11. That’s a place where a designer would probably push for, “Well, no. The 12th one actually needs to be different because it’s not […] 11th and even though it’s not consistent, in this context, we don’t need it to win. This version will deliver the usability and the utility that we’re looking for better than the other 11 will.”

In general, I think it’s smart to not get creative unnecessarily with meaningless ink on the screen like, “Let’s try it this way. Let’s change the color palette. I’m tired of this.” Those are not good reasons for […], you’re just introducing noise and it’s unnecessary. But I like that you guys are thinking about simplicity and trying to reuse templates and not looking at it as a creative tableau. Ironically, people think it’s a creative “design” tool, but at the same time with all those weapons, you have a lot of different weapons you can use in that toolkit and part of that is knowing how to use this. It’s the same thing with Photoshop, a million buttons and all this stuff. The Photoshop doesn’t make you a designer. It’s being aware of your customer’s pain and the problems they need and knowing when to use all those filters and all those different things that it can do. I like that you guys are looking into that simplicity and reusing templates when it’s meaningful to do so.

Jason: You bring up some great points and I 100% agree. My team that’s listening there, they’ll laugh because I beat it in their heads, “Context. Context, context, context.” Both in design as you’ve talked but especially with numbers in general. Like, “If I give you a number, a billion, that doesn’t mean anything, you got to have context.” I’d say the same is true for design just as you articulated. Great point.

Brian: Where does the impetus for “everybody is a data company, everybody wants to do analytics”? But then there’s operationalizing that, there’s getting buy-in, leadership behind it. Where does that come from in your org? Where is the interest in taking a 130-year-old company and getting it to care about this? Where does that come from, your influence and all of that?

Jason: That was driven by our current president because he saw it as part of a digital transformation. Obviously, this was an essential component of that. Obviously, we do a lot with analytics, but we’re also involved in a lot of other digital components that lead to that overall digital maturation. Analytics is a very, very big part of what we do but it’s not all that we do.

We serve as kind of that quarterback for a lot of the digital initiatives to help basically, guide them through the process. Because even though some of the nuances of each of this project, each one will have its own nuances, they all come back to data. Data is the currency. We found out pretty quickly that if you want to stay relevant in this day and age, you need to be digitally evolved but more importantly, as you look at it, do you [compare the 00:16:02] advantage that you can derive from analytics?

I would argue that gap is slowly closing known certain industries like manufacturing, but we probably have a little bit more runway [00:16:10] it. But for a lot of industries, analytics is becoming table stakes. It’s one of those where you can certainly expect incremental value and competitive advantage, but the question becomes how much longer. That was kind of the impetus of saying, “Hey, we got to get this going sooner rather than later.”

Brian: Do you have people in sales that are resistant to using the reporting or taking advantage of your information or is it pretty ingrained in the company culture that it’s like, “This is a tool. Why would you not want to use it?” Or did you guys have a […] getting adoption?

Jason: Yeah. I would say anytime you’re going through a transformation of this magnitude, it’s hard and I would say especially for other manufacturers. I found in general, manufacturing in general, tends to be one of the laggards industry-wise in analytical maturity. Unquestionably, it’s tough for no other reason than change is tough. You’re taking legacy plants, legacy steer pieces, legacy process, and some people has been around the company for decades potentially, and we’re asking them to change almost on a dime on their time scale how they do business.

It’s not that it’s right or wrong but what we try to point out is that, as I always say, we have to acknowledge the past. We’ve been where we’ve been, we’ve been successful at where we been. But there’s been more change in the past two or three years than maybe you’ve seen in the past 15-20 years. In order to stay relevant, you really have to be ready to evolve, not only evolve but evolve quickly. But I have to openly acknowledge that that’s hard. It’s a hard proposition for a lot of people.

Again, it comes down to change management and managing not only expectations but supporting that change. Change doesn’t happen by itself, we have to support that. That’s really what we try to coach through. The way that we try to do that is by developing a product with our customers. I’m sure as you can […], if you force something upon somebody, it doesn’t get received too well. But if you develop it in conjunction with them and do tie it around their needs, it tends to get better adaptation.

Brian: You used the word product in there and I’m interested, do you see the outputs of your efforts? Primarily, it’s BI reporting as I understand it. Do you look at that as the product that you offer to sales? Is that kind of how you see it?

Jason: Yeah. We offer a product in the form of the insight packages but it’s also the service. Service that goes with it where again, we serve as essentially internal consultant to help them along. If you take just the product-centric approach, you just deliver an insight package and you’re like, “Good luck. It’s [00:19:35]. Have at it.” What we do is we deliver the product and then we partner with them and say, “Okay, here’s what we see. Now, remember you’re talking about this going on in the channel last year and our note show that there’s been a lot of competitive activity in this area. Here’s some of the question that we have. You’re the expert, so what do you think?” What we found is that working together like that, we tend to get pretty good results versus just leaving these guys on an island to kind of figure it out themselves because they virtually always know the answer but sometimes it’s up to us using these products and then offering the service is to ask question that maybe aren’t getting asked. A lot of times, we find out that they know the answer it’s just that you kind of have to ask the question.

Brian: Is that often like, “I was using XYZ report. Could you break this down by county instead of just by whatever because I feel there’s more people living in the East side of town and the average is here or […] the whole county. I really just need this one county because that’s where everyone lives. Is that really underserved? Blah, blah, blah,” that kind of stuff and then you guys will go off and work with them for more of that detail then maybe you release that back into the product as a feature if it seems like a one-off or something. Is that how it works?

Jason: It’s actually a very fluid process. An example of what you just described is exactly what happens if hey come to us with questions. But we also do it where we flip it around because a lot of products that we create are more aggregate discussion tools. We don’t design a lot within our primary visualization package. To really get into the weeds and everything just becomes overwhelming. We have other tools like your traditional [00:21:22] pivot table to kind of dig into that stuff.

But the exact example that you just gave, they will ask us those questions, but we will also flip the script and say, “Hey, we saw that the mechanical chain in the Northeast is up 50%,” I’m just making up a number, “and at a higher level, you can see that but when we segment it out, here’s what we see. Not only when we break this down to this level, we see that’s specifically being driven by A, B and C.” That gets to where I push heavier at my team to do root cause analysis. That’s really where we provide value is by digging into it and asking questions like that. Again, operating from the lens of trying to solve a problem or answer a question or root cause something in conjunction with the business. A lot of times, we will ask those questions and at the same time, they will ask us, which is great. It’s amazing because you get the better solution faster.

Brian: I think that’s great. I’ve worked on several different tools that have varying sophisticated means of doing root cause analysis and I think it’s a really powerful way to bring some why to a what that has happened in the past. Most of the time, why is really where the money is at. The value comes in being able to understand why. A lot of times, we don’t have all the data. You can’t know for sure but a lot of times I tend to say, “Our guess, if they’re just going to make a WAG—a wild ass guess—then our guess, as long as we qualify what ingredients went into the pie, our guess may be better than any WAG.” They’re going to make one already.

If they’re going to make a decision here and go off gut, there is maybe a chance they’re right and their experience will say something. But maybe our elementary root cause analysis, which we can improve over time, will actually be better and we can get out of the total guessing game and start with something that’s kind of a macro ballpark thing. Then overtime, you can improve that analysis as new data becomes available or maybe learn about how two variables are related in the business and you can bring that knowledge into the system.

I totally hear what you’re saying. It’s a nice mix of internal product plus services and also, it sounds like it gets you guys do good discovery work as well. You guys are not just responding to questions but you’re maybe asking them questions together as a group. You kind of work through what opportunities maybe latent that no one’s talking about by asking questions using data to do that.

Jason: Yeah. In the lens that we’ve been talking through, this is really sales-centric, but this applies to any group that we interact with. We have the same level of proactive discussion with any group that we interact with. In some of these, in our market research side, it’s 100% proactive. We’re going out there scouring for information and trying to see the other things that we see. That one it’s completely proactive and now we bring insights to the business and say, “What do you guys think?” The sales one is the most fun because, let’s be honest, there’s no business if you’re not selling anything and nothing happens until a sale is made.

Brian: Right. I get that. You talked about other clients, do you work at all with the actual hardware, is there any IoT type of analytics going on with the boilers and machinery that you guys create?

Jason: We’re early in that process. We actually are getting ready to go down that task very soon. On the hardware side, we tend to not have as much involvement. That’s really more on the engineering group. I think for any manufacturer product or engineering groups probably going to be the most involved in that. But obviously, we get involved into the discussions of answering the fundamental question. What are we actually going to do with this data when we collect it? Because as you can imagine, IoT can spit out a lot of data real quick. They can become incredibly burdensome very quickly if you don’t have a plan on how to manage it. But then, if you’re going to go through the effort of managing that, you got to be able to say, “What are we going to do with this?”

Brian: Yeah. I guess the first thing that would come to mind for me would be predictive maintenance, like, “Is it going to break down soon?” I worked on a cooling company that does cooling and really as the guy told me is like, “We’re not selling refrigeration. We’re selling consistent temperature to our clients. It’s not really about coolers and all of that, so we need to deliver consistent temperature. If we don’t do that, they lose products, they can lose whatever is being stored in cold storage.” That is significant business. I’m sure for you guys, it’s heat, you want to sell heat so how do you get in front if there’s a maintenance plan or whatever, how do you stay on top of that kind of stuff?

Jason: Absolutely.

Brian: [00:26:13] IoT. One of my clients used this word one time, which I now use all the time which is like, “We don’t want a metrics toilet.” An example of you can get to a metrics toilet really quickly with every stat under the gun and how many ounces of water per minute through this pipe, that’s great because that’ll help me do, as a sales guy or as a technician, how am I going to use that information just because there’s a sensor on that pipe. It’s working something around like, “Oh, there’s a sensor. Put the data in the grid.”

Jason: I’m going to have to borrow that. I’ll give credit whenever I use ‘metrics toilet,’ that’s a pretty good one. I may actually [00:26:56].

Brian: Nice. Tell me, where does it go from here? You had mentioned like, “Oh, the competitive edge, maybe it’s closing.” Or maybe you guys feel your competitors are all kind of maybe they’re doing the same thing that you guys are doing and we are all aware of where the data can be used to drive the business. Are there other places where you see design or technology like predictive analytics or machine learning and some of these other new technologies that are out there to help drive predictions and things like that? Are you guys leveraging any of that or have plans to look to the future? What does that look like? I know you probably can’t talk about everything but maybe broadly.

Jason: Absolutely. I would say that that’s content that’s definitely, if it’s not already being done then it’s on our radar. We’ve got a pretty talented team here that goes a lot of your traditional data science turf. As you can probably surmise in this conversation, is in addition to having all skills, we’re probably the most heavily focused on the business side. As we say, we explore opportunities for a lot of this. We always look at it, again, like machine learning. Great, but we got to make sure it’s very powerful stuff. We got to make sure that whatever we’re embarking upon, because we have finite work capacity, if you pursue something, machine learning, it means we’re not doing something else. It’s not to say that it’s not important, but we really have to be able to answer to that question. Again, come back to, “This is our anchor. What are we going to do with it?”

I love this stuff. I love the stats. I love machine learning, AI, all that stuff. If you’re not careful, you can really quickly get into an academic exercise that we think is really cool. “Oh wow, look at this. We’ve got this awesome algorithm here. It does all this magical stuff,” and then the business looks at it and goes, “Yeah, so what? I don’t care. How does that generate revenue? How does that improve our margins? How does that reduce our cost? How does that enable to build the sales pipeline?” If we can’t answer those base questions and we don’t get alignment, that’s probably the most important thing is executive buy-in on exactly what we’re going to be working on, why it’s important. No, we don’t pursue it but those things are most definitely, as with any analytics teams today, I think that that content is definitely being done and/or on your radar.

Brian: You make a really good point. Sometimes I almost hesitate to ask the question. But I think it’s an exciting space in terms of predictive capability and removing viable analysis and what we call time tool time in the design world, there is there. But at the same time, you make a really good point which is again, these are tools that need to be leveraged to service an opportunity or a problem. The goal is not to go do the machine learning, the goal is to solve a business problem by which machine learning maybe applied a better […] do it, reduce cost or reduce effort, speed, something like that. I completely respect that.

I’m glad to hear that you guy are looking that as not a leading step. I know there’s conflicting signals out there. I’ve been talking to people in the International Institute for Analytics about this and at the same time you hear a lot of stuff which is, “If AI is not part of your strategy, you’re going to be missing out,” and boards just want to hear that people are doing AI. At the same time, you’ve got academic exercises going on, you’ve got people trying to take on massive like, “We’re going to shoot for the moon,” and it’s like, “You don’t even have an airplane and you’re trying to go to the moon with this thing. Show us a small win if you’re going to do an investment in AI.” It’s okay to go try it out and say, “Let’s do a small thing but let’s try to solve a business problem or have some definable output that we’re looking to do here such that we’re not just writing code and doing experiments.”

I hear there’s a problem with people putting this on their resume. It’s like people just want to have machine learning. Everyone’s a data scientist now that used to stay in analytics. [00:30:47] It’s scary in the sense of just wasting opportunity and wasting money because at some point, your smarter competitors are going to be saying, “This is a new hammer. Let’s find some nails that we can use for it. But we think […] right nails and it needs to be the right application before we whack at it. It’s not just […].”

Jason: I really like your point because again, if my peers were listening to this they will laugh because they say, “We are professionals of this trade and the tools that we might want to use might not be the right tool to use for a specific job.” I couldn’t agree more with that sentiment. It’s one of those core philosophies that I have and share with my team. Also to it is with the AI. I think that you truly made a very astute observation here and comment in that, I think a lot of companies do feel compelled to have to make significant investment in AI like today. It’s not to say that there’s not merit. There clearly is plenty of merit and plenty of potential there, but kind of your point, I really believe that it’s much more beneficial when you really minimize the risk of project and budget flow and minimize overall project risk.

You take that small bite and try a little bit, then try a little bit more. When you get to win, socialize the win, and your executives feel comfortable because I’ve done it on the analytics side. I went for a big bang approach and after nine months they were like, “Hey, man. Where’s the output?” All you need is to get bit by that once and then you realize that small quick wins are very effective because at its core, it’s really important to get executive buy-in. A lot of executives are not willing to wait nine months or a year for something when they’re expecting to see at three months. I totally agree with your sentiment.

Brian: When you talked about the wins, I totally understand if you’re close to it and maybe hard to remember those, but is there a particular story or time where something in the product and the insights that you guys put out to your customers that it was like a real win, like a sales guy said something to you or maybe an executive said something to you about how this moved the needle, like this was a memorable moment for us. Like, “I changed a customer’s mind with this,” or, “We closed the sale that we never would have been looking over here if we didn’t do it.” Do you have any anecdotes like that that you can share?

Jason: One that we had recently, again, just for confidentiality purposes I can’t get too deep.

Brian: Sure.

Jason: We did have one recently where we just basically revamped our insights packages that we distribute to our internal team. We really, really gathered feedback. We had version one, we gathered ton of feedback, kind of refined, iterated, got the feedback without making it a major release. Got feedback, refined it, refined it, and then what we did was, with a small group, we got that beta in their hand, they look at it and they’re like, “This is great. This is exactly what we need.” Because what we were doing, what we found was—I’m sure you’ve experienced this—everybody wants their own part of things. Everybody wants certain view of a report or they want certain insights or whatever it is, and it’s great. But if you have limited resources, really high-powered resources like an analytics team or data science team, you’re going to look at the opportunity cost of trying to do one of these one-offs, we were getting a ton of report flow.

Again, what I tell my team, I don’t mean to be derogatory to the DI guys in this comment, but my team’s side, I always tell them, “We don’t create value if we’re just creating reports. We create value when we’re actually partnering with business to extract insights, identify opportunities amidst all that stuff that goes well with it.” What we realized though is that, what started out as a nice, clean, three- or four-page insights package and blow it up to like 20 and [34:21] doesn’t that meet our original criteria?

Essentially, what we do is once we have the rationalization enough to say, “Okay, we’ve got all these stuffs right across 20 pages. We can actually distill it down to four pages.” It will give you the exact same information, but it might not look the exact way that you wanted it to look. The question becomes, are you willing to deal with less stuff and maybe have it look a little different, but you’ll get it in a much more concise package that you’re actually able to use and process?

What we found out is that a lot of people were doing these packages and getting the reports that they want but they weren’t actually using them to drive decision-making because they can’t see the paragraph or the block of text story before. They look at it and they’re like, “I don’t know what the hell to do with this.” We would dial that in and it just been a screaming success. It’s really nice to have it where something like that you see the evolution of it. This is just one of those things that we had, and this was kind of a side package or wasn’t a primary, but it’s become a primary now because it’s so effective.

Brian: Would you say that when you talk about reducing this, is the report like a PDF or do they access it through a browser the insights package?

Jason: Yeah, we have the options to do both. We distribute it initially via PDF, sometimes along with our comments if there’s really, really big stuff in there. We’ll say, “Hey, we see this. Here’s a driver. Here’s a supplemental package.” A lot of times it’s PDF first and then if they want to go on the web, start interacting with it, they can do that. Those are nice, but the reality is a lot of them don’t do that which is understandable.

Brian: You took it from 20 pages down to 4, is that what you’re saying?

Jason: Yeah. Same information.

Brian: This is a really good point. I’ve frequently had clients come in and they’re with data products and their concern is information overload. We’ve heard this a lot of times and the irony is that, the issue is usually not information overload. It’s usually a design problem that the information is not presented properly because sometimes, it can increase the density and increase the utility and usability, not the other way around. In fact, removing data can actually make it worse.

A basic example of that is when you’re trying to compare A and B. If A and B are not on the same, what we call a viewport like in a browser world, it would be within the browser window there. When you require someone to toggle between two screens, they have to change context and visually, your eye can process the information a lot better when it’s within proximity. Sometimes, increasing the density actually will give you a better design. It takes more care in how you do it, but it’s not always about information overload, “Oh, it’s too crazy.” They may not get it on the first time but your sales people, if they’re looking at this stuff weekly or monthly, at some point they’re going to be pretty comfortable with this.

I always tell my clients, “You need to look at the switch frequency as well because if it’s going to be used a lot, you can actually get more detailed and you can really push the, what you might see as complexity or the information density, can go up because they’re going to get familiar with the formatting. Typically, the density is actually going to probably improve the utility as long as care is given to the choices. But having that eyeball comparison without having to change pages and all of that, typically you’re going to give a better story as a broad rule. I like hearing that you guys went down in page count, up in density and in turn a better user experience at the end so that’s great.

I think we’re about done here. I don’t have too many questions for you, but this is super great. One of the reasons I contacted Jason is because I remember seeing this quote, “Jason is like a category five hurricane in the data analytics world.” I’m like, “Who the hell is this guy? No one talks like that.” I started reading your stuff and I enjoyed watching your LinkedIn social posts and things like that. Where can people find out more about you? You’re obviously on LinkedIn, I can put LinkedIn in the show notes and stuff, but are you on Twitter, any social media places they can follow you?

Jason: No, actually, I’m not on Twitter. But the best place unquestionably is going to be LinkedIn. I’m pretty involved there. I do like to engage. If you want to direct message me with questions, just talk, meetup, connect, whatever it is, I welcome that. I love the platform, it’s a great family. I just really started using it maybe nine months ago, really getting into it. It’s been great meeting guys like yourself. It’s actually phenomenal.

Brian: Cool. I’ll put a link to Jason’s LinkedIn profile on there and you guys can find him. I recommend, especially if you’re in an internal analytics type of role at your company, to follow Jason and then check out what he has to say on there. This has been great. Thanks for coming on the show. I look forward to meeting you at some point in person.

Jason: Dude, thank you for having me on here. I really appreciate it.


More Episodes of Experiencing Data:

  • Browse past episodes - each episode has a full text transcript and audio
  • Join the mailing list below to be informed when new episodes are released

Subscribe for Podcast Updates

Get updates on new episodes of Experiencing Data plus my occasional insights on design and UX for custom enterprise data products and apps.