Werner - Driving Organizational Change for More Meaningful Insights - Inspire 2017

A premier global transportation and logistics company, Werner Enterprises is among the five largest truckload carriers in the United States. It has a diversified portfolio of transportation services and the company's Werner Logistics subsidiary provides freight management, truck brokerage, intermodal, as well as international services. The company's mission is to deliver these solutions responsibly and safely, and it relies on Alteryx to help do this. In this session, you'll learn how IT and line of business (LOB) decision makers at Werner worked together to get IT away from doing ETL work and focused on larger, more impactful projects, such as their data warehouse. The shift in focus resulted in access to real-time data for more meaningful data-driven driven decisions for the LOB. You'll also hear how the organization began using Alteryx in conjunction with Tableau to address basic questions at first, and then quickly scaled their analytics practice to ask questions that could not be addressed prior to the data enrichment and analyses they conduct within their Alteryx workflows.

Video Transcription

Jeff Walters:
Before I start, I have a couple observations that I wanted to point out that I've seen when in Vegas. The first one is, how many took the shuttle yesterday, to the event? They need Alteryx bad. Literally, to get out of [inaudible 00:00:17] it's a left and it's probably six blocks to get to the Brooklyn Bowl. I think we were on the shuttle for about 25 minutes because they some back ways so they definitely could use some spacial analytics. The second is, how many people have been gambling? I know Jeff Phipps has back there. One of the things I was thinking about, do you think Vegas gets scared when an analytic conference comes to town? Do they think they're going to get wiped out? I'm guessing not, because I think they got a sweetheart deal from Alteryx and they've just been taking our money back and forth so.

With that, I do have a couple notes. I am under the weather. I'd like to thank Ramen for getting me some tea. Last night about 9 o'clock I didn't know if I was going to be able to do this. Jeff Phipps from Alteryx, and my boss, [Ship 00:01:03], can attest, I sounded like something out of Star Wars. My voice is very crackly, so hopefully I don't have the- I can't think of the 1970's show where the kid's voice cracks moment, but away we go.

So a little bit about me. There's a lot on here about what I've done, but really what I want to focus on is what [Roman 00:01:25] talked about, that's my business background. I have no IT or technical skills when it comes to working with data, but I do have business knowledge of Werner Enterprises, over 15 years plus, and that really helped me get into my role. The one thing I also did have is I understood how our data worked and how we needed to mash it together, if you will, to get inside out of it. I know there's a lot up here about what I've done and when I've taken the role over about four years ago, but it's really important to note that I think when you move forward with an organization, you have to have people who are curious and are asking the right questions, and that was something I was able to do to push me into this role.

What is Werner Enterprises? Has anyone ever heard of Werner Enterprises before this conference? Yes. Do you guys use us? I hope you're not... both customers? So Werner Enterprises is a trucking logistics provider. We're based out of Omaha, Nebraska. We're a $2 billion company. We have over 7,200 trucks and we employ over 9,000 drivers annually. We also, on a logistics side, have over 15,000 partner alliance carriers that we work with to help manage the freight. We don't necessarily move the freight all the time, but we can manage freight from an end-to-end solutions. This is a legal thing.

I have to show how awesome we are financially, arrow up. What do we do? Kind of what I talked about the other day. What we're looking to do is to do end-to-end solutions from a transportation management logistics perspective. Whether it [inaudible 00:03:03] or one of the products hauled on our trucks or one of our partner carriers or alliance carriers trucks, we would like to manage that freight from start to finish for our customers. We provide reporting, KPI reporting, we proved real-time updates for our system, we have our own TMS system that we made, that's Transportation Management System, where customers can go in and can see where their freight is currently at the moment.

Some of our initiatives, and this is what I'll be talking a little bit about today. Some of our initiatives are safety, drivers, more efficient, and then also a grow in our portfolio from a logistics. But I want to focus on the first three, because they're really the backbone of our organization, and it really starts with drivers.

At the end of the day, our drivers are what make Werner Enterprises what it is today and what makes us such a great company. Our founder, C. L. Werner, he started with one truck in 1956, and it was a dream, and he's grown it over to a $2 billion company. So it starts with the drivers. But what we want is our drivers to be safe out on the road.

The other day, they're hauling potentially 80,000 pounds of stuff between the weight of the equipment and the product that's on the trailer, so it's imperative that we give all the resources to our drivers to be safe out on the road, not only to protect themselves, but to protect the public, us, et cetera, so at the end of the day there's also a cost. But our number one focus is not to get into accidents and to create less risk for our drivers.

Secondly, the efficiency. Unfortunately, we work in a very razor-thin margin environment. Historically, for truck providers, margins could be anywhere from 5 to 10% and that's really cutting cost, and so we want to be more efficient with the equipment that we have. If you think about it, that ties right back to the driver. If our equipment is broke down and our equipment is not efficient, then that driver is not moving, then he's not happy, and then he's going leave the organization, and then we're also not getting paid for that equipment to move.

A little background on the trucking industry, I can't really give you Werner specific stats, just because it's not public, but basically, in our industry we run at 100% turnover for our drivers. Really what that's stating is that for every driver that we hire, there is a driver leaving out the back door. So imagine, earlier I said there was 9,000 we hire, employ 9,000 drivers, or 9,000 drivers at any current moment, so that means every year we're looking potentially for 9,000 vacancies to fill. As you can imagine, that puts a lot of stress, a lot of downtime on equipment.

And then also there has been increased regulation from a safety perspective. Excuse me, guys. Hopefully, we can mute the mic when I'm like belching to the side. There's been increased safety regulation. We don't always agree with that regulation. Once again, we want our drivers to be safe, but it's really hard for us to say, "Hey, the government should tell us how our drivers to be safe." We definitely have hours that we need to follow, but there's different regulations that they're trying to invoke, if you will, that potentially could be more dangerous for our drivers on when they can drive. So there's a lot of safety regulations that we're also competing with.

So what does that mean? We've been a very descriptive company for the longest time, up until about four years ago. So we started this data journey, and before we start it, I want you guys to read this phrase, because this is... we're still like this in our organization in some parts, but we were very much like this in our organization about four years ago. It's the most dangerous phrase and the language is we've always done it this way. One of the strengths and weaknesses of our organization is our experience. Our executives, our management, and leadership teams all have over 15 to 20 years experience, which is really great because they push the company forward, but also can be a detriment because they do live by this phrase, gut feel. Like, "Hey, this is what we should be doing. We should be giving driver more money, we should be doing this from a safety perspective," because it felt right to do. There wasn't necessarily data to back that up.

What I'm going to be talking about today is just kind of our path through this just to say, "Hey, here's where we are today, but here's where we started." And I ripped this off completely from Gardner, so I have to call them out. But it's really kind of the four-pronged process that people use on where they're at from an analytical perspective within their organization.

The very first one, descriptive analytics. We still live and die by this, unfortunately. What did I do yesterday? What did I do last week? Up down reports et cetera. But we have moved farther and farther away from just being a descriptive analytics company. But up until 2012, this is what we lived and died by.

The other is diagnostic analytics. Really just kind of getting more meaning out of your data in the moment. So, hey, should I give a pay raise to a driver? Is it really going to influence their decision to stay. If I make this program from a safety perspective, is it really going to have the impact that I want it to have?

The next is predictive analytics and we started to dabble on this a little bit using Alteryx and a couple other tools. Can I start to predict when things are going to happen? Once again, very driver focused. Can I start to predict when a driver is going to leave? So if you think about it from your guys' perspective and I don't know who's all in the audience, but if you're a call center, you have the same type of problem from a turnover perspective. If you're managing customers in a high-volume market, you have the same problem, like customers that potentially could leave, et cetera. Same thing. Our drivers are our customers. We have two customers, the end customer that we haul for and then also our drivers, so we want to use predictive analytics to get a better idea. Can we start to predict when these drivers are not happy and they may leave?

Then finally, prescriptive analytics. Which, every time I read the description on it, sounds exactly like predictive analytics. But those are the four phases.

Without further ado, the way we were, we were very descriptive, and there's this awesome photo of a possum and a dog wanting to fight each other. But I always call it possum fighting, because that's really the way we were. Have you guys ever seen a possum fight? It's loud and not a lot of action. That was really business and IT for the longest time. We were loud without a lot of action. A lot of complaining on both sides. You're not telling me what you need. And I'm telling you exactly what you need and you're not delivering it. And then they would come back six months later and say, "Here it is." And I'm like, "What is this? I don't know what this is."

Well, we met like six months ago. Okay, I forgot what the question was, so let's move on. But we were very much like that. Everything was extremely, extremely hard. The end result of that was bad excel, like I'd keep hearing excel hell, it was very bad Excellish-type reporting, that you still had to dig through, because guess what? At the end of the day, I would end up just saying give me a data dump. I will build it myself. Give me a data dump and I'll do it. So we would get these huge spreadsheet data dumps, whatever, and I would have to dig through it and build my own reports to try to get me in an insight, and it sucked. I'm not going to lie. It was terrible.

Other things that I had to do from a business perspective. Let's view our transportation company. You would think the easiest thing in the world would be to find out, hey, how many loads or shipments do I have left to cover? Well, when I was... for one of my customers, I would have to get on a conference call every day at 8:00, it would take me 45 minutes to get that information, so I could be on a conference call. As soon as I got that information, that information was stale. When I had my 2:00 p.m. conference call, I had to do that whole same process again. What we started to think was like, "Man, there's got to be a better way. There has to be a better way."

At the time, I was managing 40 people and I had three analysts on my team. These analysts were doing a lot of the grunt, Excellish-type report work, and then my boss, who is back there, Chip Duden, managed the whole customer service team within our logistics group, and that was roughly 140 to 150 people. Chip, being who he is, decided like, "Hey, Jeff, we need to be better with our data. I think we can do this. I'm going to go pitch something to our COO, CEO, Derek Leathers. I'm going to pitch this idea that, gosh, we need to form an analytics team."

I had no idea what that meant. I'm like, "Yeah, let's just do it. Anything to get me out of being on calls where I'm being yelled at by customers because it was taking me so long to get information and I couldn't tell them what was going on. Yeah, let's just do it."

So Chip goes to Derek, and Derek is a 299-and-a-half pound man, and you would get that. The only person who would get that is Chip. But he's a big man. He's 6' 4", he played football at Princeton, and he's bright, he speaks a lot in Washington, D.C., on the causes for transportation, and he's a very methodical person, well thought out, and very compartmentalized, not emotional. Chip goes to him and said, "Derek, I have this plan, I have this vision. I've been reading a lot on how all these customers, customers of ours, transportation companies are using data to be better. They're predicting when drivers are leaving. They're predicting when customers are leaving. They're helping reduce safety cost."

Derek looked at him and said, "Chip, you know how I love good ideas and the people that bring them and you know the Werner way is whoever brings a good idea must manage the good idea. So go do it. Go do it."

And Chip, I'm guessing, we never had this conversation, I've only had the conversation of what he told me we were going to do, said, "All right, what does that mean, Derek." You, Jeff, and the three analysts will form your own team.

So imagine that. Chip goes from managing 140 people, very secure environment. I go from managing 40 people, and now we're on an island, which we don't even know what we're going to do. We're like, "I know we can do this, but there's got to be a better way." So we started to look for different tools that would help it be easier for us to get the insight that we were looking for.

The first tool that we landed on was Tableau. I don't know how many of you guys in here also use Tableau. So the faithful moment for us is- one of my analysts and I, we went to Washington, D.C., to the conference and we met Alteryx, and I didn't know what it was or anything, didn't really want to go to the meeting, but I said, "Hey, let's do this." And we got this thumb drive, and it kind of changed our life, because it made data so much easier, easier than we ever thought it could be. In the first- and I really wanted to drop that mic- in the first drop-the-mic moment for us as a three-person, four-person analytics team, was our housekeeping dashboard. The first thing that you see up here is the cube that IT was trying to build, because still at the time we were dual railing with the BI team and IT, and then our team. But we knew we had to be quicker and faster in the descriptive analytics in order to move forward.

Housekeeping to you guys is at the end of the month, we have to close out all our shipments, make sure the correct rates are on the shipments, make sure the correct rates to our carriers are on the shipments, and there's a few other things. If you ever work with an accounting guy, they hate going back and accruing charges, et cetera, so it was very imperative, because we had a $1 million oops one month because people weren't closing out shipments in time, and then there had to be some accounting changes for that given month.

So IT said, "We got this. We got this." So they started building out a housekeeping cube. Do you guys ever use cubes in here? They're fantastic if you know the answer to the question. So they're very fantastic if I want to go, "How many loads, or how many of this did I do last week or last year?" It's really easy, it's in Excel. Boom, the answer is right there. This was going to go out to customer service people, some of them hourly, and they were going to have to go find what answer they were looking for. Most of them don't even know how to open up Excel.

The other fun thing with it is it's cubes on cubes, because there was a write-back capability that they had to create within the cube, which is a separate cube. They were on month six. I get back from Washington, D.C., I get called to a meeting where we're going over housekeeping, and I wasn't part of any of this for the longest time, so I get invited to my first meeting on housekeeping, and find out, "Man, they've been doing this for six months." I go back to my desk and I say... I bring Matt and another gentleman in and I say, "Guys, this has to be easy." At the end of the day, the rules are pretty simple, and then we found out later they're not. But they're very simple to begin with. So we made, within two hours, a proto-type dashboard within Tableau. Within two weeks, using Alteryx, because the business rules were a little bit complicated, we had a production-ready dashboard.

Here's a sample of what the dashboard looks like. It's individualized to our customer service manager. It tells them what problems they should be looking at and what they should be closing. The other thing it allows is when you click on these links, it goes right to our system. If for some reason the item shouldn't be an item on the housekeeping dashboard, because we do have these anomalies, and it drives me nuts, because I think everything should be black and white, but it's not, we've created using Alteryx a write-back, where we go to a table that we've created in our database, which will then just take it off the report going forward. They can write in the number and what the issue is, but it would disappear from the report. This right here gave more insight, and we went from having 4,500 items open on average, to just under 800 in the months following the release of this dashboard. Pick the mic back up.

So where are we today? Today what happened with that, when we released that dashboard, we released a lot of different dashboards. You guys have used Alteryx and Tableau, we were able to build a lot of really slick, cool things, is vice-presidents and executives went to Derek, unbeknownst to us, and said, "Why in the hell is Chip's team not running all of reporting for the organization? We've gotten more out of that team in two to three months than we ever have in the last 10 years from the reporting group. They should be running it." Well, that's a lot of friction. Really the theme of the story is actually going to be a makeup story, so it's not just bashing IT, but at the time IT did not want to let go of these things.

Derek calls Chip and our CIO at the time into the office and was like, "Hey, guys, this is what I'm hearing. I want you two to work it out and figure out what the best solution is." So what happened is we got all of the data delivery for the organization. What that means to you is we would manage all the reports, the dashboard builds, building up data sets for Tableau, doing analytics, et cetera, for the organization, and then IT would manage all the really technical stuff, which they should be doing, which is keep the damn data into an area that I can use it. Don't worry about what business is wanting. Because at the end of the day, I don't want to make reports, I don't want to make dashboards, which has really been a theme. Now I know why I'm speaking, because literally everything that I've been to so far has been the same theme as let business give business the tools to make their own thing. Teach them how to fish. Let's solve the hard problems. At the end of the day, I want smart people working on hard things. We had a lot of smart people that were building really mundane, up-down dashboards or reports, or whatever the case may be. I want to get that off our plate and have business do it.

So today we have a fairly mature self-service model, though we do have some needy people that we work with. We've created a... when using Alteryx, we've minimized IT support on things that are already in our warehouse, so we can actually do a lot of the blending ourselves and create data sets that we can give out to the business. The other thing it's done is, the really astute business people, they've gone away from, "How many loads did I move last week," or how many drivers who leave that started asking harder questions, "All right, why did I only move that many loads. Oh, here's my data so now I can start to look in. Man, this is a seasonality thing." They've been able to do that, where before they were just trying to get the report off their plate. This is where we are today. The next thing, kind of in that self service, where I want to get done with building out reports, these are some of the things that we are creating.

This is called the Big 4 Tableau Dashboard. Earlier I told you guys how horrible it was to be Jeff, like in 2011, and 2010, when I had to come in and spend the first 45 minutes of my day and the first 45 minutes after lunch getting information out of our system. One of our really big Tableau users kept pushing us and kept coming to us saying, "Jeff, I need real-time data." I'm like, "I don't think that exists." I just really didn't want to deal with the issue. "Jeff, I need real-time data." Finally, we got a meeting with the new IT regime, which we're working really well with now, we said, "Man, Eric, how do we do this? Anthony keeps coming and asking for real-time data. I have no idea what he wants it for. Who really needs real-time data?"

What Anthony was trying to solve was the same problem that I was dealing with, because the people that work for him were dealing with the same thing. They couldn't get information out of our system quick enough to make decisions, and when they got that information, they did it manually, they passed it on, and they never made a decision on what they just collected, because they're like, "Man, I just this off my plate so I can stop being bugged."

So it would go up, it would go up to a VP, and they'd say like, "What happened here?" And it would just start the whole cycle again. So Anthony wanted real-time data. What we were able to figure out in our talk with Eric from IT, is we can get it, but we had to build some of the complex business rules, because once again, nothing is straightforward in data, and so that's where we use Alteryx and Alteryx refreshes this data set every 15 minutes or so, and now the directors and the vice-presidents and our van division can come in and look at it and never have to ask a question of, "Where are we at with said customer on the number of shipments they have left?" They can now go to this dashboard. They don't have to go to our legacy iSeries/AS400 system to find this information. They could actually look at it, and if they get on a call with a customer, they have all that information at their fingertips.

So where do we want to be? Really good question. We absolutely want to be better with our data and our boss, Chip, really presses us. He hates it when I tell them the really sweet dashboard that we put together that tells about something we did last week. He wants to get more out of the information, more out of our data, kind of like what we were talking about of solving driver turnover, making our drivers safer on the road, and I do segment analysis on our drivers and our customers. So we want to be a little bit descriptive, but we want that descriptive to live in the business unit as much as possible. We have a lot of Tableau license, we're going to start to release Alteryx license out to the business. We want them to just do that. We'll teach them how to fish. We do want to continue to be diagnostic, and I've a really cool story about that.

We are in a meeting and everyone thinks when drivers leave, it's all about pay, and it's not. In our industry as a whole or in general, drivers typically turn within the first 90 days, and that's an industry thing, it's a weird thing and I think it's because the emotional handoffs from each group is my... we haven't figured it out, but that's my gut, as I hit the mic. But everyone always thinks it pay. They're not getting enough money upfront. There's been a lot of study, though, that say it's not pay. But inherently our leadership team thinks it's pay.

We have a student program and these students, they're new to trucking, they go through the school and they get their CDL, and before they can start driving for us, we put them on a truck with a trainer for... we're going to do it more in performance-base now, but it used to be hour-base, roughly 230 to 250 man hours, which is right around five to six weeks, depending on how much they're driving during that time. Then they get their first truck, and then most of them want to go home. They've been out on the road for five to six weeks, they want to go see their family or whatever the case may be. There was a thought that, hey, should we... because during that time they get paid a salary, if you will. They get paid a minimum weekly. They're not getting paid mileage pay, which is typical in the trucking industry. So there was a thought that said, "Hey, if these drivers go home, should we just pay them what their student salary is?" "Yeah, that sounds like a good idea, I think we'd save a lot of drivers doing that." "Let us take a look at that. Let's just take a look at that."

So we took a look at it, would it make a difference? There was nothing in the data, we couldn't even find a predictor in the data of why drivers were leaving and why some were staying. But I could tell you that pay was not a predictor. If we paid the guys more money, would they stay, because it was not showing up in the data. So we went back and said, "Hey, it's not pay." "No, it's got to be pay." And everyone wants it to be pay, because it's so easy, you don't have to do anything, you don't have to find a... solve a hard problem. It's so much easier just to write a check and make it go away. But it wasn't going to be that. We said, "Man, it's going to be a lot of money. A million plus if we do this program." "I'm telling you, it's not pay."

We had four meetings where we had to defend that and I remember one of them we had was healthy friction, but Chip was like, "Guys, $2 million decision, it's not pay. We don't know what it is and why these guys are leaving, but we can tell you it's not pay." And sometimes that answer is better than the actual predictor, because in our eyes, we're saving money, because I do have a feeling if we would have just done it, or if it was 2012, again, we would have just paid these guys and we wouldn't have seen a result from it.

Other things that we want to do is be more predictive. Last year we rolled out an accident model and we actually made a testament for the whole year with a really small subset of drivers. This year we actually rolled it out and "institutionalized" it. Now our driver managers are making calls on a monthly basis. The way the model works is it predicts if the driver is going to get in an accident within the next 28 days, so on a monthly basis, these driver managers are making calls and we kind of have talking tracks, depending on what is showing up in the data. Some of it could be like if they're young drivers like, "Hey, do you feel comfortable out on the road? How you doing?" Et cetera. If it's a more experienced driver it could be, "How's family life going? Are you doing okay?" Really general talking tracks. The results have been astounding. Unfortunately, I don't know if I can share them, but they have been astounding on the really set of drivers that we've talked to. Most recently, and we really- this is what we wanted to present on here, but we finally got it rolled out on Monday, so two days ago, is a speed dashboard.

Love Jeff Phipps, he's a great sales person, he kept hounding us, and saying, "Hey, what can Alteryx do for you?" We said, "You know what? We've got this problem. We want to get an idea of when our drivers are speeding." I know how fast they're going based on the equipment in their truck, but I don't know necessarily are they actually speeding or not, because we just want the drivers to know that, hey, sometimes, like big brother out there, hey, we have this data. Don't know what we're going to do with it, but we have it, so slow down.

So Jeff did a great job of working with Tom Tom and we got an additional data set through Alteryx, and this week we were able to roll out our speeding dashboard that really focuses on the- every week the top 5% speeding offenders from the previous week. Once again, there's a talking track with our safety department and driver management to speak with the driver and say, "Hey, you need to slow down, Dude. You need to slow down. Here's where we've seen that you've been speeding. Here's the roads." We have it connected to Google Maps so they can actually pinpoint the location of where the truck was.

With that, my final slide, and this is just a generic slide. I get asked every once in a while when I speak, is what do I look for in a business analyst. I do think the best business analyst comes home grown, rides the load, because they're going to have a background of your organization and typically they're going to be hiding somewhere that you don't know where they're at. I mean, honestly, they could be doing some mundane job, and so the very first thing that I look for, and this is typically when I hire entirely, is the ability to care. Are they asking the questions? Do they really care that the organization is going to move forward? It's very hard to determine, like, hey, do they care or not? But the way I do it is if people are asking the right questions, so for us it is, "Hey, what do you think we should be doing for driver turnover? What do you think we should be doing for our safety? Here's an idea I have." If people are coming to you with questions like that, specific to your organization, in my mind they care and that's just something you can't teach.

The other big thing is just people that tinker a lot. I have a lot of people on my team that don't really have a technical background. Our best Tableau person, he left a little bit ago was an English major. For the most part, the people that are doing the analytic work on my team, don't have a technical background, but they like to tinker, and it's something about this millennial generation, they like to stretch the limits of what each of the tools that they have. When you start to see that in an analyst, harvest that. Give them ideas. Say, "Hey, do something really cool with Alteryx, and bring it back to me, whether it's job related or not."

Matt, for instance, tried to use Alteryx and Tableau to build the best fantasy football team ever. He was still learning the tools, but he tried. The other thing which really goes hand in hand with the tinkering is just innovation. Those people that- and this is the most over-used term- but thinking outside the box. So people that are really just stretching the imagination of what is possible.

I'm telling you, a year ago, we didn't think doing a speed dashboard was possible. But with the help of a lot of people at my team, we were able to come through and make that. We didn't think an accident model would be possible, but once again, with people like Matt, we were able to come up with an accident model. So we have a few more things up our sleeves and hopefully if we speak again, we can kind of tell you what those things are.

The other is number four on my list, and these are all in order in my opinion, is the aptitude for understanding business. You can be taught a little bit, right? Like if you have to hire someone off the street, you can teach them what you guys do from a business perspective.

Really, the last thing is the aptitude for technology, so like skills based. At the end of the day, the people that I've seen in our organization who have the best technical skills, whether it's SQL or not, are actually the people that are the least analytical in our organization, and I think that's what created some of the disconnect before our team was founded.

The other thing, too, is you can teach technology. You can teach people how to write code. You can teach people how to use Alteryx or Tableau. These are things that are all taught, but if you look at the things up top, those things aren't really taught. Those are the skills you want, the soft skills. It is hard to measure if someone has them, but those are the things that you want in your analyst and they can't be taught. You can't teach work ethic, you can't teach people to care. That's, in my opinion, inherent. But the bottom things you can definitely teach.

So without further ado, I'll take any questions. Before I do this, though, I've been made fun of a lot from my friends, they keep calling me a keynote speaker, and I'm not, so I'd like to take a selfie, if you don't mind. I took one early on and they're like, "Who are all the people that paid money to go see you speak and there was like two of you in here." They're going to think that I trumped it. Awesome. Thank you.

At this point, if anybody has any questions, I will bring the mic on over.

Crowd question:
So, two questions. The first one, on one of the earlier slides, you said like 300 plus percent increase in data engagement. I was just curious how you were measuring that.

Jeff Walters:
Yeah, so the way we measured that- so the question was the 300% engagement, data engagement. And where I had measured that on was the actual housekeeping dashboard, and a couple other big dashboards we use. I could measure that because I had a housekeeping report that was built and I could see what the usage was and then I could also see what the usage of our dashboard was. So it went from- I can't do the math in my head, so it's not going to equal 300%, but- it went from like five people a week looking at the housekeeping report to like 30 to 40 a week looking at it, which that enabled us to close down the- what did I say- 4,000 plus open items to just under 800.

Crowd question:
And second question, so my company has had Alteryx for 10 years, really didn't do too much with it for the first seven, and only my team has really taken advantage of it, and we have a lot of the classifying that you're talking about with IT and our business intelligence team, making cubes, doing Excel SQL dashboards. That's still kind of going on and so with the change in your company, has IT or your BI team kind of moved towards the Alteryx or the analytics side or has that just- responsibilities that they used to have, have moved out of them and into your team?

Jeff Walters:
Yeah, so good question. So if I heard the question right is, so when our team was formed, kind of where did the ownership of reporting and analytics kind of fall. So before our team was formed and just shortly after our team was formed, reporting and traditional BI and analytics actually, fell under IT, and at the time analytics was probably six to eight months old, and it was managed in IT. We didn't think that was the right decision. We begin Chip, Jeff, and Matt. We just didn't. But we are an army of three, right? So we're going against the mighty IT. There was a lot of head-butting. But here's what happened is, the greatest weapon that we had was honestly, at the end of the day, was Chip and Jeff, and that's because we understood what the business problems were. That's what you need in an organization. You want to make that change, you have to have business backing. Chip is the vice-president, so that helped a lot, but we were able to deliver things fast, and it wasn't us who went to our CEO, [inaudible 00:36:39] CEO, Derek Leathers, to say, "Hey, they should manage it." It was other VPs and other executives who went to Derek and said, "Why in the hell aren't they managing this, because we're getting more results from them."

So my personal belief is reporting a dashboard should live solely in the business groups as much as possible, so that descriptive, give them the tools- if you hear like Coca Cola or Ford or Sony or those guys talk to me, they believe in the same thing, is give them the tools, teach them how to fish, and have the smart people work on hard things.

So for you, I think you just need to get a business champion to help you and then keep just giving them results and giving them items, and saying, "Hey, look at what we've done," and then I think you can make a change or momentum ship within the organization.

Crowd question:
Thank you.

Crowd question:
To build on that, right, like you can have the executive buy in, but what would you say was like the secret sauce in actually continuing to work with the IT departments, because they can frequently stonewall you or not give you access and that kind of stuff.

Jeff Walters:
We're going to air some dirty laundry. The question is, what was the secret sauce of us working better with IT? When the change was made for our team to take over all the data delivery that reporting, there was six people in the BI group that did not like it, and they left. They didn't want to make a change. They wanted things to remain. They wanted that first slide that I showed you, they wanted things to be done the same way. When they left, it was like a huge burden just being lifted off us. So there was that picture of the possum and the cat drinking. We have a really great relationship with IT now from a data perspective. We can sit down with them or we can request data, but it took certain individuals leaving our organization, so I wish I had a better answer for you what the secret sauce was, but sometimes you just need people to leave in order to move forward.

Crowd question:
It sounded as if the application of analytics was primarily to support operations. Are you doing anything to apply analytics for strategy execution?

Jeff Walters:
So if I heard the question right is- because what I showed you guys today was kind of like we're putting dashboards onto a bad application. So we're using analytics or reporting to help solve in the moment things, because we can't get the information out of our application, and I believe your question was, are we using analytics to be more strategic? That is our goal, we've started to do that within the last 12 months, once again with like the predictive safety model. We want to get more into market segmentation for our customers. We're very retail centric. Are there other customers that we can go out and take a look at? That would make sense. We know over the last eight years in transportation, the markets have changed drastically, and it's been hard for a lot of transportation companies to change, including us. It's just- Chicago used to be just loads and loads of freight and now it's all dried up, and it's taking everyone a little bit to get used to that, so we want to get more into that market segmentation for our customers, and then also market segmentation for our drivers. So, yeah, we want to hire as many drivers, but we also want to hire what drivers make sense.

One of the things I didn't talk about in here, we do have the Alteryx data set, and we just started to get into the mosaic data, and that's been really interesting over the last couple weeks we've gotten into it. It's things that we've been able to see as what type of mosaics from a trader-student relationship make sense. And it's alarming. I mean, you can see these two mosaics really get along and these two mosaics don't. But then our data... but then also get an idea of how to advertise to these guys, because within each mosaic group, it gives a description of like, "Here's the best way to communicate with these people." Honestly, some of the things with the messaging we do, doesn't resonate from a mosaic perspective to some of our bigger, bucketed drivers. Mosaics are right, because I've fallen to the active group. Does that answer your question?

Crowd question:
Do you have any philosophy or advice for managing workflow? Not managing the actual workflow, but so-and-so created a workflow today, that's really cool. How do you manage so many and so many different people creating workflows?

Jeff Walters:
So the question was how do we manage our workflows? That a very good question. Right now we're very centralized with Alteryx in our company. It's only my group, and my group is roughly 10 who use Alteryx. So from that perspective, it's pretty easy to manage because we all create our own thing and put it on there. With that said, though, I challenge these guys on a quarterly basis to go through and figure out what is being used or not. Most of our workflows are supporting Tableau dashboards, so I can go out and actually see if Tableau dashboard are being used. But right now it's not the best answer, but I can do it easier because it's very centralized, but we are looking to put Alteryx potentially out to the business units, and we will have to have a strategy on how to manage that, just based on all the conversation that I've heard here.

Crowd question:
I think you mentioned that you had somewhere in the neighborhood of 1,000 Tableau dashboards that were out there for the organization. Could you talk about the QA process that you have for your group to go through, kind of before something is released to the entire organization?

Jeff Walters:
Yeah, so first off, the question was what's the QA process to institutionalize a dashboard? So with it our Tableau environment- we've just changed the sights up- but for sake of this, we have four different sights. We have a production sight, which we just call the Werner sight, we have a prototype sight. We have an at-hawk sight, and then we have a customer facing sight, so our customers can come in and get their dashboard. The at-hawk analysis is- that's where most of our dashboards are- and that is just a mess, graveyard... but it's up to the business owners of each of the folders to manage the content that they have on there. For us, when we build out the dashboard, we throw it on to the prototype sight and then make sure that everything is working, so at the end of the day, my stuff that I'm hitting is already data that has gone through the QA process, so I'm already hitting institutionalized data, if you will. But what we do is we put into the prototype, we don't have extracts run on it, so it's kind of stale data to a degree, unless the user wants it refreshed. But when a user or a director or whoever buys off on it, and we're good with it, and it's doing everything they want, we'll then put it onto our Werner sight or our production site.

Thank you. Any other questions? Okay, and with that I would like to thank Jeff for giving us a sneak peek into the Werner analytics journey. And before everybody goes every which way, if you could pull up your app and rate the session, that would be wonderful, so that we can continue to improve, and thank you for attending.

Jeff Walters:
Thanks, guys.



Pour Démarrer