What is the Digital Workforce?


What is the digital workforce? What does automation mean in real life? What is the data's role in helping create the digital worker? How will digital workers impact the day to day of the human? Let BP3 answer your toughest questions.


Lance Gibbs: Shall we start-

Chris Kimmitt: Okay. I'm Chris Kimmitt, I work at Wesco Aircraft as the head of data Architecture and business intelligence. Do you have any other questions in the [crosstalk 00:00:15]-

Jordan Bohall: I'm Jordan Bohall, I'm the vice president of analytics for IH Mississippi Valley Credit Union and I'm an adjunct professor for a master's in data science program for Western Illinois University.

Lance Gibbs: I'm Lance Gibbs, co-founder of BP3. And I have a background in data and statistics and statistical analysis. So before we get into the digital workforce bits specifically, I thought one question that might be interesting is just trying to getting a sense of the last day, day and a half now. You hear the word automation to the point now, and intelligent automation and whatnot quite often, but just trying to get a sense of where it really fits in organizations. Obviously with everybody here, you're here for a reason, right? There's something going on there. But one of the questions I had was... maybe, Chris, start with you and your organization, where does the term automation really fit? Is it a board level thing? Or is it not quite board level, or executive level, or whatnot?

Chris Kimmitt: In my organization right now it's at that board level. We're looking at efficiencies, that we can get up how we can scale the organization as a whole so that we can meet the spikes, we can meet that new business without throwing headcount at it. That may be temporary head count in nature. So it's at the very top.

Lance Gibbs: What kind of [inaudible 00:01:40]... In your business, Wesco Aircraft, what do they really do?

Chris Kimmitt: So Wesco Aircraft, we're supply chain for the aircraft industry. So think about any hardware parts that you see on any kind of an airplane. Boeing is a major customer, all of the commercial airlines. And so we supply hardware parts. We also supply electronic harnesses and chemical products for anything aerospace related. That includes space and rockets as well.

Lance Gibbs: In a global distribution?

Chris Kimmitt: Yeah. Global distribution.

Jordan Bohall: So in my organization automation definitely sits at the exec level and senior leadership, with the rest of the VPs. There's been a bit of pushback from folks, say our underwriters or any of our retail staff, that are afraid of automation like you so often hear. And so right now it lives at more of the senior and exec level. Absolutely.

Lance Gibbs: Okay. Yeah. For us the term automation is something that we've always used in our world of consulting and advisory and services. But the definition can vary, right?

Jordan Bohall: Yeah.

Lance Gibbs: Some of that's driven by technology, some of that could also be driven by particular, just global efficiencies that start to happen in the industry that we're working with. Even outsourcing, right? Aspects of businesses or aspects of the supply chain, people would call that automation. For them that is effectively what's happening. They put it in, put it out there and get something back. So it's been an interesting, I think, ever evolving definition.

Lance Gibbs: All right. So talking about the digital workers. I love Beaker. So this is, I think, a great euphemism, right? We got our data scientists and we got Beaker, [inaudible 00:03:27] assistant. I'll let either one of you guys jump in on this. What do you think has been data's role? As it has evolved, we talk a lot about analytics and insight. What has been that data's role in helping create this notion of the digital worker?

Jordan Bohall: I'll jump in. We're going to talk, I'm sure a little bit about data literacy, digital literacy. And so for the digital worker, I see this issue as goes data literacy across most of our workforce. And so the rise of the digital worker, I can see as a direct response to folks not necessarily having the skill sets that we need to be building APIs or any sort of automated solution. Absolutely.

Chris Kimmitt: And on my side, we more or less have used data to figure out our underlying processes in a lot of cases. Places where we thought we had processes before, we tried to robotize found out that they weren't really processes at the time. But we've used the data to tell the story of where things are escaping and what it looks like so that we can help move over to that bot world with those learnings.

Lance Gibbs: So in that kind of notion of the bot world, I know that's been kind of more the analog for a digital worker as the non-human actor of it. But we also talk about having robotics as an example, but there's also going to be other technology where it sits alongside the human in getting work done. Has data been a big driver, you think, of this kind of evolution itself? Proliferation of it? Democratization of data, as an example?

Chris Kimmitt: Yeah. I would say it's pushed us into an area, I guess, that we were kind of uncomfortable with. Our data was centralized in IT. We've been slowly pushing out to the masses and helping them understand how their processes are working better, which helps them look for new opportunities. So it's been a big driver in that aspect for us. And as we move to where bots can really sit and help augment some of those, what we would call... I'm trying to think of the word. More of the commodity type tasks to help them, help them scale.

Jordan Bohall: Yeah, I completely agree with that. I see a lot of the automation pieces that we've been doing in my organization. And there is a talk that scared the crap out of me yesterday because it was so close to my heart and it hurt. But nevertheless, where we are with our current automation journey, we've been getting rid of these silos and trying to make data available as much as possible and collect data from all these individual silos. And so with the bot revolution, again from another presentation, I'm not sure how long the bots will actually last in industry. But nevertheless, it will help me push forward my automation initiatives so that people get more and more accustomed to the various automation techniques. Absolutely.

Lance Gibbs: Yeah. I think one of the things I've seen a little bit of, and maybe it's just a function of the maturity where the market is, some of our clients and people that just... not even our clients, people I've run into and met with other companies. I wouldn't say competitors to us, but maybe larger entities and how they're looking at it. And one of the things is even if they talk about automation and the notion of a digital worker at a senior level.

Lance Gibbs: One of the things we're also talking about is, where is the obstructionism start to happen? Because there back in the days when we'd talk about, "Hey, we're looking for process improvement, process optimization initiatives." And we were going to sit down with the business, "Hey, let's talk about this." The first thing they're doing is they're looking at you thinking, what about my job?

Lance Gibbs: And so coming at them and saying, "This isn't about evaluating whether or not you can do your job. We're just looking for inefficiencies and where we can shore up." This becomes much more overt. Which is, "No, actually you are talking about my job or my team's job." And that's a big shift for people and how they're going to actually start to adopt or even embrace this. Again, how obstructionist might they be in doing it? It's going to be an interesting, I think, dynamic. I don't know with your organization if it's the same, people kind of, "Hey, this is the way we do things. We like it this way. We're not looking for bots running around."

Chris Kimmitt: We had a little bit of that, but at the end of the day, it wasn't scalable. So we had to look for bots to try and augment some of that. We can only throw headcount at it for so long. And a lot of those same tasks that they were trying to protect early on were the ones that they complained about the most as well. So it was kind of a funny situation.

Lance Gibbs: Brings us to our next question here. Poor Beaker. So how will these digital workers impact the day to day work of the human? As both an academic and an operator in the business, how do you see it?

Jordan Bohall: So I'm going to piggyback on the last question you asked. And with the pushback we're getting, of course we have people who were taking chunks out of complete workflows and that's all they know. And so when we have our process improvement coach come through, they're fine with that because she's a very nice person and she can approach them in the right sort of ways. But then when I come in on top of that and say, "Yeah. Okay, all these process improvements we just made, we're going to automate that with the bot." They're terrified.

Jordan Bohall: So to answer this question, I hear so many times at various conferences or in the various trade publications that, "Oh, we'll just upskill our employees. We'll upskill them." And with the rise of the low-code RPA solutions, for example, that's not a big gap to achieve. And so what I see happening is there'll be a bunch of people who will make that jump to be able to do the low-code scenarios and build their own bots that they'll have running in the playground with them. But at the end of the day, they'll have to have somebody, if it's a manual bot, to sit there and still click it to make it go. And they'll be able to work much more quickly, but it might be a same sort of skill set that we need. So there'll be a bit of a stratification.

Chris Kimmitt: And from our side of the house, what we were seeing with a lot of the areas we've already dug into. And I'm primarily thinking order entry and customer portal type interactions where we're having to scrape information out of customer portals and load orders into our system. A lot of that was being handled by a lot of our people that we would prefer to be CSRs and customer touchpoints, versus doing manual data push and pulls.

Chris Kimmitt: And that was one of those tasks that I mentioned before that they didn't like. So it started out as kind of iffy, a lot of obstruction and a lot of pushback on it. As we put in the first few use cases, they saw the light and moved in the right direction. It's helped get those people in front of the customer a little bit more, better customer service on our side. So it's been a positive journey so far.

Jordan Bohall: So we have a similar situation. We call them human APIs for lack of a better term. What were the backgrounds of your CSR agents and how did you get them over that hump to that next level?

Chris Kimmitt: A lot of the backgrounds, they were traditional CSRs, or they were the account reps that we had in place before. What ended up happening over the years was those commodity tasks, where they were going out there and having to do a little bit more of the background work, grew and grew and grew. And so what we hired them for was not what they typically were doing at the end of the day. So that job changed over time. Now we're trying to pull it back and make it scalable, quite frankly. Because the more customers we bring on board and the more of these interactions that we have to have, the harder it will be for us to staff up for them. So it's a revisit of their initial job.

Lance Gibbs: Yeah. I can definitely see that. So what are the skills needed to manage these digital workers? Now it's one thing when you're leading them, it could be talking about CSR or something, maybe it's a contact center, maybe there's 300 people in there. Let's say you don't need 300 people. The fact of the matter is it's a subset of that. Or maybe you've got a group where that's like 15 different groups of folks, they're maybe segregated by geography or by product lines or offerings or strategic accounts. And folks are very used to that kind of interaction or how you manage that. But what kind of skills are needed when maybe half of that workforce really is digital bots? How do you take a CSR, a CSR manager, and what do they need to have?

Jordan Bohall: Yeah. So when I teach at the university, it always astounds me that my students aren't what I would consider digitally literate just as much as my employees are. And these are people trying to become data scientists. And why don't I consider them literate? Or why don't I think that they could actually manage digital workers? It's because they lack a very, very basic skillset, critical thinking, creative thinking. It sounds so simple. Who called us digital elites, yesterday maybe? We're all elites.

Lance Gibbs: Oh, Craig. Yeah.

Jordan Bohall: Craig called us digital elites. So we're awesome. But at the end of the day, we're not great at critically thinking, and we're not great at logical thinking and reasoning, and we're not great at seeing entire workflows from beginning to end of process. And so these three areas, I think are absolutely necessary to manage our digital workforce.

Lance Gibbs:
Are they going to get that in the world of academia? Is that going to have to be something homegrown in organizations?

Jordan Bohall: Yeah, definitely in organizations. Organizations have to really ramp up what they mean by critical thinking. And you'll think this is kind of dirty, but my background's philosophy. Not CSR engineering, like I'm guessing most of you. And so I see the organization being responsible if they want to manage the workforce and not really lose head count as much as people, doomsday, say it will. They'll have to really focus on that.

Chris Kimmitt: And whenever I look at this, immediately, I guess, the data literacy side pops in my head. So I may be jumping a little bit forward, but just knowing what their data's telling them too. I'm thinking once we move into the cognitive use of bots and other things, they're going to have to know how to read, to interact, and to argue with that data. We're going to be throwing all kinds of different information points at them that they're going to need to know how to interpret and make use of them.

Chris Kimmitt: Learn how to do that storytelling so that they can bring it to the business and convince people this is the way that we need to go. So there's just a broad, just better understanding of the data. And it seems like over the years we've taken a step back, like you allude to. And I think the quicker that we adopt technology, it makes that cycle even worse. We have less and less understanding of what's going on behind the scenes in those aspects, the more and more things move. And the quicker that we jump around in technology.

Jordan Bohall: I liked the idea about the storyteller. What do you mean by that?

Chris Kimmitt: Just being able to take those data points and paint the picture and tell the story to the organization. What does this data relating to us? It's one thing, and I see it all the time with a lot of our workers, where they'll bring a giant spreadsheet in front of the CIO and plop it up there and go, "Here's your answer." And everyone looks at it confused-

Jordan Bohall: That works for some people. My CFO loves it. Loves it.

Chris Kimmitt: He's not a statistician, is he? I had a former CIO that was a nuclear physicist. And that's what he loved. He thought that all reports should be a giant spreadsheet. The bigger, the better. But being able to visually, either a storyteller or something to get it to the audience so that they understand what you're trying to convey to them. A lot of times, whenever you show spreadsheets in front of a lot of groups, the whole point of it's missed. They may make a decision that's completely arbitrary to what you were trying to push forward, just due to the way that it was presented. So to get those workers more and more used to telling the story, how to relate to their audience, how to get their points across.

Lance Gibbs: Yeah. Seeing it quite a bit, where you have a organization you're working with and ambitions are great and their ambitions are high and they realize we're not exactly bringing them necessarily the keys. This self realization journey they go through, like we can't just fix this in a vacuum. This is going to take me working with... If I'm policy servicing, it's going to take me working with the claims folks, the warranty folks, or maybe the sales people to really make this happen. We're going to have to reach across the aisle, whatever euphemism, you want to create bridges, again over the silos, over the gaps. And data was an important part of that. We all had a process map and that was pretty quick. Everyone got the notion of here are the boxes, here are the participants and the lanes, and here are the systems we interact with. What was frequently missing was the data, the shape of it all.

Lance Gibbs: Processes aren't two dimensional, they're not just boxes and lines. There is a lot more to that and data represents that. So how do you create a cohesive story? And it seemed like that started to gain... that's why I'm hitting on this, because you said maybe we've taken a step back. I agree. Because as that started to be much more talked about and discussed, how we need to bridge the process. They don't care about your org structure, they cut across it. End of story. How do we start creating these bridges? Then you've got a situation where there was a proliferation of technology in the world of SAS and logos popping out of Weworx, and every other thing every week. And it made it very easy for a particular business area to swipe a card or create a transaction where they can basically subordinate IT and their involvement and they can be more self-driven.

Lance Gibbs: And there is a lot of freedom in that. And to be able to say, "Oh, I control my own destiny. The pressure's on me to be able to show X, Y, and Z, or maybe get more performance out of my group." "Hey, there's a great new little SAS application we can use. And it's low-code or no code or whatever it is, or we just plug it in and boom, it gives us the answer." And even with some of the reporting. I mean, that was kind of the reason Tableau, we started getting [inaudible 00:19:40] to begin with. It was the ad hoc reporting and visualization, so that all sounds great. But what happens and what we've seen as they've been able to be more in control of their destiny and maybe not rely as much on others, it's actually hardened the silos.

Jordan Bohall: I was going to say it pushes the goal posts back.

Lance Gibbs: It does. So now you have to fight again to re-establish a level of relevance, because they've relieved some of their own pain points through a very quick and dirty vehicle. And maybe it's good enough, which I'm a huge fan of. Pragmatism should rule. But then what's happened is there's not a real compelling reason for me to talk to the left or to the right, "Hey, that's their problem." And so we're back again, what it felt like, it took a long time even get some of that progress made with it. And now I'm just wondering with some of the skills needed to manage digital workers, or even having digital workers, is going to be one of those things where what does that do to the dynamic of an organization? And I don't think we know because we've not seen scale? We just haven't seen scale with any of this stuff just yet.

Jordan Bohall: Let me ask you guys this. We've been playing around the idea of like a VP of HR architecture, or some sort of architectural manager. And their whole goal is to align each one of these pieces so that we don't recreate these silos once again. Align all of our IT, align all of our technology, every other process. Is that something that you've seen work or fail?

Lance Gibbs: I've seen mixed results with having the... and I'm not sure it's the same role, but I have seen where you've got an IT organization. And of course 85% or more that IT's organization is not to innovate, it's not to go check out the new stuff, it's keep alive. Because all this stuff is great and wonderful until the systems go down and then it doesn't matter what you got. So that's where they spend their time. And of course the business operations and other constituencies within the organization, non IT, have needs. And that conversation is always of met with some level of friction. Where do they fit in priority? And who can yell the loudest?

Lance Gibbs: I mean, this is how literally how some of these things have gotten [inaudible 00:22:08]. The squeaky wheel... I haven't seen squeaky wheels, I've seen boulders crashing through organizations to get what they needed. But now seeing the introduction of those business IT liaison. And their world is not necessarily just, "Well, I'm an intermediary. So what's my value?" It's, "Well, I'm trying to mediate, I'm trying to help codify what those needs are and where we can take this to the organization." And then that gets bubbled up to the CIO, CFO, CXO, whatever it may be, where priorities are made. And I've seen that with mixed results. I think there's a lot of culture to hash out with it, but I've seen good and bad with that. Do you have anything like that at Wesco?

Chris Kimmitt: We just recently built that business liaison. Before then, it was a good portion of my BI team that was filling in that slot just because we had that frequent interaction, almost daily interaction, with all of the silos going across in order to get their reporting. But now we've formally broken out that business liaison group. What we're trying to push, because we were always good at roadmapping, and having a strategy is trying to instill that into each one of the silos so that they build that roadmap and strategy.

Chris Kimmitt: Then we can overlay all of them and get that cohesive [inaudible 00:23:35]. That way nobody's fighting and stepping on the other ones. Just for their pet projects it goes into a larger, I guess, document that goes up to ELT. And they're the ones that go through there and each person gets two or three items of their choice, and then the other groups get what they need. So that's how we're managing it right now. It's still too early say if it was the right way to go or not, but it kind of seems to be getting some pretty good traction at the moment.

Lance Gibbs: Anybody else out in their organizations have something similar to what Jordan's describing, what we're talking about right now? Just a show of hands, anybody have that kind of intermediary in their operations? You do?

Jordan Bohall: How's it working out?

Audience Member: [inaudible 00:24:24] because I am new to the job as [inaudible 00:24:29] director. We actually [inaudible 00:24:31] on the business [inaudible 00:24:35]. So it is like the intake process and working with the business is our job and we work with IT, [inaudible 00:24:38]. [inaudible 00:24:48].

Jordan Bohall: Okay. So you're still right here? Yeah.

Lance Gibbs: Yeah.

Chris Kimmitt: But that's a good point. We pulled a lot of our liaison from the business side to build that bridge, that way we weren't going into that adversarial IT role and just embedding ourselves in there. And that's been tremendously helpful I thought.

Audience Member: [inaudible 00:25:17].

Lance Gibbs: Good. Did you hear it?

Jordan Bohall: Yeah. Go for it.

Lance Gibbs: So she has somewhat of an advocacy side, it doesn't report. It's not like [inaudible 00:25:24], really on the business side, but kind of supports and advocates for the business, with IT.

Audience Member: Yeah. I report to the head of strategy.

Lance Gibbs: The head of strategy, so it's part of your-

Audience Member: Yeah.

Lance Gibbs: Okay.

Jordan Bohall: Nice. Thank you.

Lance Gibbs: Since you started on this, I thought we could just put it out there. Show of hands real quick before we get into this. Data literacy, anyone know that term? Heard that term? Show of hands. You can raise it. It's okay. There are many people in here and you're not on camera. There you go.

Jordan Bohall: Do you have a definition for it?

Audience Member: Not one I can come up with on the spot.

Jordan Bohall: I've found that there are over a hundred different definitions for either data or digital literacy coming from some of the best journals in the world. So don't feel bad because nobody seems to know what's going on here.

Lance Gibbs: [inaudible 00:26:14] on data literacy? How you view it, maybe you start with that?

Chris Kimmitt: Yeah. To me, it's just the way that you... like I mentioned earlier, you read, interpret, interact with the data, argue question with the data. Just the overall, being deep enough into the data that you have a good comfort level with it.

Jordan Bohall: Yeah. So I think data literacy is going be what citizenship requires to be active and participate in the 21st century and beyond. However, there are... what? 52% or 51% of the world's population that doesn't have access to the Internet or a computer. So that means 48% or 49% of us who do, given that we don't really know what data literacy or digital literacy is at this moment, if we are moving down this hard automation track, this hard AI and all that kind of thing. Then in order to make a living or to survive, we're all going to have to truly become data literate. So it's a massive issue. It's probably one of the biggest issues of the century.

Lance Gibbs: Yeah. I think there's very tactical, very specific instances where for, instance, I talked earlier about the process maps and data being part of that. And people say, "Oh, I understand the data cycle time, throughput, all that." I'm like, "Those are operational measures, I get it." But do you know the difference between continuous data, attribute data? Do you understand where these particular data chunks would live? Do you understand how to do stratification? Do you understand how to do disaggregation? Do you understand the difference between dealing with normal distributed and non-normal distributed data?

Lance Gibbs: So when you're getting this data and you're trying to read and interpret what it means, these are some of the elements that... Sure, fine. If you have a background in statistics or economics where this would all be played in, then you're going to get it. If you don't have that background, the organization, I think it's incumbent upon them to have to start doing that. But even before then, you're going back to academia. We still have, it's come a third now, they're technically literate. They run rings around me.

Lance Gibbs: I've given up. By the way, I've completely checked out. I am the worst IT help desk my mother-in-law and family could ever call on again. Back in the day. I was really, really good. No problem at all. Today? Nope, not touching it. Windows 10? I got nothing. And so there's enough movement out there now that it's hard certainly for me, speaking for myself, to keep up with what's relevant out there, what's meaningful out there, and certainly how data is being used. How ignorant are we about data? How ignorant are we about the intent and the use of data? Look at Cambridge Analytica, look at Facebook.