Hello, everyone. My name's Matthew Bidwell. I'm an associate professor of management at the Wharton School and a faculty co-director of the Wharton People Analytics Initiative. I also teach courses on people analytics to both MBAs and executives. I'm really excited to be here today or whenever you're watching this, talking about AI and people management. As Petraeus described, managing people is complicated. People have free will. Not only do we need to hire them, we need to motivate them. We need to persuade them to do the things that we want them to do and they very much react to the ways that we manage them. Figuring out how to manage them appropriately is key to running a successful organization. Now, in this course, our main goal is to describe how new technologies run artificial intelligence and machine learning can create new opportunities to find better ways to manage people. Before we get into that, though, I think is actually useful just to say a little bit about how organizations have been managing people so far. After all, before AI organizations had all these problems and they were thinking very seriously about what they could do to get the best out of their employees. Now, clearly, in order to really get into how organizations manage people, that's 10 hours. I want to spend 10 minutes. Okay. But just very briefly, go through how organizations solve about some of the main challenges that they face and and try to address this. I want to talk about three in particular that are critical to organizations and we're really going to focus on over the course of these videos to talk a little bit about hiring, a little bit about managing workers engagement and a little bit about attrition. Let's talk about hiring. Hiring is a really fundamental process in managing people. After all, if you bring the right people into your organization, you don't need to worry so much about training them. They're more likely to do the job well. Hopefully they're even going to be more engaged if you hire the wrong people or they stay and do a terrible job or they leave immediately and you have to hire all over again. Either way, being able to figure out how to hire successfully is really important to organizations. Peter is going to spend a bunch of time discussing how AI can be applied in this area. I want to spend a little bit of time before then just talking about, what are the conventional ways in which organizations hire people. I think it's fair to characterize it in some ways a tug of war between very structured processes for figuring out who has the most potential and managerial discretion around how to hire with a side order of cost control. Let me talk a little bit about each of these. First of all, the structure process, so we can think about hiring in some ways as a measurement problem. A lot of AIs about data, how we measure things here in any hiring. What we're really trying to measure is applicant's potential. We're trying to figure out if we bring them in as employees, how good do we expect them to be? Now, there's been basically a century of research by psychologists into this question, trying to figure out how we can assess people and predict their potential in different jobs. This has given rise to a bunch of best practices the organizations try to employ. We know that three things are pretty good when it comes to predicting performance. First of all, most importantly is what we call structured interviews. What's a structured interview? It's when organizations first figure out what attributes of people are most important to do the job. Then they work out what questions can we use that will actually assess whether people have those attributes and they define a set of questions they're going to ask every candidate and then they also have a rubric. The answer to each question they define beforehand, what's a poor answer, a medium answer and a good answer and they run every candidate through this, score them accordingly. Structured interviews, very good in some settings, test of cognitive ability, broader job knowledge tests can do a good job of figuring out who has the abilities necessary to do the job and you need some evidence that personality test can also be effective. HR departments will adopt these practices and try and use them to do a structure of assessment of people to figure out who has the most potential to perform well in the organization. They'll also try and avoid the things that we know tend not to work so well in particularly unstructured interviews. That is interviews where we basically just sit down with people, ask them a few of our favorite questions, have a back and forth and try and get a good feel for them. That's the structured process side of hiring. The other side I said, was managerial discretion. What do I mean by that? Well, the HR department can put all of these nice structure processes into place. But at the end of the day, in most organizations, the HR department is not actually doing most of the hiring. For some high volume entry level roles, and it might go through a structure recruitment process, but most of the time new hires are actually being hired by what we call hiring managers, the front line managers for whom they're going to work. What this means is that even though the H.R Department might have all these structure processes there use depends on whether or not the managers involved actually want to apply them. Often it seems they don't. We often find that managers resist these structure processes, partly because they find them very stiff and formal they get in the way of a smooth interaction with people when you're following down a set rubric on a questionnaire. But also, I think because we tend to believe we're very good judges of character, we're not. But we tend to believe that we really know what people like. We form opinions of them quickly, and then we spend most of our time trying to justify those decisions. If you believe you're a great judge of character, why would I need to follow this very structure process when instead I can just chat with them? Also we find a lot of managers may not follow these processes and instead basically just prefer to ask people what their spirit animal is and take it from that. The third piece that I mentioned about hiring is in addition to this back and forth doing structure process, trying to get people to follow them, there are also elements of cost control. You want to hire the right people, but you also want to do it efficiently. Organizations are also spending a lot of time worrying about their cost per hire, how much are we spending in terms of external spend on advertisement and recruiting? How much are we spending internally? The time of our recruiters and so on in order to bring in each hire, is that too much? How do we bring it down? They may also worry a lot about how long it's taking to fill the job with track time to fill the job, because the longer to fill the job, the longer it's staying empty and the more disruption that's going to cause. We look at these as well and try to figure out how do we get our hiring done properly? Yeah, I think Peter will discuss a little bit. There can be a challenge there when inputs are easy to measure and outputs are hard. Maybe we focus too much on cost per hire and not enough on quality. Some organizations will also try and figure out are we doing a good job of this? For example, you might survey supervisors after three months or six months and say, okay. We hired this person, would you hire them again? You might also look at attrition after three months or six months also using this not only do we try and get the right people hired, but we can also start to look at the overall structure of our hiring process and get a sense, is it doing the things that it's intended to and through the use of all of these levers, organizations really try and make sure they hire the right people effectively. A second area of people management that we're going to talk about in this course is engagement. Again, coming back to this fundamental issue people have free will. We need to motivate them to do the things that we want them to do to make this organization successful. Now, there are lots of levers that we can pull around motivation, a big one is job design and structuring the exact tasks that people can do. The culture of the organization and the way that people are led also helps to define their relationship with it and how excited they feel about coming to work every day. A lot of things we can do. Fundamental getting it right, though, is actually knowing how engaged people are in the first place. Organizations spend a certain amount of time and effort trying to figure out are people engaged? How do they feel about this now? Obviously, it's not always easy to know what people are thinking and feeling. The standard way that we try to do this often is through employee engagement surveys. I'm not sure how many of you have taken these surveys, but basically a lot of companies will pretty much every year send out a long survey to their employees. They will ask you, how do you feel about working here? What do you like about your job? What are some of the downsides? How do you feel about your manager? What's your opinion about you pay your benefits, the company? All of these things. Often they tend to use very specific sets of questions developed by psychologists that really enable them to clearly tap into specific opinions about each of the attributes that they might care about in their work. Okay, why do this? First of all, it gives us a good sense of how are we doing overall or are people excited? We do it from one year to the next, you can start to examine trends. Right? Are we getting better are we getting worse? Even more important, you can look across different groups. What are the parts of the organization where people are engaged? Where are the departments where they're disengaged? What can we do about that? Chimney's engagement surveys are reasonably effective. If you want to know how people, how engaged people are, you can do a lot worse than just asking them. That said there can be some problems. Some will come back to a couple of times this year. There's obviously a gap between what people are thinking and what they tell you they're thinking. For example, people might not always tell the truth. I remember back in my labor economics class, one of the things that I really picked up was that anybody from your company ever asks you, are you paid enough? The answer is just always no, no, one would like to be paid more. People may be a little strategic about, trying to make out things aren't great in the hope that people improve things for them. On the other hand, they can be worried about retaliation. If I'm in a small group and people are asking me about my manager, am I sure that my saying my manager is terrible is not going to get back to her and then get back to me? You have to be a little careful. The other challenge with these surveys is also they're expensive, right? I mean, if you're getting everybody in your company to fill out a survey, that's an awful lot of person hours that are going into that. Even just the effort to make sure people fill it in is not trivial. I think to avoid those expenses. A lot of organizations will therefore only do this once a year, which, again, if you want an overall sense of how we're doing, is not a terrible idea. But if what you're trying to get is an early warning sign of problems, you're only getting that once a year. When things are changing very rapidly, this happened recently during the pandemic. You don't want to wait till next year to figure out what's going on. That get the long time between them and the cost of doing this, turned out to be quite a challenge when doing these engagement surveys. A third area that I'm going to talk about at least during this course is attrition. People leaving turnover, various different ways that we can describe it. Why do we spend so much time talking about it? Firms really care about it, because it's a major source of cost for them. I've just mentioned it's really important to hire the right people. Well, if I hire the right people, but then they leave, I've got to go out and hire all over again. There's a high cost of just replacing people. But I think even more than that, there's a lot of disruption that comes from turnover. The people who work with the person who's left, they've got to build new relationships with the people they have to work with now. People who are coming to replace them, they don't understand the organization. They don't know those people. It takes them a long time to get up to speed. There may even be some very specific knowledge about our products, about our clients, that other people in the organization don't have and walks out the door when people leave. Attrition tends to be very costly in organizations and they spend a lot of time thinking about what can we do to drive that down. Now, again, there are a lot of levers that potentially we can pull to address attrition. We can think about who we hire. Are we hiring people who are likely to stay? We can think about our pay package. If we pay more, hopefully people will stay more. We think about all the things that drive engagement. If we have better leadership, better design jobs and so will people say, "There are a lot of levers that we can pull." One of the questions we want to know, which ones should we pull? Then we need to figure out, why are people leaving? Broadly I would say there are two approaches that organizations have taken to this. One is just figure out where attrition is. Organizations will traditionally track their attrition rates, they're basically calculated as the number of people who leave in a given period, maybe six months or a year, divided by the number of people who you employed at the beginning of that period. That could be a small number, it can be a very big number in some areas, say, areas of retail, fast food and so that can be over 100 percent. Again, when you track these attrition rates over time, you can start to look at trends, if it's edging up, can we do something about that? Probably more than that. By looking at where attrition is high, which departments, which jobs, which groups, we again can start to get some idea of what might be driving it. Why would it be that this department leaves more, and what does that tell us about what we need to do about it? Even more direct as a way of understanding why people are leaving, is exit interviews. Exit interviews are where people sit down with people who are leaving and ask them why are they leaving? You can get insight from this. Maybe a lot of people are telling you they're leaving because they got a better paid job. Maybe people are telling you they're leaving because another role involves more career advancement. Maybe they're leaving because their new job has a shorter commute. By looking at what reasons come up systematically, organizations are able to start getting a sense of what they might need to do in order to reduce attrition. Again, exit interviews are not perfect. There are few limitations, so one is you can't even get everybody to respond in an exit interview. If some people just abandon their job, which can happen a fair bit, you never get a chance to speak to them. Even more than that, there are big concerns that when people do these exit interviews. A lot of what they're telling you is not what's true, it's what they think you want to hear. People may be reluctant to criticize others in the organization, bring up poor management, those things, because they don't know if they're ever going to want to come back. It may be more convenient to talk about money, whether or not that's really the reason they're leaving, less clear. Although exit interviews can give you some insight, there are some real debate about exactly how seriously you should take the answers. Given the critical importance of employees to the functioning of any organization, firms have put tremendous amounts of thought over the years into trying to work out, how do we track the basic dimensions of the employment relationship? What are some of the levers that we can pull to improve it? You see some of the things they're doing. Each of these approaches has important limitations. Managing people is hard. Partly the limitations come to the fact that employees complex and everyone's unique. Some things, particularly attitudes, just really hard to measure. But there are other challenges as well. Often getting data is costly when it comes to engagement surveys, those things. Often we rely on managers and supervisors to make some of the core decisions when it comes to hiring, even when it comes to interpreting exit interviews and we worry that they're not necessarily super reliable. They may have their own biases, they may have agendas that they're trying to pursue that are not exactly the same as the organizations. Their discretion can also create, in some senses, complications. We have a bunch of things, they're not perfect. Now, I'd like to be able to say that advances in AI and machine learning, allow us to resolve all of these problems that you see with managing people. Like to be able to say, but obviously it's not actually true. A lot of these problems continue to be issues no matter how much technology we throw at them. Nonetheless, it is also the case that new technologies have created opportunities to do things differently, to find new ways to assess who's a good employee, to find new ways to assess are people engaged? Are they likely to leave, those things? They don't solve every problem, but they do provide these complementary approaches. What we want to do next is really start to explain how these new tools work, what is machine learning, and then go on to explore how they can be applied within the field of people management to develop more effective solutions to a lot of the problems we've described.
