Data Masters Podcast
October 14, 2020

It's about solving a business problem, not the data

Kathleen Maley
Former Head of Consumer and Digital Analytics, KeyBank

Kathleen Maley was most recently Head of Consumer and Digital Analytics at KeyBank where she led a large team of analysts who used data and analytics to shape strategies. Before that, she spent more than a decade at Bank of America in a variety of roles that helped the bank use data in areas such as risk management and pricing.Kathleen shares why she stopped using the term data driven, why aligning around a business problem is key for data and analytics teams, the role language plays in communicating with business colleagues, and what she’s learned as a female analytics leader.

I'd rather read the transcript of this conversation please!

Anthony Deighton: This is Data Masters, and I'm Anthony Deighton, Chief Product officer at Tamr. My guest today is Kathleen Maley, a leading thinker in the area of analytics and data strategy. Most recently, Kathleen was the head of consumer and digital analytics at Key Bank, where she led a large team of analysts supporting the bank and making data driven decisions. Before that, she spent more than 11 years at Bank of America and a variety of roles that helped the bank use data as a core asset in areas such as risk management and pricing.

Anthony Deighton: Some topics we'll discuss today include how analytics and business teams can work effectively together, the importance of language for analysts as they engage questions around data, and what she's learned as a female leader in analytics. Kathleen, welcome to Data Masters.

Kathleen Maley: Thank you so much. It's a pleasure to be here.

Anthony Deighton: So maybe to start us off, and maybe by way of introduction for the audience, maybe you could share a bit about your background and experience in moving from being a statistician to a model developer, and then as an analytics leader.

Kathleen Maley: Yeah, and I'm actually going to back up one step if that's okay, because I think my first career has actually been very, very useful. As I've made that journey from statistician to analytics leader, I was a teacher. And what I've discovered is that so much of what I do now is about teaching, communicating, meeting people where they are so that they can receive the benefit of what I do. So yeah, I decided after teaching for a number of years that I wanted to go back to graduate school, I earned a master's degree in Applied Statistics, decided I wanted to do something with it. I was invited by Bank of America to join, started in model development, and thought that was the cat's pajamas. I was later asked, I was bribed really, to take a job in model risk control. And I said yes, because it was going to be my first executive level position.

Anthony Deighton: So Kathleen, I really love this idea of moving from educator to practitioner to the work of doing, and now in a way you're moving back to educating to bringing this knowledge you've gained back to the industry, would that be fair?

Kathleen Maley: That is exactly right. And I see it in two streams. And what I learned first is that we analysts leave school, not having a clue how to be useful. We know our algorithms, hopefully, we know some basic statistics. I personally, as a statistician, think that's very important. We know some programming, we know how to get to our data, manipulate it, create some views that help us understand it, but we don't know how to be useful.

Kathleen Maley: My entire job is supporting the objectives of a business leader. How do I learn his strategy? How do I engage with him using language that is familiar and comfortable to him? How do I approach him in a way that is inviting instead of in a way that might put him on the defense? Those are all really important things for doing this job that nobody ever teaches us.

Kathleen Maley: Also, sort of what's the point of the job? The point of the job isn't to create charts, the point of the job isn't to create a model, the point of the job is to teach my business leaders something about their business they thought they knew but maybe didn't have quite right. It is to give them a tool that allows them to make the same decision over and over in an automated yet safe way. It is to give them tools that answer the questions, the same questions they'll have over and over. It's never about a dashboard. It's about answering a question. It's never about what's interesting to see, it's about what's useful to know. How do I reorient my thinking around that purpose? So that's the first stream of work, training my analysts how to be useful.

Kathleen Maley: What I discovered in the last few years is actually there's just as much work on the side of the business. Do they know that, that's my job, or do they think my job is just to get the data? When I've given them a solution, an algorithm that needs to be implemented so they can leverage that algorithm, is there a set of expectations that goes with that, that actually the work they have to do, the change management, the adoption is actually much more than what I had to do to build the algorithm? And do they know what some of those key milestones are?

Kathleen Maley: So it is just as much about teaching my analysts how to be useful as it is teaching the business how to leverage this really special asset that they have. And that's something I've been working on, in particular in the last few years.

Anthony Deighton: Yeah, this reminds me many years ago, I had the pleasure of dining at a very famous restaurant in Chicago called Charlie Trotter's. And actually, it's like, weird turn of events met Charlie Trotter himself. And he said, something that's always stuck with me, he said, "I know more about food than you do. And my job as a restaurateur is to have you, is to provide for you an amazing dining experience, right?" So and it's his way of saying what you're saying, which is, you know more about this than the business leader does. But it's your job to guide the business leader through the process to be almost like a Sherpa on this journey of understanding their data and making better decisions with it.

Kathleen Maley: Yes, but I'm going to extend your story there. If I absolutely cannot stand chicken, he might convince me to give it a try. But if I really don't like chicken and what I really want is steak, and I make it really clear that what I want a steak and he brings out chicken, I may not like that dish. And so he's got to work with me. If I want chicken, if chicken is what matters, give me chicken. But if I want steak, give me steak.

Kathleen Maley: And so again, it goes back to I support the objectives of my business leader. I don't tell my business leader that his objective should be other than they are. [crosstalk 00:07:02] more to that story when you talk about influence and what I'm seeing in the data that he may not. But yes, I know absolutely the data better than my business partner ever will. He or she will always know better what they're trying to accomplish with it.

Anthony Deighton: So we often hear this framed, when a business leader comes to an analytics team, and we'll say, "Can you get me data on this? Can you get me data on sales, or I really need the data about what's happening in the southeast region." But you have some really specific steps and really specific ideas about how an analytics team can respond to that, that's the person at the restaurant saying, "I want a hamburger. Give me the thing I desire," as opposed to, here's, "I'm hungry, and I really like these kinds of foods." Right? But maybe help with, what are the steps that you suggest for an analyst team to address those kinds of questions?

Kathleen Maley: Yeah, the first thing is, no matter what my business partner tells me, and I have this conversation with the analysts who work for me, no matter what my business partner asks for, I literally interpret it in my mind as an invitation to help. I don't care what they say, what I hear is, "Kathleen, I need your help with something." And so when I hear, "Kathleen, I need your help with something," what I'm able to ask then is, "Help me understand what you're trying to achieve, so I can better meet your need."

Kathleen Maley: And when I'm working with a new analyst, when I've taken over teams, or when I'm building teams, one of the first things I say is, "When your business partner comes to you and says, hey, Kathleen, can you get me this data?" Because by the way, this is not a criticism on them. Most analytics teams have trained then that, that's what they're supposed to ask for. They're supposed to ask for the data. Doesn't work as well, but that's what they've been trained to do.

Kathleen Maley: So they come and they say, "Hey, I need this data." I'm not allowed to say yes, no matter what. If you have to pretend that you are losing connection on your phone and hang up on them, I would rather you do that than say yes. You absolutely cannot say yes. But you also can't say no. And these poor analysts will look at me like, "Kathleen, you're asking the impossible. I'm not allowed to say yes. I'm not allowed to say no, what else is there?" Well, help me understand what you're trying to achieve, so I can better meet your needs. It's an invitation to a dialogue.

Kathleen Maley: I also, because this is sometimes again about changing muscle memory. If I believe my job is to say, "Okay, what data do you want? How do you want it cut? Over what time period? Do you need a trend? Is a snapshot okay?" Then those are the questions I'm going to ask. If I'm now being asked to understand business strategy, if I'm now being asked to respond, help me understand what you're trying to achieve, so I can better meet your needs, I need a different set of questions to follow that up with. So the new set of questions isn't about the data, the new set of questions is about what this individual, this business leader is thinking. Are you worried about something? Do you have a hypothesis? Was there a debate about something? Did you see a new piece of data that doesn't appear to align? It's all about getting into their head.

Kathleen Maley: And then it's about in some cases, checking where we would go. What would you be able to do with this? If I provided you this, would it be useful? How would it be useful? It's all about uncovering what they're thinking, and where they will go with it so that I can really go along that journey with them.

Anthony Deighton: Yeah, what I think's really powerful about that is the idea of doing that right up front, at the beginning, creating that goal alignment, even before you begin doing data extracts, and trying to manipulate the data into a good place.

Kathleen Maley: Absolutely. Because how will I know what data I need to pull together, or even how to pull it together. The same data can be presented in multiple different ways. It doesn't matter. It's not about the data, it's about what's the question I'm trying to answer. Am I looking for an opportunity, or am I afraid of a risk? And that tells me do I worry more about false positives, or do I worry more about false negatives? It's exactly the same data. How do I orient the question ... Sorry, how do I orient my data, which is what I use to create an answer around the question that really matters?

Anthony Deighton: Right. So creating that goal alignment up front, and then really being crisp and clear on the business questions, which then lead you to maybe the data questions, and frankly, the challenge and hard work of, collecting that data, bringing it together, finding ways to clean it and present it in a way to address that question. Maybe that's a good segue to this question of language, because one of the things that strikes me as you talk about this, is, you're talking more like an English professor than a statistical professor, meaning the language of this is so important, it's so important to ask questions in the right way, and ask the right questions. And maybe you could share a little bit more about the language of it.

Kathleen Maley: Yeah, that's an interesting one. And you're right, I believe very, very strongly in the language for a number of reasons. There's some psychology to it. So am I putting my business partner at ease? I'm I making it really, really clear to my business partner that she still owns her strategy, I am not here to challenge her strategy, I am here to understand? And then to provide ways of helping that individual shoot higher, and get there faster.

Kathleen Maley: It's so important that my team feel at ease, that they know I've done the job, that they know, I know what I'm talking about. But I think more than anything, the importance of language isn't about the language itself, it's about the thinking that is involved. We represent thinking through language. For example, I've moved away from the phrase data driven, data driven strategy, data driven culture. It's not data driven, it's business driven. We're using data to support that business, but it's still business driven. It is about the outcome. If I spend all of my energy talking about the algorithm I created, then we still have no idea what we get for it.

Anthony Deighton: Yeah, I think that's totally fascinating, this idea of, because everybody talks about this, the data driven enterprise. And your point, which I think is spot on is actually thinking about maybe a strategy driven enterprise that uses data to support the business decisions that you're making.

Kathleen Maley: Absolutely.

Anthony Deighton: Maybe just in the spirit of making this really practical for people, are there examples from your past that shed a light on this?

Kathleen Maley: My favorite is the one I've shared, right? Help me understand what you are trying to achieve so I can better meet your needs. I think the other is, I intentionally do not use the phrase analytics initiative, there is no such thing as an analytics initiative. There's only a business initiative. And I've encountered this problem. When we talk about an analytics initiative, well, when the analytics are done isn't the initiative done, isn't the work over?

Kathleen Maley: And I've had business partners look at me and say, "Thanks for the algorithm, Kathleen, but now what? All you've done is create a bunch of work for me. Now I need to change my strategy, I need to write new scripts, I need to train my sales center to do something different, thanks a lot." And it's a psychological hurdle that doesn't need to be created. It's not an analytics initiative. And you're definitely not going to be done when the algorithm is done. It's a business initiative. You want to change the way your call center operates. Okay, I can do some exploratory analytics up front to tell you where your risks are, and where your opportunities are. I can create an algorithm that routes calls differently. But at the end of the day, it's still a call center strategy. And there's a lot of work involved in implementing a new call center strategy. So the initiative is end to end.

Kathleen Maley: And I think, again, it's this unconscious, subconscious, psychological hurdles we create when we talk too much about the data as the solution. The data is an enabler-

Anthony Deighton: But the data is just the beginning of solving the problem, it's not the solution in and of itself.

Kathleen Maley:... that's right. And so reorienting, again, the way people think about this, setting an expectation, I believe in being explicit now. Sure, that's part of just because of the way my brain works. But when I enter a new organization, I tell my team, "Look, this is where you are today, this is how we're going to do it differently. And this is what it's going to feel like." I talk to the business leaders we support and I say, "My team is going to come to you with a different set of questions, please be prepared for that. Here's why we're doing it. Here's where it's going to get us." It takes it out of the subconscious puts it into the conscious. Language allows us to do that.

Anthony Deighton: Yeah, it makes it really explicit, as opposed to implicit and therefore, in the front of people's minds. Now, you've also spoken a lot about, because we've talked so far, really about the upfront piece where we're getting goal alignment or understanding the business problem that we want to address. And in the back of every analyst's mind, of course, they're thinking about algorithms and techniques and data, and how to put things together. But then there's the section of the project, as it were, where you're presenting conclusions and results and answers to those important questions. So assuming we've gotten goal alignment and now we're giving, presenting back answer. Of course, the tool of choice here is PowerPoint. And you've also spoken a lot about mistakes and ways that people could do a better job of presenting results. Maybe you could share some of that.

Kathleen Maley: Yeah. So analysts are trained in school to demonstrate their process as a way of proving, if you will, supporting the validity of what they've done, and the data they've ended up with. That's what they're trained to do. Completely useless in the business environment. So what I prefer, I tell my team, "Look, first is to know your audience," which is, am I proving the validity of my work to another analyst, in which case, I'm going to want to go through my process. What was my sample? How did I select it? What was some of the techniques I used to analyze it? What's my P value maybe?

Kathleen Maley: But a business partner doesn't care about that, nor should they, I don't want them wasting their time. If I'm doing my job, well, if I'm a qualified analyst, then they shouldn't need to. What they're really worried about is what was the question I asked him? What was the answer? So I ask my teams literally to work backwards. So think about everything that you had to do. The first thing is, you know your question, if you're bringing me a set of slides, and I say, what is the business trying to do, and you cannot answer it, you need to go back to the beginning. First thing is, what's the question? Go through your process, that's the process you had to go through. It's all relevant. It's all really, really hard and it takes a long time. But now you have an answer and what your business partner cares about is the answer.

Kathleen Maley: So you have to start with the answer. What is the so what? Why does it matter? We hear that a lot, but this idea of putting literally work in reverse order. Then, and this is the other really critical piece, only include the data that is relevant to your answer. You will have done as an analyst a ton of stuff, you will have looked at a ton of stuff, none of which matters to the answer. Only include those pieces of information that are actually relevant to the answer.

Kathleen Maley: I'm a big fan of one slide. If I cannot get my story down on one slide, I might not know the question, or I might have not actually drawn a conclusion. Analysts are also not taught how to draw conclusion, we're taught to produce data, which is not the same thing as studying data, and drawing a conclusion. Analytics is not a noun. Analytics is a verb, it is the studying of data to create information. So if I can't draw a conclusion, I can't say what's the so what, I can't say what's the headline.

Kathleen Maley: And I also want to know, how do I know if my work was well received? How do I know if I did a good job? If we're spending an hour talking about the analysis, I've not done a good job. If we spend five minutes talking about what's the conclusion, and the rest of the meeting, our business leaders engaged in where do we go from here, what's next, then I know I've done a great job, because they're not talking about my work. They're talking about the conclusion, and what it means for their business. So knowing also, what are those signals that tell me I've hit the nail on the head.

Anthony Deighton: Yeah, I think this is brilliant. And I think it also comes back to this idea that in an engagement with a business leader, you might have three months of work, and you may look through hundreds of different data sets. And the result may be, the conclusion is actually quite simple, and may only involve one or two of the data sets that you evaluated. But because we work so hard on it, we want to justify our existence, we want to show here's all the things I did. Here's how much work this took. And to your point, business leaders don't care nor should they.

Kathleen Maley: Right. Or, and I wonder sometimes, I'm sure that's the case sometimes, that I want to show my work, and it was so hard, and I want to get kudos, well, I can do that. That's my job as an analytics leader. I wonder if at times, it's, "Gosh, I'm really not sure what the answer is. So I'm just going to throw it against the wall. This is what they asked to see." I think defining the question is the most important step in the process. Because if I don't define that, I can't draw a conclusion. And now I am just going to overload, well, this is interesting, this might mean something, this could be, and it just becomes this soup of charts and tables.

Anthony Deighton: Right, which just goes back to where we started, which is if you frame the question as I need this data, versus I have a question about my business that needs addressing, then, if the conclusion is, here's a bunch of data presented beautifully, is in fact, you've answered the question. So to your point about beginning the process with a goal alignment around the business problem, that in a way forces the conclusion to be a business conclusion.

Kathleen Maley: Yes. And if I know what question I'm answering, I'm going to know what information is relevant. In some ways it solves that secondary problem.

Anthony Deighton: Yeah, again, going back to this idea that the whole alignment up front is the key piece. So we've talked a lot about the job of the analyst and their responsibility in this conversation, but the nature of a conversation is that it's a two way street. What responsibility do you put on the business leader? And maybe related, and we can maybe talk about this afterwards, but the chief data officer inside an organization, how does their role changing in this? But maybe let's start with those business leaders, again, what responsibility do they have in these conversations?

Kathleen Maley: Yeah, well, the first thing, it is a responsibility, I don't generally think of it that way, or phrase it quite that way, right? Because everything is about not me making demands, but me offering a very, very valuable, scarce resource to these leaders. And so it becomes a conversation about if we want to maximize return on investment, my team, generally speaking has a lot of growing to do, they're going to have to change the way they operate. This is what it's going to look like. And the part I need you to play to help us get there looks like this. And that's going to be ... Again, it's always an invitation, this idea of language, and emotional intelligence, if you will, which I think is incredibly important for a job like this. It's all about relationships.

Kathleen Maley: And so first is again, explicit, I'm not going to make them guess at anything, I literally have a one pager that takes leaders through, look, this is where we are today. Does this make sense? Everybody's frustrated on both sides, you're not getting the return on investment. You want your analysts to be proactive, everything takes too long. And that's usually the starting point. Well, here are the stages, here's what it's going to look like as we go through.

Kathleen Maley: And you can see in the early days, a lot of emphasis on me asking my analysts to operate differently, but guess what business partner? It's step two, and step three, this is where you're going to start to feel some differences. I'm going to ask that we have a strategy session. I'm going to ask that you start to include these team members as parts of your team. I believe in a centralized but fully dedicated model, which means that I care for the analysts' career, I care for the analysts' hard skills and soft skills, but they don't work for me, they work for the business. And so this fully dedicated model allows them to develop deep domain experience and knowledge, they know the business strategy, and so they need to be part of the team, they need to be in team meetings, they need to be part of the conversation, they need to, when there is an initiative, a project being kicked off, you would have all the key members of your team, that's going to include your analysts.

Kathleen Maley: So I talk to them about what this process is going to feel like. I talk to them about how things will be changing for the analysts themselves, and that they're going to be trying new things. I talk to them about if this is going to work, if we're really going to be able to provide value for you, you got to include us as team members, you got to open up your strategy, you got to be prepared when my analyst for the first time says, "Tell me what you're going to do with this?" It's going to feel like a challenge, because they're still learning how to say, "Help me understand what you're trying to achieve, so I can better meet your needs," but rest assured, that's what they're trying to do.

Kathleen Maley: Be open, be ready for something that's different. Call me if you ever have a concern. Don't let it wait for three months, and meet with me, be open, this is a transformation, this is a change, it's not going to be perfect from day one. So there's a new phrase that's been coined servant leader. This is how I approach the job and everything has to be in the context of, "Look, I want to help you aim higher and get there faster. Let me be your user guide. Don't touch my power tools. You don't need to run the power tools, I'll run the power tools, but you direct me to the right projects in the right way."

Anthony Deighton: Yeah, so it's about moving from being a service bureau, to a business partner, and really putting that partnership up front.

Kathleen Maley: Exactly. And you know what? Putting it in that context, they have an HR partner, they have a finance partner. This isn't different from that.

Anthony Deighton: That's actually a great point. And relating this back to the chief data officer role within an organization, how do you see that changing in the spirit of exactly what you said, creating this leadership level position around data and business strategy?

Kathleen Maley: Yeah, so this is such a tough one, because chief data officers don't agree on what the job is. And so if we who do the work, don't have a standard set of language and don't agree on what the job is, and don't agree on how to approach the job, what hope do our business partners have? So that's the first thing. I'll give you my opinion, I'll give you my [crosstalk 00:28:48] on this. But that's the first thing, I'm very sensitive to the fact that there is just not, this is not a settled space yet. We are still figuring it out.

Kathleen Maley: But I think of a couple things. First of all, a chief data officer has different responsibilities and should have different responsibilities from a chief analytics officer. Most organizations, many organizations, I should say, have made it the distance to chief data officer. And again, if we who do this work don't agree on definitions, how is there hope for anybody else? I don't think of it as the business side, the analytics side, I think of it as a three legged stool.

Kathleen Maley: A chief data officer is typically about data quality, data management, data accessibility, very, very critical stuff. Many chief data officers are now moving into the space of analytics. But often that looks like we're going to do some automated dashboards, we're going to do some automated analytics. I don't think there's any such thing, that's another conversation. For those who really and truly have a blended role, meaning they're thinking about data, and they're thinking about analytics. Those are two very, very big jobs. But I think at that level, and a strategy at that level is, how do we actually make it work? How do we coordinate around all of these different business strategies? What if we have business units that operate in silos, but we, as a centralized function, see efficiencies that reach across?

Kathleen Maley: This is actually one of my favorite things about the job, I sit in the middle, everybody needs data. And I have very, very clear line of sight into competing business priorities all the time. So part of my job is to bring these leaders together and just make them aware that it's time to have a conversation. That is a part of the role, I think of a very effective chief data slash chief analytics officer. We are that central point, we see everything that's happening, we've got to echo that back.

Kathleen Maley: My focus can never be about the things that I want to do. My focus always has to be about what is the best for the business? Where are these business leaders helping, or headed and how do I help them do that? That means I cannot have my own agenda. And I've seen both, I've seen analytics leaders, chief data officers, who are really comfortable with this idea of helping others meet their goals by aiming higher, and [inaudible 00:31:31] faster. But there is just as often this dynamic that the chief data officer has his or her own independent agenda. And I think that is where things start to break down. It's got to be always about what is the business trying to achieve, and how do I help them do that?

Anthony Deighton: Got it. And I mean, maybe to really put a fine point on it, would you agree with the assertion that a chief data officer ought to manage both things like data quality and data availability, but also the analysis and meaning making out of that data, really combining in a sense, this chief data officer and chief analytics officer role?

Kathleen Maley: I think in a smaller organization, it's absolutely a necessity. And you will probably have to have one person with that shared responsibility, and that's really tough. And then each organization has to know for themselves is our bigger problem right now, the fact that our data is a mess, and we need somebody really who understands the technology, and a strategy for around data and data management? Or do we have enough control of our data, and now we're transitioning into, now we want to monetize on it?

Kathleen Maley: Then you probably need somebody who is more experienced in data usage, creating information and meeting through data. They're very different skill sets, typically different personalities that come to those jobs as well. And so it is, every organization has to decide that for themselves. For larger organizations I do think it is time to start thinking about the difference between a chief data officer and a chief analytics officer, and move beyond the two sided coin and start thinking about the three legged stool, we need good data, we need good analytics, and we need the business leaders to have the right partners in those positions, so that they can shoot higher and get there faster.

Anthony Deighton: Right. So shifting gears slightly, a lot of what we've talked about, and what you've spoken about is the role of analysts in the context of the decision making process, the business decision making process. I'm curious if you find this as more challenging for women, because they're often the minority in analyst roles or potentially in leadership roles. And I'm sure you've learned some very practical experience as a female leader in this space, and maybe you could share some of that.

Kathleen Maley: Yeah, this is, it's a timely question, it's a difficult question, because of course, I don't know what it's like to not be a woman in analytics, because each of us only gets one shot. I can't go back and run the experiment as a male. But I'll talk about things that are maybe not directly related to my role as an analyst and just what it means to be a woman in leadership, a woman in corporate. I've had the experience where I've been shushed in a meeting literally shushed with a finger, I've had the experience where one of my business partners repeatedly referred to me as my dear, I've had the experience where a very senior executive was looking hard down the V-neck of my blouse talking to me about how much he appreciates the necklace I chose to wear. I've had those experiences.

Kathleen Maley: I've had the experience where I was told I needed to demonstrate more thought leadership. And then I was told that I wasn't needed in a strategy meeting, that they would come and tell me what I needed to do when it was settled. I've been told I could go and talk to, I made a change on my team that needed to be communicated. And I said, "This change needs to be communicated." And the answer was, "Okay, you tell that executive over there, but I will tell this executive over here." This split down the line between I was essentially told I could tell the female executives but not the male executives.

Kathleen Maley: Those things happen, but I don't think it's unique to analytics. My experience in analytics, I think I've had a bit of a leg up. My mother was a math teacher. Math was just a thing that we did. My undergraduate degree is in math. I went to a woman's college, and I walked into my first calculus class, I didn't think I would major in math. But I had a math requirement. So I walked into my first calculus class, there were 48 other women, and myself with a female professor, Mary Louise Cookson. There was one man in the class because he went to a college down the road, and it fit in his schedule a little better.

Kathleen Maley: And I walked in that room, and it was the first time in my life, I was surrounded by other women who were doing this thing. And I'd had some trouble with calculus in high school. In hindsight, there were other reasons for that. So I was retaking it, because I thought it would be an easy way to get my credit. And I walked in, and I thought, "Wait a minute, why couldn't I do this." And that was it. I mean, three months later, as a freshman in high school, in college, I knew what my major was going to be.

Kathleen Maley: My very favorite professor, Helen Grumman, taught me number theory and abstract algebra. And she is still a mentor. She's phenomenal. She's part of the reason I went to University of Michigan for my degree in Applied Stats. So I have had very, very good role models and very, very supportive role models.

Kathleen Maley: However, I was asked to attend a conference not too long ago, and I looked at the marketing materials, it was all white males. I'm fortunate that I'm bold, I'm confident, I'm willing to swim upstream for the things that are important to me. Sometimes it can get tiring to have to swim upstream.

Kathleen Maley: I remember a male executive saying about his all white male leadership team that he thought diversity was important not at that level, but at lower levels. And he wanted to see the next generation trained for those positions of leadership. And I wondered, what was wrong with my generation? So, it's a tough, tough place, because there's no good way to talk about gender differences. A woman who mentions them, is a complainer, can't handle it. A woman who doesn't mention it, or keeps it inside, continues to reinforce the dynamic that exists. It's a tough one.

Kathleen Maley: I will say, I have had men support me throughout my career in a phenomenal way. I think the men who have been most supportive are the men who've been willing to engage in the conversation and not dismiss it. If I've presented a scenario, is he willing to say, "Huh, I'm gonna think about that." He may not draw the same conclusion. But at least he's able to step out of his own experience, and begin to see this experience from my perspective, to at least acknowledge the reality of it, the differences of it.

Kathleen Maley: Because it happens, when I was just in that meeting, I was the only woman in the room, there were 20 men, no one even ... There was one other person who noticed and he came to me after he said, "I can't believe I never seen anything like that in my life." Because, again, is a woman in this position. I question sometimes like, "Was that just something that happened? Was that just him? Was that because I was a woman? Would he have done that to a man?" And so in that case, it was actually so helpful when this man came to me after to say, "I can't believe I just saw that, what are you going to do?" And he said, "I wanted to say something so bad, but I know, you and I know you're going to handle this in the way that's best for you." That was so helpful.

Kathleen Maley: I think it's most men are not misogynists. But very, very few men have walked in a woman's shoes. And so are they willing to try and experience that for a little while, or are they open enough to hear and to think differently, and to make different decisions, to look and say, "Okay, if we only have white males, is there something we can do to change that?" Because we know as statisticians, that when your data has no variability, you cannot do anything with it. We depend on variability in our data, and yet we see leadership structures that are totally homogeneous. So those two truths, again, as an analyst, as somebody who really relies on logic, those two truths are counter to each other. How is it that variability in data is incredibly valuable, and yet we don't see variability in our leadership frameworks? [crosstalk 00:41:45]

Anthony Deighton: Yeah, I think that's fascinating. Just to draw these two worlds together for a second, you make very clear this idea of being explicit about both the language and questioning model for an analyst and a business leader. I think the part of what you're saying here, which I think is very valuable, is to be explicit and create language around discussing gender roles in leadership positions, but also how it affects the role of the analyst, and to your point about being shushed in a meeting, it's that's exactly wrong, right? That's exactly what leadership needs to create a framework in which that is not permissible, so that all those voices can be heard, you can create that variance in opinion and debate, which ends up making better business decisions.

Kathleen Maley: And I think it's really, really important to not just say, "You're bad, because you did that." I think it's really, really important to begin to expose good people to behaviors that when they step back and view it from the outside, they think, "I don't like that. I don't like that I did that. I don't like that I think that way. Oh! My gosh! Yes, I do see how I'm participating in this." It's not about bad people versus good people. It's about a very, very narrow perspective and experience that happens to be in the majority and hold the power, and making a deliberate choice for these good people to say, "Wait a minute, we've got to think differently about this." And to your point, we got to make it explicit. It's not good versus bad. These are a lot of otherwise good people who just haven't lived my experience.

Anthony Deighton: It would be really helpful if you share briefly, a really practical example of where all of these things we've talked about came together in a very real world example, a data challenge that you faced anywhere in your career and life. So maybe if you could share a very practical example, and then I think that would help us.

Kathleen Maley: I've got two, one that was earlier in my career, and one that was later in my career. So I don't know if you want me to go through both of them. And the reason I mention both is, one was what really illuminated me to the fact that analysts need to work differently. The other was, we need to engage our business partners differently.

Kathleen Maley: And so I'll start with the latter. I was asked, the head of the contact center called me and said, "Hey, Kathleen, we released all of this new digital self serve functionality, we expected our call volume to come down, call volume has not come down. We have not budgeted for it and we are running really hot, we need some help."

Kathleen Maley: Perfect, I knew exactly what he's trying to do, and we'd already been working together. So he knew exactly how to engage with me, he knew what that process would feel like. I spent a few weeks doing some exploratory analysis, put to bed some misconceptions about who was calling, why they were calling, what they were trying to do, why they weren't using self serve. We identified some real roadblocks and decided that in this case, implementing a new strategy, a new call center strategy for a certain subset of callers was ultimately going to get us what we needed to take out all of these calls.

Kathleen Maley: So now he involved me building an algorithm. Piece of cake. Spent two weeks, got the algorithm said, "Here's your algorithm," and the answer was, "Kathleen, I thought you were going to help me, all you've done is create a bunch of work for me." Well, without an algorithm, if he'd thought I need to implement a new contact center strategy, he would have known exactly what to do, and he would have had his team on point to write new scripts, to figure out who was going to be in the pilot call center, make sure those people are trained, know what monitoring needed to happen.

Kathleen Maley: But somehow, because we kept talking about this analytical solution, solving it through analytics, when we got to the algorithm, there was no one there to take us to the next step. And there was a lot of frustration. And again, it may not be right, and it may not be logical, and it may not make any sense, but humans are not that way. And so we created this psychological hurdle for ourselves. We overcame it, we worked through, it could have been easier. At the end of the day, it took him about three months to implement this new strategy. Not unusually long, when you think about the pilot group and the training and all of these things. But it was worth $6 million annually. So it's worth it.

Kathleen Maley: But again, it's about how do we approach from the start, from the outset, and make sure there are very, very clear sets of expectations on both sides. Make sure actually that the business partner has staffed appropriately for the initiative, has funded the initiative, because otherwise, even if I build the algorithm, it's a zero return on investments, it's a negative. Right? If they are not prepared to take the next steps to actually implement and adopt, then we just wasted a whole bunch of time and money on developing the algorithm, even if it's a great algorithm. So that's one.

Kathleen Maley: The other was much earlier in my career. I had a business leader come to me. I was at that time doing client experience analytics. A business partner came to me and said, interestingly, another call center example. And he said, "Hey, Kathleen, I just made this change in the context center, cost me $7 million, I need you to give me the data that shows that client experience has increased." And so this was a gift. This particular request was a gift, because he did not come to me and say, "Kathleen, I need you to get me this data." Had he done that I would have given him the data because that was my job. My job was to give him the data. And I would have asked, do you want to trend it? How do you want to cut? All those.

Kathleen Maley: But he didn't, he said, "I need you to show me that what I did made a difference." And natural curiosity. Well, what did you do? And when did it happen? And how did it work? And he was very forthcoming. And I started looking at this analysis, and I said, "Oh, my goodness, I see what happened. You spent $7 million on a whole bunch of stuff that didn't matter. But the one thing you did that made a big difference was how you selected your people, and what special skills those people had that ultimately are what made the difference."

Kathleen Maley: Now, I had a very good manager at the time who said, "You do not go to him and tell him he just wasted $7 million, what you do is you say, hey, don't spend another $7 million. And don't grow the [inaudible 00:49:10] quite as quickly as you'd hope to, because you're going to want to do a few little things that aren't going to cost you any money to make sure that you actually grow it differently. Because the thing that's making the difference is different than you thought it was."

Kathleen Maley: I went to him individually and privately. I positioned it as, you're so smart. I was very young in my career. "You're so smart in how you did this. By the way, this other thing, don't focus too much on that anymore." So I had a lot of help with the messaging and bringing the story back to him. It was so all true. It was still positioning him to make a much, much smarter decision, but it also allowed him to manage his previous decision.

Kathleen Maley: I feel very good about that. I don't want to call anybody out unnecessarily because they took a shot at something and it didn't work out exactly as expected. That was a huge learning experience for me. Once I entered this field, that was the first major pivot on how to do this job well, how to engineer better business outcomes, and strengthen relationships with the business leaders who were responsible for those outcomes.

Anthony Deighton: Yeah, and again, I think what you hear in that example, and both of those examples, is this idea that the form matters as much as the function, the language of it really matters in the way to support people in their decisions, but also aligning strongly to that business strategy. I think you really hear those come through in those examples.

Anthony Deighton: So with that, Kathleen, thank you so much for your time, and this thoughtful conversation. I hope everybody has been able to take away a number of nuggets from our time together. And with that, I thank you for joining us on Data Masters.

Kathleen Maley: Thank you so much. I really enjoyed it. I appreciate the conversation. And I hope that your audience, that your listeners are able to take something away. As a teacher, that's always what I hope for.

Suscribe to the Data Masters podcast series

Apple Podcasts
Google Podcasts