189. Numbers Need Narrative: Use Data to Influence and Inspire

[00:00:00] Matt Abrahams: When it comes to communication and persuasion, numbers can be your friend. But you don't want to use numbers in a way that numb your audience. My name's Matt Abrahams, and I teach strategic communication at Stanford Graduate School of Business. Welcome to Think Fast, Talk Smart, the podcast. Today I am excited to speak with Miro Kazakoff. Miro is an author, entrepreneur and senior lecturer in managerial communication at the MIT Sloan School of Management. He is the recipient of the MIT Sloan Teacher of the Year Award and is also the author of Persuading with Data, A Guide to Designing, Delivering, and Defending Your Data. 

[00:00:41] Miro, welcome. I look forward to our conversation. We've known each other for a while and it's great to finally have you on the podcast. 

[00:00:47] Miro Kazakoff: Yeah, I'm so excited to be here. 

[00:00:48] Matt Abrahams: Great. Should we get started? 

[00:00:49] Miro Kazakoff: Absolutely. 

[00:00:51] Matt Abrahams: So, data are often seen as objective, but you argue that they can be a powerful tool for persuasion. How do you recommend using data to craft a compelling narrative that influences the decision makers we communicate with?

[00:01:06] Miro Kazakoff: So data is objective, but fundamentally people are not. So even though I believe that objective reality does exist, whenever you are communicating, you are going through some sort of subjective filter that people are putting on all of the information that's coming into them. And so the fundamental thing, and I think almost all your guests bring this up, is that you have to have the right answer, but the key is going to come down to understanding your audience. So I actually think the way to craft compelling narratives is actually what it starts with is empathy. But it starts with your ability to understand for the people who are your audience, who are they? The best way to find out about that is to go ask that person. And I think a little bit of the mistake people sometimes make is they go ask the person, oh, what do you want to see? And what should this data look like? 

[00:02:03] And sometimes that can work if the person really has an understanding of how they take in information. But the questions that I prefer are questions like, what kind of decisions do you make? How do you make those decisions? What would you need to know in order to change the decision that you're making? And that the people who get good at this is actually not so much the people who can talk and draw graphs well. But actually the people who can listen the best and who can best use that information to understand, oh, okay, this is what's going to get through to that person. 

[00:02:38] Matt Abrahams: So, as you mentioned, knowing your audience is a theme that goes across many of our episodes in the power of listening as well. So, what I'm hearing, and not necessarily what I expected to hear from you in terms of this answer, is that the best way to be persuasive with data is to best understand how your audience makes their decisions. And then think about what are the best tools and information to provide, and that'll guide you in the direction of the data to use. You know, one of the challenges of using data in communication presentations is making complex information understandable. You can numb people a lot with a lot of numbers. What strategies do you suggest for making data accessible without losing the impact or negatively influencing your credibility? 

[00:03:26] Miro Kazakoff: One of the first things is to recognize something called the curse of knowledge.

[00:03:30] Matt Abrahams: Yeah. 

[00:03:30] Miro Kazakoff: Which is this idea that once we know something as people, not like we've memorized a fact, but like recognize a pattern or understand how something works, that it's very hard for us to understand what that looks like from an audience point of view. And to just know that that affects all of us all the time, it gets worse the more you know about a problem. And so it means that understanding how this is going to look to your audience is actually a really challenging problem for everyone, and it cannot be ignored. And the depth of your knowledge doesn't help you see it better in any way. If anything, it helps your ability to solve the problem, but it hurts your ability to see what it looks like from the audience.

[00:04:15] And so the key to it, I think, is other people. Is actually getting, and especially with data, getting it in front of other people. And what's neat is it can basically be anyone. So when I work with companies, what we try to do is we get together people who don't work together on the same problem day to day, and we give them the skills to help analyze each other's work. These coworkers are sophisticated people who understand how the industry works. They're often, you know, scientifically or have deep engineering training. And when they say, oh, I don't quite understand what that means in this context, that can be really helpful for people to see. All right, I see why you wouldn't understand that. And so it turns out that almost anyone can do it. 

[00:05:03] Matt Abrahams: The curse of knowledge and the curse of passion, I would argue, loom large in any communication, but especially when you're using data. I like to say that the only antidote to the curse of knowledge and the curse of passion is curiosity and empathy. You have to be curious enough to think about the other person's perspective and empathetic enough to want to do something about it. And you highlight that, focus grouping, testing, checking it out, all really important. We have choices to make when it comes to dealing with data. We can present that data in lots of different ways. Have you found in your work, in your research that certain ways resonate with people? For example, is it a visual representation? Is it actually showing the numbers themselves? Are there specific bits of advice you would give for how we actually show the data to make it impactful? 

[00:05:50] Miro Kazakoff: The one thing about showing the data that I think goes across everyone is the right way to show the data is very, very clearly. Which unfortunately doesn't provide you a lot of practical advice on what you should actually do on any given day. You can think of a couple of levers. Probably a table is often going to be the best way to show data to them. It's a very dense way. This is like rows and columns of numbers. It's a very dense way to get a lot of information across. And then the mistake that people make is as they talk to people farther away from that immediate work group, like senior managers in the company or outside stakeholders, to think that that same model is going to work because it worked with people who know this problem really well. 

[00:06:36] As you move away, you're almost certainly going to want to move to graphs, which are better at a couple of things. Graphs are really good at showing comparisons quickly. So actually looking at the number fifteen and the number thirty, very quickly we can know that those are double. But it is way faster to look at one bar that is twice the height of the other bar and to see that that number is double the other number. And people worry that you lose precision when you do that. But the farther away you move from people who understand this problem in detail, the more the speed and the clarity is important. And whether it's thirty point one and fifteen point three, they're not actually super relevant. 

[00:07:18] Matt Abrahams: Right, and in a moment we'll talk more about the story you tell around the data, but you can make that difference known through the story as long as people see visually the distinction. Before we get into the storytelling aspect, I'm curious to hear your thoughts on context for data. Because people throw out a lot of numbers, they do a data dump. But they don't necessarily give the context that helped that information mean something. What are your thoughts or advice on context? 

[00:07:47] Miro Kazakoff: Here's my rule on context. And I've cribbed this rule from Barbara Minto, who developed something called the Minto Pyramid, which is a way that I often teach about how to sort of visualize logical argument. Which we will talk about in a moment when we dip into storytelling. Which is that context should only be things that this audience already agrees are true. And I'm talking about the context at the beginning of the presentation. That one of the mistakes that people make is at the beginning, they either introduce things that are not widely agreed on. Like maybe in your work group, we agree that that radio marketing campaign last quarter wasn't very good.

[00:08:29] And so this analysis is about what should we do instead. You walk into the marketing group that ran that radio campaign, you cannot just be throwing that off casually to dive into what we should do instead, you actually need that analysis for them to focus on whether or not that campaign succeeded. So there's one part that's context at the beginning, is things that we all agree on. And then context should come as close to the data that it applies to as possible.

[00:08:59] So because I think of academic norms, people want to load in a bunch of context in the beginning. Both of things that the audience doesn't agree to or information that's going to be important later on in the presentation. And what you're asking your audience to do is to load all of that information into their brain and hold on to it for one, three, ten minutes. And there's just frankly almost no chance that what you're doing is so important that they should do that. It's just, it's not a reasonable thing to ask of folks. 

[00:09:30] Matt Abrahams: In many ways you're distracting them because you're asking them to hold something in memory while you're giving them important information. So what I take away from that is timeliness of context setting is really important. And I want to make a distinction, there's the context that's sort of the psychological readiness piece, which is what you're talking to. I think there's another form of context, or at least I use that the word context to identify this, to help people just grasp what you mean. I'll give you an example. Many, many years ago, I worked with a very senior leader of an international bank, and as they were presenting, they gave the data, the number of how much money goes through their bank every day. Astronomical number. I'll never see that much in my bank account. And I asked, what does that number mean? Well, they did some quick math and came back and said, It's roughly twenty-five percent of the world's money. That context for that data helped me appreciate the magnitude. So there's that other level of context as well. Do you have thoughts about that level? 

[00:10:31] Miro Kazakoff: Yeah, one of the biggest things that we teach people, and it's the thing that people most often come back to me and say this was game changing, is a way of when you present slides in the workplace and graphs, which is to orient. And then actually orienting is the most important part of explaining data. And it's totally counterintuitive that a good orient is where you actually narrate the audience through every single visual element. What does the X axis mean? What does the Y axis mean? What are all the colors mean? What does every single element mean? So that the audience can visually process the slide, audio process the slide. Because if it's on the slide, the audience is going to try to process it, it's just how brains work. 

[00:11:18] And in that sense providing context allows, when you're through all of that at the end, for everyone at the same moment to get the critical piece of information. Which is now that I understand what's being displayed, what does that mean? And it means revenue is declining or our segmentation, you know, model needs to be changed. And then what people think is like, well, that's way too much. And the answer is, that's because you're putting too much information at once on the slide. And if you actually have this discipline to be like, when we get there, at the right time, I'm going to go through everything, it's going to give you the critical discipline to put up less. And to really focus, coming back to the audience, on like, actually what's the part here that the audience needs to know. 

[00:12:06] Matt Abrahams: That advice I think is so wise. It's, in many ways I think in analogies, you're a tour guide. And you're taking your audience on a tour of the data that you're presenting, and you give them the information that's relevant in that moment to help them understand. This is not a conversation about slide design, but I want to put an exclamation point after what you said. The biggest issue I see when people craft slides is there's just way too much information. I think the mantra we all should have in slide creation is less is more. And we should challenge ourself to put the least amount of information needed, that's needed. Because you can overwhelm people. I really appreciate that advice and that notion of orientation is really key. So we've been teasing and promising to talk about storytelling. So let's do that. You emphasize the importance of storytelling when presenting data. How can we turn raw data into a persuasive story and what are the key elements that lead to success in a data driven story?

[00:13:01] Miro Kazakoff: The thing to focus on first is getting focused on what's relevant. 

[00:13:05] Matt Abrahams: Right.

[00:13:05] Miro Kazakoff: Having really clear logical support. And so one of the tools that, that we teach, I talk about in length in Persuading with Data, my book, is this idea of a Minto Pyramid, which is, there's many tools to do this. But this is one really flexible tool to just get your logic really solid and clear. And then when that's clear, that's the time to bring in specific examples in the form of story. And certainly when you're talking to huge audiences, like if you're the CEO of a company or you do big things, you're almost entirely going to go on story. And, you know, if you're talking to maybe a deep scientific technical engineering audience who literally works on this problem every day, maybe you're not going to go into story at all. But for most folks trying to make quantitative, well grounded decisions, the needle's going to point a little more towards that technical end and that actually their problem isn't that their story isn't exciting. It's that their logic isn't clear. 

[00:14:07] Matt Abrahams: I appreciate that distinction. Before we end, I'd like to ask you three questions. One I create just for you and the other two are similar across all the interviews. Are you up for that? 

[00:14:16] Miro Kazakoff: Absolutely. 

[00:14:17] Matt Abrahams: So, I assume you're like me and you have some pet peeves or concerns. Is there one issue or pet peeve that you have around how people use data and tell persuasive stories that we haven't discussed, that you would like to maybe use this as therapy and have a little catharsis? For me, for example, it's people who give lots of data serially, like one after the other and make no connection among the data. And then you're left there going, what's the point you're trying to make? 

[00:14:46] Miro Kazakoff: The thing that is bugging me the most these days is a combination of two things that we brought up. But I see them united together so often, which is the problem of storytelling when the data isn't right. And going and using the story of how I did this analysis. And so people, because they hear that storytelling is really important, will try to find some sort of story that isn't relevant to me. And then it involves them going into a level of detail that sort of doesn't, that doesn't serve. And missing critical logical information at the end. So the other thing that I think often goes with this is in order to drive the story, they have these headlines we were talking about that go beyond the data. And I cannot stand when I give a piece of feedback or I see something and someone goes, but it's a good story. That is when I just, if I had enough papers to throw them in the air and like sulk out of the room, I would do that. It's unprofessional. So I don't, but in my head, that's what I'm doing. 

[00:15:56] Matt Abrahams: So stories are used in lots of venues and there are some stories where their connection with reality and what is shown in research is not as important as the stories that we're using to make business decisions to actually guide strategy. And so I totally agree. When somebody just creates a story that's not based on the data, it can be frustrating. Let me ask you a question number two, who is a communicator that you admire and why?

[00:16:25] Miro Kazakoff: Michael Lewis. Writer of books like Money Ball and The Big Short. And the thing I love about Michael Lewis is that his prose doesn't call any attention to itself, that you feel like you have this clear window to the thing that's happening on the other side and the prose isn't in between you and the thing that he's describing. And I actually think of really good data slides as being like that. They don't call attention to the data. They're a window to the phenomenon that the data is trying to show. That the best version of a data slide is just clear as glass, like the prose of Michael Lewis.

[00:17:03] Matt Abrahams: I love how you said it gives you a window to whatever it is you're discussing and you don't get caught up in the way that it's being presented. And that's in terms of the words you tell in the story and the data that you use. Question number three, our final question. What are the first three ingredients that go into a successful communication recipe?

[00:17:21] Miro Kazakoff: And again, my point of view comes from sort of analysis and things. I think the first one is assume you're blind. Assume that what you see is not what other people see and what you experience is not what other people experience. And then with that assumption is the second ingredient. Which is the classic one, which is to know your audience, because if you know that you're not experiencing it like they are and that will help, you know, you need to go spend some time understanding how they experience it. And then for data, my third one is focus on structure. Get your logical framework right, get focused on what this audience needs to know and really clearly what is the evidence that is needed to support this fully, but not beyond that. And if you have that, then the graphs become easier to draw and the tools like data in ratio become easier to draw. And there's plenty of books that can sort of teach you those things. But really, the fundamental insight is knowing you don't see it like them, understanding who your audience is, and then getting the underlying structure correct. 

[00:18:31] Matt Abrahams: And the only way you can get that structure correct, I would argue, is knowing your audience and starting from a position of, we're not the same, there is difference here. I appreciate those three ingredients and can see how they would help not just with presenting data, but with communicating anything. Miro, this has been fantastic. I knew we would have a lovely conversation. The detail and specifics about how we can think logically and present data in a way that supports our point of view, sometimes using story, sometimes not, really helpful. I appreciate your time. 

[00:19:02] Miro Kazakoff: Thank you so much. It's been such a pleasure and so much fun to be here. 

[00:19:06] Matt Abrahams: Thank you for joining us for another episode of Think Fast, Talk Smart, the podcast. To learn more about communicating with data, please listen to episode 49 with Chip Heath. This episode was produced by Jenny Luna, Jermaine Hamilton, and me, Matt Abrahams. Our music is from Floyd Wonder. With special thanks to Podium Podcast Company. Please find us on YouTube and wherever you get your podcasts. Be sure to subscribe and rate us. And follow us on LinkedIn and Instagram. And check out FasterSmarter.io for deep dive videos, English language learning content, and our newsletter.