Telling a Compelling Story with Data

Anita Brick: Hi, this is Anita Brick. And welcome to CareerCast at Chicago Booth. To help you advance in your career. Today, we're delighted to be speaking with Cole Knaflic, who's been analyzing data and using it to tell stories for over a decade through analytical roles in banking, private equity and most recently as a manager on the Google People analytics team.

Cole has a B.S. in applied math and an MBA, both from the University of Washington. Cole, thank you so much for being on the call. So excited. We're going to talk about your book, Storytelling With Data. I know you're super busy, and I know you have an addition to your family at home, too.

Cole Knaflic: Yeah, thanks for inviting me to do this. And yeah, things like this are perfect that I can sort of schedule around feedings and naps.

Anita Brick: Sometimes people get confused about data and story and how they fit together. So when you think about the best, what are 2 or 3 characteristics of the ideal?

Cole Knaflic: When story and data really come together in bringing story and data together, you get credibility. We're in a world where the amount of data is increasing constantly. With data, we can make smart decisions. So weaving data into your business case or into your project, or your communication helps lend that credibility to it. The data helps guide our decision making. Stories can help make the thing that we want to communicate memorable.

Humanize the data and help it be something that will resonate with our audience. Everybody's sat through a presentation before where the presenter puts up a list of facts or bullet points, and proceeds to read through that list of facts or points, facts on a screen, or data in a spreadsheet. These things aren't inherently memorable. Our brains aren't formed to be able to hold on to things like that.

But if instead we can take those facts or take the important pieces of them and weave them into a story that has a plot and twists and an ending. Our brains are hard wired to hold on to that. By combining the credibility you get with data, with the memorability, the ability to resonate with your audience, you get with story makes this a really powerful combination.

Anita Brick: There are a lot of questions about all right, so what are the nuances? It makes sense. The data brings credibility. The story creates an emotional connection. But how do you actually do it? There are some great questions. And the first one was from an executive MBA student. And he said I like that you emphasize context in your book and it is essential to know this. What if the group members you are presenting to have very different priorities? It really makes it hard for me to hone in on a specific context that includes all of them. Any advice? Greatly appreciated.

Cole Knaflic: The situation is inherently complicated, right? You have a diverse audience, which means it's going to be impossible to make a single communication that's going to be pointed directly at any of them. So oftentimes what happens when this is the case is we end up building a communication, you know, whether that's a PowerPoint deck or something else about our project or about our work, instead of making it about our audience.

If you really want to be considering your audience first and foremost, thinking about who they are, thinking about, you know what will resonate with them? What do they care about? What keeps them up at night? And then think about how you can frame what you need them to know or do in terms of those motivating factors. Now, it's challenging in and of itself, but it's even more challenging when you have a diverse audience who are going to have different perspectives and different things that they care about.

There are a couple of thoughts I have when that's the case that you're encountering. One is to consider whether you have the ability to split your audience into different segments and actually communicate with them separately and form different communications that are pointed directly at your specific audience for each of those different segments. Now, that's the ideal. That's not always possible.

Another approach could be to have a single communicate, but have segments within it that are pointed towards specific segments of your audience, and then you just want to make sure when you're going through it, or if it's something that's being sent out that you're making sure to call attention for a specific audience on what parts in particular they need to pay attention to.

A third strategy in a mixed audience can be to really focus on who the decision maker is, or who the decision makers are, and craft your communication primarily for them, since those are going to be probably the most important in really getting your message across.

Anita Brick: I mean, I think it's also trial and error. The more you practice, the better you're going to get it.

Cole Knaflic: Absolutely. All of this stuff is a lot of trial and error. Absolutely. Learning from some experts who can hopefully speed that along as well.

Anita Brick: Here was a question from a weekend student. And she said, and I would say that this student is pretty representative of a lot of people at Booth. I love data so much that I can lose track of time. That said, to a growing degree, my audience is much more focused on the story the data tells and how to apply this. What would you suggest someone like me start to get out of the weeds and into more strategic things?

Cole Knaflic: You have to get in the weeds when you're analyzing data, because that's how you get to know it on a level where you can use it to inform action and drive change. So that's absolutely part of the process. The risk is then not stepping back out of the weeds when you're communicating with your audience, which is what this question directly gets at. What way is you just forcing yourself after you spend so many hours or a day analyzing specific data, forcing yourself to step back away from it for a bit. Actually take a break, go for a walk, or turn your attention to a different project or read something that's totally unrelated. Give your brain a break, then start thinking about the data often.

It's great to do this without the data right in front of you. Position yourself in front of a whiteboard, or get out a blank piece of paper and start writing down the main message that you want to get out of it. Are the main things that you're learning from. The data that you want to get across to your audience can sort of help you start to frame what eventually will become the story that you tell.

With this data. You're always doing this, honing in on the details and then stepping back to look at the big picture as part of your process. The big picture may change or pivot depending on what you learn. As you're digging in deeper, you always want to make sure you're not being biased by the story that you framed out.

Let it change if the data continues to be analyzed because you are elsewhere. Another great strategy can be once you've analyzed your data a bit, you know it fairly well and you're getting ready to the point where you want to start building your communication to tell it to somebody else. Would you just grab a friend or a colleague and tell them verbally about what you're learning?

There's something about the process of verbalizing what you're learning and what you're seeing. That then makes it easier to make that jump from the weeds to the strategic, because you start to self filter and find that the less important stuff doesn't make it into the version that you're saying out loud.

Anita Brick: And that makes a lot of sense, because you're forcing yourself to move in and out of where you've been doing a deep dive. And sometimes, like you said, you just need a break, especially if you love data. Another question along those lines from an evening student. He said, While I'm principal, I love your idea of thinking like a designer, but I'm a quiet guy. What can someone like me do to bring a bit of design when I don't think like an artist?

Cole Knaflic: Don't discount your eyes' ability to recognize smart design. If you think back, everybody would draw and color as a child. And there's something as we grow up where most people stop doing that, but that innate ability is still there if you don't trust your inner design. I then collect examples of you, whether it's graphs or other sorts of communications that you see others do, that you think are effective and start to emulate some of the things that you're seeing that are effective.

And if you're ever unsure, get input from others. Ask specific questions. If you're having doubts about something having to do with your design. Or often when I'm working on a specific visual and I'm iterating to try to get it to the best that I can get it to, I'll do what I fear to the optometrist approach, where I'll have my latest and greatest version of the graph, and then I'll make a copy of it, and I'll make one change to that copy of it so I can look at the two side by side and say, which is great, which is better A or B, and continue with that process, always keeping the latest and greatest

so I can go back to it. If the change that I introduce makes it worse, and asking myself each time and using that as a way to visually see them side by side and iterate to get to a visual that's going to be really effective. Also helpful in this case can be to start with a blank template, or start with a suite of graphs that are already designed in a way that are going to be easy for your audience, in case it's helpful.

You mentioned my site before it storytelling with data.com. I have an Excel template there that has just a basic line chart and bar chart. Some of the more standard ones that have all of the nuances done when it comes to getting rid of background colors and grid lines and the things that don't need to be there. Folks can feel free to download if that's something that they find helpful.

Anita Brick: Thank you. That's great. I'm a big believer in having templates that you can always add your own touch to.

Cole Knaflic: And it's really funny because I thought forever, oh, I should just make a template for myself because I do the same things every time. You know, I change the font color here, I change the color here, I get rid of the grid line and that I save myself so much time and so I finally made the template. But I find personally I don't ever use it because for me, starting with the ugly Excel graph and figuring out, okay, now where do I want my audience to look? How do I manipulate this to make it really effective?

Anita Brick: So here's someone from the opposite side. And he said, I know I'm a bit of an outlier. I'm a good storyteller. When I'm able to keep the input qualitative, what are some small steps I can take to bring the data into the story?

Cole Knaflic: Yeah, this is an interesting question because I usually come at it from the other side, but that's awesome if you feel great when it comes to the storytelling piece, and then you want to think about where you can start weaving data into that story? One of the main things you get with data is credibility. It helps people believe your stories.

So there's definitely some positive impact that you can get by bringing data in. So it isn't obvious where to do this. Think about befriending a skeptical quant in your circle and get their input. Have them figure out where they can poke holes. Would be places where you could potentially pull in some data. Now that could be through an example.

It could be through a stat, it could be with a graph. That'll depend on the specific situation. And it's going to feel forced at first because it's not part of your normal process. Just like for those coming at this from the quantitative perspective, trying to tell a story with data is going to feel very forced. At first it will feel uncomfortable, but it's by doing it, forcing yourself to do it, practicing it, that you get better with it, and over time it becomes natural.

Anita Brick: It's a good point, my thought and jump in. If you disagree, do this where the risk is really high. Experiment in the low risk presentation when it's just three people, rather than it being a whole giant big group or we are now presenting all this to the board of directors of your company. Start with the easy stuff where few fail. It's not such a big deal.

Cole Knaflic: Yeah, that's excellent advice and especially if there are places where you could try out new things and it would be appropriate to then get feedback from your audience, is a space to start trying some of this stuff. For example, in an MBA program.

Anita Brick: Very good point. You know, sometimes even if you are doing some volunteering with a nonprofit, and I guess most of the time you're often volunteering. If you're working with a startup, those are other places to experiment and stretch.

Cole Knaflic: Out, or you're worried about doing this with your internal team and in trying out some of these things again. Yeah, where the stakes are low, if it feels too uncomfortable to do sort of on the main stage with a big audience and get more comfortable and sort of home and iterate to a place that feels good.

Anita Brick: This question has come up again and again, and I mean in other episodes, starting in the place where you can build your confidence. So one thing and it is super important, an evening MBA student really wanted to know how you applied your own approach to data and storytelling when you were interviewing?

Cole Knaflic: I thought this was such an interesting question.

Anita Brick: I mean, I think this is a question that probably everyone has this evening. Students said, having read your bio, you've had a robust career in data analytics. How did you insert data into your interview stories to really highlight key, relevant accomplishments?

Cole Knaflic: Yeah, so incorporating stories into your interviews with prospective employers is an excellent way to illustrate your skills, because you're saying, here is evidence that I can do this, but also that you can articulate it clearly, package it in a way that's going to be understood. So there's so many great benefits to this. And it's funny because when I first read this question, I was thinking about when have I done this?

Have I actually done this? Because most of my research on storytelling has come since I've had jobs. But I was thinking about my interview process at Google before I started working there. And, you know, you can always anticipate that you're going to get questions that start out something like, tell me about a time where you were, tell me about a time when you know that there would be some questions like that where they are looking for evidence and specific scenarios.

I do recall making a list of important or noteworthy projects I'd worked on, accomplishments and different things that I'd want to potentially highlight or reinforce through those. For me, I had that list in my head going into the interview process. When you've got that fresh in your mind, you can pick and choose different scenarios, different projects that you've worked on or have been through, and insert those as a way to sort of just leave into some of your interview responses. Again, it's evidence that you both have the skill that they're testing for when they ask the question, and you can talk about it in a way that's going to be effective.

Anita Brick: I know that people have different frameworks for stories, but how do you make sure that you put the data in? Because some people have crafted their stories without the data, how do you integrate the data into that framework of a story?

Cole Knaflic: Anything that you can quantify without it feeling overly forced? Because data can come in in easy ways in an interview, like, you know, how many years did you work on something or work in a particular role? Just the descriptive stats of being aware. Now you don't want to make your responses sound like a list of facts. Also, thinking about your story that you're going to tell, or some of the accomplishments that you want to talk about, or the scenarios you want to talk about.

If there are ways to restate to ensure that as well. Not all data stories have to have data in them, which sounds counterintuitive. So, for example, there was a question, something like, tell me about a time when you were working with data and realized you'd gone down the wrong path or had made an error or something like that, but there was one time I was working at Google and we were developing a dashboard.

And I'm going to be vague in this because I can't talk about specifics, and you can do that in an interview as well as say that that's what you're doing, right? They don't just think you're withholding details. This dashboard for the decision making process that we were going to be going through, we'd start it out just with boxes, placeholders for the data that we were eventually going to want to have on this dashboard.

And there were a couple pieces of data that were readily available. So we plug those in. They ended up being quite large because this is the only data that we have, and then we had these tiny placeholders for the other data. Then when we got the remaining data, we plug them into the tiny placeholders that were there. You can maybe see where this is going.

We ended up with a dashboard that had a lot of visual preference for two pieces of data that made them look like they were the most important, when really they weren't. They were only big because we'd have them on hand before we had the other stuff. So we realized this sort of late in the game, but early enough to be able to resize things and size the similarly important thing.

Similarly, so that we wouldn't be putting heated verdin preference on some of this data that wasn't necessarily more important than any of the rest of it. So notice in that example I didn't talk about any specific numbers, but I'm still talking about data in your interview process where you can think about examples of when you've worked with data, you know, whether it's making a mistake or learning something new or driving change. These are all going to be common things that prospective interviewers will be interested in knowing. So having example stories that you can talk through on hand will be helpful for that.

Anita Brick: Got it. No, that's good. I went to a presentation and maybe it's because I was reading your book to Word clutter, just like flashing in my brain as I was looking at the slides. And actually an evening MBA student asked, clutter is a big thing in your book and in life. To me, it's clear when there's too much information on a visual, how would you approach a colleague who has a messy graphic representation of the data that he and I are presenting? I don't want it to go out this way, and I also don't want to ruin the relationship. What would you suggest I do?

Cole Knaflic: You want to have a good product, but at the same time you don't want to offend your colleague. You want to approach this carefully. Depends a little bit on your relationship with the colleague. If you can get in there and make some changes directly and they're going to be okay with that, then definitely go for it. If they're going to take offense to that, which I'm guessing based on the way this question is worded, if that's anticipated to potentially be the case, then you want to be a little softer about your approach.

It can be things like, you know, maybe you're practicing going through it and you can say things like, hey, you know what? What if we try and make some suggestions? That way if things are really cluttered and you're wanting to take sort of one easy step at making them feel less so, start by taking everything that's not as important and pushing it to the background by making it gray.

You can, in a lot of cases, get away with leaving the same density of information, but making it really clear what's less important and draw the other stuff out through the use of color or bold or size to really make it clear on a dance slide where your audience is meant to look. Another strategy can be, and especially if you're dealing with cluttered stuff here, put it up on the big screen, because by doing so, things on the big screen always look more dense than they do on our desktop or on our laptops.

When you get on the big screen, it's sort of a different opportunity to be able to say, hey, and especially if this is the way you're ultimately presenting it. Hey, wow. You know, when I stand up in the back of the room and I look at this slide now, I'm not sure where to look, I think we can make a couple of minor changes to make this way clearer for our audience are a couple of strategies there, but I agree with sort of the sentiment of the question, which is that this is a challenging situation you want to approach cautiously, offering suggestions, offering to help, maybe with if there's any content that hasn't yet been finished, maybe offer to jump in and do some of that directly, then you may be able to bring your colleague around to say, oh, hey, I really like how you did that and try to understand where the density of information is coming from, because they feel like they have to have every bit of information on the slide. Is it because they're less comfortable with their tools? Right. Each of these different reasons will help you sort of get at how you might be able to change that behavior.

Anita Brick: Got it. There's a related question to that, this time from an executive MBA student. And she said, in your book you talk about pretentious attributes and that they should be used sparingly. My colleague believes more is better on making things stand out. What are good ways to choose key pretentious attributes without going overboard?

Cole Knaflic: Great question. And it's sort of almost the opposite of what we saw before. Treatment of attributes are things like color and size and the position on page ways of drawing your audience's attention to certain parts of your visual. Now, when used sparingly, they can be really excellent for focusing your audience's attention. But when they're not used sparingly, then you lose the contrast and then you lose their pretentious value.

You can imagine, you know, if you see a rainbow in front of you, it's difficult to focus and quickly on one of those colors versus if everything is gray or everything is black, and then there is a single color. It is used very sparingly for things in that single color. Draw your eye. So in any visual or on any slide, think about what is the most important thing you want your audience to look at and think about.

Using color only there and now protective attributes you can layer. You could put it in color, you can make it bold or you know, if it's in a graph, you could make the lines thicker. You can use size, especially with words. Color and size together work really well for signaling. Important depending on how dense the information is. Sometimes you can go to the next level of what's the next most important stuff, and then maybe draw slightly more emphasis there and then have everything else pushed to the background.

But again, the danger in emphasizing too many things or overemphasizing one piece that makes it look really important when it's not, you know, going back to the dashboard example that we talked about earlier, we don't put you in a situation for success. You always want to be thinking about, how can I make the information as easy as possible for my audience to get at? A lot of that is thinking about how they take in the information, and making it really clear implicitly through your design choices where they should look at it.

Anita Brick: And alumni have a question. And he said, we are not using PowerPoint slides as much as we used to. And it moved to more nonlinear presentation tools like precis. When you think of presentations like that, what visuals do you recommend that can be absorbed easily and not make the participants dizzy? I mean, some people overuse the pivot and turn and flip and all of that.

Cole Knaflic: When it comes to tools specifically, my view is any tool can be used well and any tool can be used not so well. Each of them has good features and not so good features. For me, the presentation software is simply the vehicle for getting your information up on the big screen. I always start with a totally blank slide and anything that goes on that slide, I'm adding to it. I'm not using any of the built in templates that lend themselves too easily to, you know, death by bullet point or, you know, have these crazy transitions where things are flying in and zooming out this layer of glitch that doesn't do anything to your actual story. The same thing. Irrespective of the tool you're using, you want to think about your audience and think about what you need them to know or do, and then how you can use the tools at your disposal to do that in the most successful fashion.

So build your presentation and then put it up on a big screen in an empty conference room, and stand back and click through it and observe how you're feeling when you're watching slide transitions, or when you are going through the different nodes within. And if there's anything that feels gross, take it out. For me and PowerPoint, the only acceptable animations appear, disappear and transparency goes.

I think you can use it effectively to help focus your audience's attention. And when you have a done slide and for whatever reason, attempt to reduce the density but want to be able to show your audience, that's one piece at a time and direct their attention. That way, having things appear and disappear or go through levels of transparency can be a way to achieve that. And you want to think about how you can use some of the capabilities of these tools to make your story and your data easier for your audience to consume, and not throw in this layer of glitz just because you can, or just because the tool has it there.

Anita Brick: Got it. Are there types of grass that are better for big screen presentations than others?

Cole Knaflic: What else you always want to think about, where you want your audience to look. And different kinds of graphs will achieve that in different ways. You want the graph itself to sort of fade to the background. What the grass helps you achieve is that moment of understanding with your audience. So whatever graph is going to help you best get there most quickly, want to be the right graph for you for your specific situation?

Anita Brick: Okay, good. So do you have time for two more questions?

Cole Knaflic: Sure.

Anita Brick: Okay, good. So when Alan said I'm all for data front and center, I just joined an early stage company and data is only used around financials. How would you suggest I begin to nudge senior leadership to bring data into more areas of the organization, especially as we plan and execute a go to market strategy?

Cole Knaflic: Great question. One way to start to move that direction is just have really specific examples of where you could bring data in, whether that data that you think you should be collecting, right, so that you can do something with it in the future. And then you want to think about what specifically that you could do with it in the future, and what is the benefit to the organization? Can you sort of boil it down to a cost benefit analysis. There will eventually be some benefit you get out of having data. What is the cost of that and where is it worthwhile or is it bringing in external data? Right. It's our competitor's data that would help you further plan your home marketing strategy.

What are some of the costs and benefits associated with that? And I think just having really specific use cases for what you could do if you started bringing in data in other areas of the organization, can be helpful for framing up the decision making process around doing so.

Anita Brick: That makes total sense. I think, again, it's doing it in ways that are perceived as lower risk as opposed to, okay, we're going to do all of a sudden we're going to do this and we're going to judge everybody based on this. That would just yeah, probably frequent.

Cole Knaflic: Low cost, low risk, start with the low hanging fruit. Start by asking some soft questions to gauge whether there's appetite for this before you go full bore. Have we considered collecting data on this or using data? When we think about this, is there an appetite to do so? Is it something you'd like me to research? Start softball, be an easy sell at?

Maybe like, oh hey. Yeah. You know, we've never actually thought of that. That's an interesting idea. Why don't you go do some research and report back to us, and we'll figure out where we might want to do that. Or if you're finding there's more resistance, then you want to see if there are some friendly or more accepting folks at the organization, especially if they're influential folks who you might be able to get on board.

Or again, examples from other companies can sometimes be useful. Part of it is thinking about, well, why isn't there data used in other places today? Is it just because the company's been focused elsewhere and hasn't had time to think about it? Or it's grown so fast and hasn't had time to think about it? Is there anything specific that's keeping the organization or individuals from wanting to pull data in? And sometimes if you can get to the root cause first, they'll help you figure out how to best position yourself when you want to start having some of those conversations so that you'll be successful.

Anita Brick: Not only as you were talking, it could be that they never had the skill or the talent to do that.

Cole Knaflic: Think that's often the case, right? Nobody's been thinking about it or nobody's had the bandwidth to do anything about it.

Anita Brick: And maybe that's one of the reasons why they hired the alum. So we'll have to find out.

Cole Knaflic: There you go.

Anita Brick: Hey, there we go. So I know we've talked about a lot of things and thank you for sharing your insights and your wisdom around this. So what are 2 or 3 things that someone can do to bring data and that compelling story together?

Cole Knaflic: Number one, and this holds true for any type of communication, whether it has data in it or not, would be. Start by pausing and consider your audience. Think about who they are, what motivates them, and how you can really frame your communication to meet their needs. Really, everything we do when it comes to communicating with data should be for that audience.

You want to keep them in mind throughout the design process and in general, try to make things easy on them. It brings me to my second piece of guidance, which is to use color intentionally and sparingly. Color out of all the attributes, when used sparingly, is probably your most powerful tool for directing your audience's attention. Do you want to think about any graph?

For any slide, think about what's most important on that page or on that slide on that graph, and use color there, and you can push everything else to the background by making it gray or making it black, and just use color really sparingly to really show your audience where to look. And then thirdly would n't ever just show data.

Always tell a story. We would at least put the words around it to make the data make sense. So every graph needs a title, every axis needs a title. Even better, if you add words that say here, the audience is what you're meant to get out of this graph. You don't want to assume that two different people looking at the same data visualization are going to walk away with the same conclusion.

It means if there is a conclusion you want your audience to reach, which there should be. If you're at the point where you're communicating to them, then you want to make that explicit by stating it in words, either in your voiceover or on the communication directly, stating by doing those three things, thinking about your audience, keeping them in mind throughout the design process.

Using color to help direct your audience where to look, and then using words to explain to your audience why they're meant to look. They're right. What's the context with the takeaway? What actions should they take based upon this? You really put yourself in a position for success when it comes to communicating successfully with data.

Anita Brick: It's great. The whole field is still very much in the nascent stage, and it's nice to see that you are in the forefront of this, and your book really does exemplify what you're talking about. It's clear, it's concise, and the visuals and the text go very nicely together. So thank you for writing it.

Cole Knaflic: Yeah, I had fun with it. I mean, it was after working with hundreds of different companies and realizing, hey, the core lessons are always the same. And, you know, everybody can learn from the examples from the various industries because it's the same underlying principles that's driving all of them.

Anita Brick: Well, thanks for doing it, and thanks for making the time to be with us today.

Cole Knaflic: Absolutely. Thank you for inviting me.

Anita Brick: Of course. Back to the template. There is an Excel template on the website, which is StorytellingwithData.com. And there are other things there too. So if you get a chance, go take a look and thank you all for listening. This is Anita Brick with a CareerCast at Chicago Booth. Keep advancing.