Demystifying Data Storytelling: An Expert Panel Discussion

May 28, 2020

As data visualization becomes a more prevalent means of identifying insights and tracking progress throughout an organization, how do you ensure that the data you’re presenting is telling the right story for your primary business objectives, and in the right way at the right time? 

Join us for a virtual conversation and Q&A session with:

  • Brent Dykes, Chief Data Storyteller, AnalyticsHero
  • Bryant Hoopes, Principal Analyst, 33 Sticks
  • Lea Pica, Data Storytelling Advocate and Founder of LeaPica.com
  • Peter Nettesheim, VP of Data Strategy, ObservePoint

To discuss best practices and ask questions about how to:

  • Identify the valuable data stories that need to be visualized 
  • Shape the data story into a narrative that aligns with your business objectives
  • Deliver your story in a way your audience can understand and utilize
  • And more


Peter Nettesheim (00:00):

All right, welcome everyone. It's good to have people here. We know it's an interesting time, but it's fun to be able to do these things virtually. Welcome to our discussion today. Today we're going to be talking about demystifying data storytelling. And ironically, just last week I read an article that said the next chapter in analysis is data storytelling. So this seems to me to be most appropriate. So welcome. Let me introduce who we have here. It's actually a pretty incredible group that we have. First, Brent Dykes, who's the author of Effective Data Storytelling: How to Drive Change with Data, Narrative and Visuals. Brent has more than 15 years of experience, enterprise experience, with analytics at Omniture, Adobe, and Domo. So, Brent, thanks for being here.

It's great to have you. Thank you. Next, Bryant Hoops, he's been in digital marketing for almost two decades now. And, as a principal strategist and analyst at 33 Sticks, he has tons of experience. It's going to be great to get his opinion. Bryant, welcome. Thanks for being here. And last but not least Lea Pica, who is a data storytelling advocate and the founder of Leapica.com. I think most of you know her, she's an experienced digital analyst who now teaches data storytelling to thousands of digital analysts and marketers and hosts. It's a popular podcast. You guys probably know it already, Present Beyond Measure. It's the Present Beyond Measure podcast, and she's also the creator creator of the Pica Protocol for delivering data stories that inform decisions.

So Lea, thank you for joining us. So, throughout the discussion today, just make sure that you submit questions, and we'll answer as many of those as we can, as we get to the end of the discussion, but it would be great to hear and see what's available. We'll be making this recording available and so we'd love to have your comments to make this even better. But let's jump into our discussion. I'm thrilled and excited to hear what these experts have to say as by way of introduction. My name's Peter Nettesheim, I'm the vice president of data strategy at ObservePoint. I have a long history of dealing with and working with data, but I think really we're here to hear these experts today.

So I'm excited to hear from them. So first and foremost, to start off, we see data visualization and dashboards being adopted more broadly that's happening, that's occurring data is increasing and the technology handle it is, is also improving, but they're still not telling a story. And so let's talk about, to start, the difference between dashboarding and visualizations and data storytelling. So first Brent, I'm going to ask you, what exactly is data storytelling and how is it different from dashboarding and data visualizations? So Brent, what's your take on that?

Brent Dykes (03:26):

Yeah. Thanks Peter. So one of the things is obviously with data storytelling, the visual component, the visualization of data is a core element of it. But what we're really trying to do with data storytelling, that's a little bit different than the dashboard. Most cases, the dashboard's about monitoring data. And so obviously that's important to kind of spot insights or potential insights in your data, but then when we want to communicate an insight to other people, that's where data storytelling comes in. Cause we take that insight, we visualize it in a way that's going to communicate. We add the narrative element to that data in terms of how do we structure the data and then how we give the context and really lay out the insights in a way that's clear and understandable for people to follow.

Peter Nettesheim (04:16):

Interesting. So as we talk about that, we understand a little bit better, what it is. Lea, let me ask you, where does this need to dashboard and share data primarily come from? What is this all about? And we'll get into a little bit more about how to data storytell, but I'm interested in your take on where this need comes from.

Lea Pica (04:36):

Well, the way that I see the need originating is from one of two places, and that can help determine the ultimate format of a dashboard versus a data presentation, where there is either a need expressed by a stakeholder or a group saying we would like to know what's happening with the campaign. We want to know the results of an optimization test. We want to know something for some reason, so you can construct a dashboard around that or a data presentation around that. And in a moment, I have an analogy to help distinguish between those two. But then the other one is from you, the data practitioner. Maybe there's something that you've seen in your data that you feel needs addressing. There's an issue. There's a win that you'd like to celebrate, or maybe there's some decision that you want to help influence like the adoption of a new platform or trying to secure additional analytical capabilities. So understanding those two directions can help determine the sphere of influence you want your presentation or a dashboard to have.

Peter Nettesheim (05:43):

Interesting. So I what I love about what you were saying is that you seem to start with an end in mind. There's something that you have to get done. There's some information that needs to be conveyed. And with that end in mind, it turns into the data storytelling, which is different than dashboarding.

I love that. Bryant, let me, let me ask you—boy, there's a lot of information out there. And when we get into data storytelling, how much data do you need and what type of data do you need for this? And Lea and Brent, feel free to chime in as well.

Bryant Hoopes (06:14):

Yeah, thanks, Peter. There's so much data available in all of the analytics platforms. And you know, we use Google Analytics, we use Adobe analytics and others as well. And the reality is that any story can be built from one single report. And oftentimes I see, in practice, dashboards that are built that have literally hundreds of report lists or visualizations, and, you know, there's so much information there that it can be overwhelming for somebody to know where to pull the story from. And really, I always start with just the one report.

Oftentimes as we deploy data tracking methodologies, we have a question in mind, and that's why we're deploying, tracking in place. So you go back to that question and say, okay, why did we collect this data in the first place? And how is the data that we're collecting relating to that exact question? And then, that report that it generates—that one report—can then be sliced in all different ways from segments to time-based charts to any other types of visualizations where you're looking at that one report through multiple lenses or multiple different dimensions, and that then becomes the basis for starting your report or your storytelling interests. The other thing that I would say too, is that you know, oftentimes we can get stuck as analysts thinking, we don't have enough data and we need to implement something more and we need to implement something more before we start the story. And that's just a fallacy there. You're never going to have enough data. And I've yet to do an analysis myself in which I had a complete data set, you know, complete in that it was clean and it was perfect and it had everything I needed, but the reality is that the data was there. And I can use what I need to tell that story.

Peter Nettesheim (08:11):

Interesting. Lea or Brent, would you add anything to that? Would you add on top of that?

Brent Dykes (08:18):

I would say that, you know, one of the things that we do is when we're exploring the data, we want all the information, all the data that we can possibly get, right. So we can really understand the problem. But then when we transitioned from exploratory analysis with visualizations to the explanatory side of explaining what we found in a clear way, that's where we want to strip out a lot of that data, because you know, everything that we used in our exploratory phase can be noise when we go to explain it. And so there's a pivot that we go through when we're doing the exploratory, where we'll consume as much data as possible to understand the problem or opportunity. But then as we pivot to explaining it to other people, that's where we may need to remove some of that noise limit, what we focus on and not try and tell too many stories at the same time, right. We want to have one story that we're really telling in a compelling way. And the analogy that I would use is when we tell stories, data stories, we're not telling Cinderella, Snow White, and The Wizard of Oz all at the same time, right. We pick one of those stories and tell that story really well. And so whenever we have too much data that could potentially support different stories, that's where our job as the data storytellers is to strip that out and really focus on what direction we want to take people.

Peter Nettesheim (09:39):

Interesting. Lea, anything that you would add on top of that? Is noise something that you run into that, like they had mentioned, you've got to get down to the end result of what you're trying to do, and you've got to make it concise and clear, and you've got to get noise out of the way? Is that something you've run into in your experience?

Lea Pica (09:57):

Yes. And actually I have an analogy I'd like to share as well, that I think would help people distinguish between the role that a dashboard plays and a data presentation with a story. Because I find that often when we try to present a dashboard, that's where we get bogged down with the noise. So one of my favorite analogies for explaining this is that dashboards were created so we didn't have to become a car mechanic to understand the vital systems of our car and be able to learn how to make simple, very important decisions to keep it running. You don't get in your car and your dashboard displays a screen and says, "Lea's in the car. Lea would like to go to the store. Will she get there? How will she get there? What will the ending be?" No, you get in. And it says, "I'm almost out of gas, my tire pressure's low."

And I'm empowered to make simple decisions that I can do on my own without understanding. And if the dashboard's constructed well, that's what it empowers the stakeholder to do. But to bring that into a data story or a presentation, that's when you're bringing your presentation or story to a car mechanic, the data practitioners, the car mechanic—they know the ins and outs of what has to be done. And it's your role to help explain to the car owner, what has to be done to keep that car running at its peak performance, without trying to explain where the carburetor is and where the flange is. And I don't know car technology, really., but that's the general idea is—what do you need in order to convince the car owner of the absolute most important thing they need to do next to keep that peak performance going without trying to explain to them how a car is built. So for me, that's the key to understanding the difference between those two delivery mechanisms and also eliminating that noise.

Peter Nettesheim (11:56):

Nice. Excellent. Well, I appreciate your explanations on that because one thing I've noticed is as trends emerge, it's often critical to understand what we're talking about. Otherwise it just becomes a buzzword and things go out and it doesn't really take hold, but this concept of storytelling really, I think, has the possibility to impact because it's about communicating and taking action on data that comes from information. And that only happens when you get down to a concise end result and communicating in an effective way. So I love it that you guys are going down this road. Let me ask another question. So now that we understand a little bit about maybe what data storytelling is versus dashboarding let's talk about how do you even build a story? How do you start with this? For example at ObservePoint, we help marketers understand which content pieces are having a greater impact on desired results, like generating leads or something like that. And so in this example, if you found that videos have a greater impact than white papers on generating high quality leads, well, how do you even start to build a story to tell the executive team what you found and that some type of action needs to happen, or that you might need to make a shift? How do you even begin that? Lea, what's your take on that? How do you even start that?

Lea Pica (13:22):

Well, for me, one of the biggest pieces that tends to be missing from data stories when presented is the actual story. So a lot of us aren't taught to understand what the mechanics and the actual structure is of something called a narrative arc. So all stories, if they're told, well, have a natural arc to them that take the observer through a journey of transformation of where someone was and where they want to be, or where they could be or where they are. So if you're trying to prove a point that videos are better than white papers, if you go in there and just say, videos did this much better than white papers. Yeah. That's not how Game of Thrones is presented to you. They don't go, hey, this is a show about dragons. No, they take you through a journey of a character where you have a setting to start with, which is the exposition, what you are talking about. Then there's a sort of rising action. Like, "Well, we tested, we did a test and we saw some really interesting things that we didn't expect." That starts to build anticipation. Then there's the climax. Like, "If we don't start replacing our white papers with videos, we could lose this much engagement on our site," or whatever you're tracking. And then those recommendations start to bring the story to a close, with a falling action saying, "But don't worry. We have a plan for you." And that plan feeds into a feeling of resolution. Like, guys, we have this plan, now we know who's gonna take action. And you walk out of there feeling like there was a journey and a resolution.

Peter Nettesheim (14:52):

Interesting. That's fascinating that you're able to relate that to things that we know in everyday life. When we watch a movie, when we read a story, it's not entirely that different, but we tend to not do that. We just present, like you said, well, videos are better than white papers, so why aren't we doing this? And suddenly, do we even wonder why we were having a hard time getting that initiative across? Bryant or Brent, what would you add to that? How do you even start? How do you build a story that way?

Bryant Hoopes (15:27):

Yeah. I mean, I like where it's going with the Game of Thrones analogy. And I might add to that a little bit. When you see a dragon, you can either realize that dragon is a small impact to your civilization, or it's a big one. And movies and TV shows and books do a really good job of helping you understand what is the overall impact of the story that they're trying to tell. And we need to do that, as well. As analysts, we need to size that impact up for the stakeholder. Let them know that—hey, this story I'm about to tell you, as I'm framing this up, has to do with 50% of our customer base, right? That's going to grab their attention and that's going to be something that helps them understand how important it is and how impactful the data analysis that you're about to take them through is going to be to the business. Whereas, you know, if you have a story that only impacts half a percentage or 1% of overall revenue, you know, it's probably not worth your time to investigate much of that. And B it's not worth, you know, the company spending resources on fixing that, unless it's something that's truly broken. So I always try and make sure that as a storyteller, I'm giving that concept around impact and sizing and scale before I get into any part, any other parts of the actual story itself.

Peter Nettesheim (16:52):

Yeah. Interesting. Any other thoughts on that?

Brent Dykes (17:00):

No, I agree with what both the other panelists have said. And I think, you know, before you start any data story, as Bryant said, you have to evaluate, do you have an actual insight? That's a value because it takes work to create a data story. And then, you know, like Lea said, it's that arc that we need to create around the story. And I think a lot of the times, we need a hook for our data story. So that's, you know, in a traditional literary example, that would be like an inciting incident where something interesting that breaks the status quo kind of starts the journey of the character and this, in our case, it could be—wow, I didn't realize that the videos outperform white papers by so much and what's going on there.

And then all of a sudden that leads into a story. Then some data stories can be very short. You know, again, it depends on the amount of information that we have and the complexity of what we're sharing. And then other times it may take a lot of, you know, we need to share a lot of contexts. We need to build up and kind of do that slow reveal, building up to our "aha moment," as I call it, the kind of the climax of our data story. And then at the end of the data story, it may be that we don't know what the answer is. We don't know what to do with what we found. But whenever possible, if we can offer—hey, here's three solutions, you know, maybe we don't want to ditch the white papers entirely.

Maybe we want to have, you know, we want to do video, obviously promote them more prominently on the page, but still retain the white papers. Or maybe it could be another option to completely scrap the white papers and totally go a hundred percent into video. But then that might mean where we need to do additional analysis to support those recommendations, to support the options. And sometimes what we're all trying to do is facilitate a discussion with the executive to make sure that—hey, you know, let's talk through this, let's decide what we want to do with video compared to the white papers. And, you know, that may be just what happens in that case with this story.

Peter Nettesheim (19:09):

Interesting. So I like what you said there in that you weren't indicating that you're going for an end result, regardless. Which, the risk there is that you start to introduce some kind of bias or you use data to maybe not tell an accurate story or a picture. That's not what you guys are saying at all. What you're saying is you need to present information in a clear and concise way, or your message isn't going to get through. How do you make sure that you don't introduce bias or embellish too much? What's the process that you've found as you go through that, to make sure that you're telling a truthful story, but it's one that's impactful? That's organized, that gets to the point and allows people to make the decisions, but also represents logically where the data is going with things? How do you handle that situation?

Brent Dykes (20:00):

I think it's very rare that we can be totally objective, right? So we need to acknowledge what our biases are and maybe even expose them when we present our data story. Say, "You know, I've always preferred video over white papers," you know? And so that's kind of buy in, but I think we need to take a position. I think if we leave it just kind of vague, you know, just not really taking a stance, not really guiding people down a path, then your audience can be open to interpretation. They can maybe misinterpret the data—who knows what could happen? And so I think, again, we want to be truth seekers. You know, I think a true data storyteller who has integrity will seek for truth. And we'll try to provide that because at the end of the day, our end game is not to deceive or trick or get our way one time, because the next time when we come back and the people feel like—well, last time you tricked me, last time you didn't show all the right data, or, there was a lot of bias in there.

Then you lose your credibility as a data storyteller, which completely undermines your ability to tell future data stories. So I think it's always in our best interests to acknowledge bias. I mean, it's, it's hard to remove it, but always try to. You know, I'm always focused on driving action with our insights and that's where a lot of insights die because they aren't clearly explained, they don't really connect with the, with the audience, and then they don't go anywhere. They don't do anything. So I think it's our responsibility as data storytellers to make sure that we manage the bias and try to be truthful with what we share.

Peter Nettesheim (21:46):

Excellent. No, I love that. Leo or Bryant, anything you'd add on top of that?

Lea Pica (21:52):

Yeah. I must echo Brent's sentiment that I don't know anyone who's able to truly remove their own bias out of the data that they themselves compile and also the people who are receiving it. So I think accepting and allowing for those biases, the more aware we become of them, is really important. There's a fantastic book called Good Charts by Scott Berenato. On my podcast, we discussed data ethics, being an ethical data communicator, and looking at the agenda behind why we are framing data a certain way, because unless we're looking at the raw numbers, any way that we manipulate that data visually is going to create a bias of some kind. So we have to ask ourselves, could we sleep at night with the way that we're presenting this information? Could we back it up in a court of law, or if we were under cross examination, would our arguments crumble, because we're just not putting enough integrity or, you know, backing into the statements that we're selling? Just as someone said in the comments, we are selling these ideas, essentially. So really examining your own bias, which is going to be there, and also making space for the bias of the audiences, which will be there, and seeing where they meet. That's where I think things can remain the most productive.

Bryant Hoopes (23:15):

Interesting. That's fascinating. And I add to that as well, that the tools we have at our disposal today allow us to at least start the data analysis from a nonbiased way in that, you know, there's attribution modeling or there's anomaly detection or things like that that are built into our tools now, in which these tools are finding outliers for us without any bias. And then at which point we can then start doing our analysis and interjecting where we think we need to go with that. But having that capability today where, you know, 10 years ago, we didn't have those built into these tools, allows us that opportunity to be much quicker and find things at least initiating it from a non-biased standpoint. And then I also say too, that bias is not always a bad thing either, right?

The executives or stakeholders you present to have a bias, your customers have a bias, you have a bias, your boss has a bias. Your spouse has a bias, right? We all have biases. And one of the things that you can do too, and Bryant's already mentioned, and Lea's already mentioned it, as well. But I'll reinforce it in maybe a different way is that as you expose your bias and maybe showcase how the analysis broke your bias or changed the way that you looked at something, there still is a lot of credibility with those that you're presenting to in which you say—hey, I thought video was really cool because me, me, me, me, me, and although these things are about video for me, in reality, it was this other type of animation that was the case. Those are the ways that as storytellers, you build a ton of credibility and you're able to capture and win your audience over in ways that, you know, if you just ignore it and act like you're a robot, you know, they're going to see right through that.

Peter Nettesheim (25:10):

Yeah. Interesting. One of the comments that came through which I thought was fascinating as well and I agree with is that you can also list your assumptions in your gaps in the data. I think maybe it's human nature to try and look like we know everything, but as soon as we do that, I think we get into the element that you all are saying, and that is, you don't know everything, and you have to recognize that you have biases and you have to recognize that this is just how it is. I don't have an agenda here. Yes. I have an opinion, but let's have a discussion about this. And I think, I can't remember if it was Brent or Lea who said it, oftentimes it's just the introduction into it, the discussion. And to me, that's what I think effective data storytelling can do is it presents the data in a very clear way.

And you as experts can really help us all understand, okay, how do we present that? So we start the discussion. But I love what your opinions have been on this and where you've gone with that. Good comments that were coming through, as well. Before I move on to another section, anything you'd add on top that you didn't get a chance to say? Awesome. Okay. So now that we've talked about data storytelling, we've talked about, you know, what it is, we've talked about some of the differences between dashboarding and what it becomes. Now, my question gets into how do you actually deliver that story? Or, or how do you make it effective? And so probably Bryant, I'll start with you in that. What are some of these best practices when you get into data storytelling, and what are some pitfalls to avoid? Cause maybe that's the better area. I know many of us have stepped into those pitfalls. So how do you deliver the story?

Bryant Hoopes (26:58):

We've actually talked about a lot of them here in exposing bias and presenting a compelling story off of just the amount of data you need. Maybe I'll share one thing directly answering your question of what not to do in certain situations and then another more tactical or pragmatic approach and tip that I can offer as well. One of the things that I see people that are new into the industry, or kind of beginning that journey from being an engineering type, to wanting to look into the data and sharing it, is we often want to justify all of the time that we spend looking into the data.

And so oftentimes I see presentations in which there's so much data and talking through the process we went through and why we thought this way and how we thought that way. And you know, all of the sudden, you know, 40 minutes of a 60 minute presentation is your own journey as an analyst and not the data story or the journey of your customer. And that's something that you really want to avoid. If there's enough curiosity from people after a meeting of—hey, how did you come about that insight? You know, those are good side conversations to have, but don't make that a big focal point about the story that you're actually talking through and going through pragmatically or very practically. If you have the opportunity to present your information in a presentation, make sure that your data visuals can stand alone, assuming that you don't have the opportunity to present.

If you've got a good enough story, it's going to grow legs and it's going to travel around throughout an organization or even to agency partners and other people out there, so make sure that each slide on its own can stand and tell that story just from what's visually being presented, assuming you don't have the talking points there. I know that's a little counterintuitive to what a lot of people talk about from doing different presentations. But again, we're not talking about a keynote presentation at a huge event. We're talking about an analysis and a data story. And in that type of a situation, you always want to make sure that those slides can stand on their own and that presentation can stand on its own without the narrative.

Brent Dykes (29:16):

One of the points I'll just jump in here for is that you do need to be careful though, because when you are presenting, when you are there as the narrator of your data story, you don't have to have as much text because you're going to mention those key points yourself. You're going to highlight them, you know, obviously with excellent visualizations and everything else. However, when you're no longer there and you're sharing it out through email or PowerPoint then you might have to add some annotations. You might have to have more commentary because you're not there. And so the danger that occurs is when an analyst says, I don't want to create two presentations. I don't want to create a version that I present and then a version that I send to everybody. So I'm just going to use the version that has all the text and all the annotations.

And then you can face plant on that because all of a sudden now, as you're presenting this slide that has all of this additional text that you're actually now saying as well, it is going to produce noise and it's going to have a less impactful experience for the audience. So you need to be careful, and I would recommend creating two versions. One that maybe has a little bit more text in it, too, a little bit more commentary and text when you're emailing it or sending it to people to kind of consume on their own. And I would have a stripped down version that when you're presenting it physically or virtually, but you're there present, then you don't need as much as the text. And so that would be one recommendation that I would have when you're telling your data stories.

Peter Nettesheim (30:55):

Interesting. So, "I don't want to put the time in to do it," isn't a good enough excuse to not make it happen. It sounds like to me, oh, Lea, was there something that you wanted to add on top? I know that we've been discussing here. Is there anything you wanted to add on top of that? And then we'll get back to Bryant and some of these pitfalls with the rest of you as well.

Lea Pica (31:15):

Yeah. You know, I'm going to concur with Brent on this one, too, where I actually developed a method to create two separate living documents, one for a live presentation, and one for the handout, because if you're doing the live presentation slides, well, they're very simple and they require you as the narrator to deliver them, or else there's no need for you to be there and no need for the meeting, but your voice is not there and guidance isn't there when you're sending that document afterwards. But I actually, I can share this too. If you're interested in the comments, I created a process with a blog post that shows exactly how to leverage the first document to not spend a lot of time creating a whole second one. There are ways to save time with that, but it is important to create two documents that serve the needs of the environment of both online and offline audiences.

Peter Nettesheim (32:10):

Interesting. I love the idea where you're talking about "serve the need." So the thing that you're doing really has the end in mind, and if it's a communication vehicle, when you're not there to explain, well, then you better have assumptions included. You better have some followup information. If you're there to talk, then you probably shouldn't have all the information that they could just read on their own. I love where you're talking about—know your audience, know what you're doing, know where you're going with this, as all three of you have been have been indicating, as well. So I think that's great.

Bryant Hoopes (32:44):

Real quick. If I could. Yeah, it really does kind of come back to that too, around the audience. You know, I think that there are times in which, you know, having multiple documents or you know, maybe a handout or like a one sheet handout that shows kind of the key takeaways are needed. And other times you work with, you know, I work with a lot of different clients with different personalities. Maybe they don't need that. Right. And so, you know, these are all strategies to take away and hopefully what we're doing is we're giving new insights and new tools and ways to think about it. But ultimately you, as an analyst, know who that stakeholder is and you know the audience that they might share it to and what they'll need. So always keep that in mind and put yourself in that position. Lea used a good example earlier around the whole cross examination thing. Think about that. What is the cross examination you're going to get as you presenter to the peers of your stakeholder, what cross examination would they have? And anticipate that, put those into the slides themselves, or have an appendix with some additional data that you can easily pull out and actually have available.

Peter Nettesheim (33:53):

Well, awesome. Can I ask you guys, can you give me, and I know I'm putting you on the spot a little here, but can you give me an example? We were getting a lot of requests asking for examples of how to tell a story. Can you give maybe one example where you did it really well, and then an example where maybe you didn't do it really well and highlight why it happened that way and why one was good and why one wasn't great? I know I'm putting you on the spot a little, but could you talk to that just for a little bit?

Brent Dykes (34:22):

Yeah, I can, I can jump in with an example. This is actually from my book that I share, but when I was doing my MBA program, I was preparing for a midpoint presentation with a senior executive and senior VP of eCommerce. And I stumbled across some insight that showed that the shipping policy that we had at the company actually probably wasn't really based on customer feedback, wasn't really that critical or important to customers. And so I had this data point now that I felt was important, that department important for this SVP to see, but it wasn't actually directly related to my project. And I made the fateful decision to include it in my presentation. And so fast forward to the day when I'm doing this midpoint presentation with the VP, I get to the slide and this insight and he yells out bullshit and I'm like, oh crap.

Brent Dykes (35:23):

I just completely screwed up my possibility of getting an internship at this company. And the lesson that I took from that was I had taken an insight that obviously was counterintuitive to him. It was again going against the grain of what the department already had and what did I do? I failed to actually build a data story around that insight. I just kind of put it out there saying, Hey, you know, I, as a naive MBA student, I thought, hey, I found something interesting. Maybe he'll look at that. And they'll say, that is interesting. We should further investigate that. We should see where that should go. No, he's just like—oh, that's bullshit. Get it outta here. What are you talking about, intern? You know, but if I knew then what I know knew, if this truly was important, I would have talked to my manager. I would have said, Hey, you know, is this something worth exploring? Yes, it is. Okay. Then I'll build out a data story around that insight and really present it in a way that hopefully that SVP would actually consider it and potentially either explore or look at doing more investigation into the idea that, possibly, our shipping policy needed to change.

Peter Nettesheim (36:42):

Interesting. Great. Thank you for that. Lea or Bryant, anything that you can share from your experience where either you've done it well or not done it well, that would help give some concrete examples of what data storytelling is. Yeah, go ahead, Lea.

Lea Pica (36:58):

Yeah. And I'll bring it back to this thing you keep seeing, which is thinking of the end result in mind. So before I went on my data storytelling journey I would respond to the ask of just come and present the latest results every week of a search marketing campaign. So that's what we did. We had the same report every week. It would produce the same kinds of numbers, except if there was something extraordinary, like a holiday, and we would drop every number we had available. And that was the end. That was it. So I can't say that I recall tremendous action or excitement or engagement happening as a result of that. But the watershed moment for me came when I was just discovering storytelling, the aspect of that, for an example, where we saw a lot of people abandoning our site search functionality, where they would search once and they would leave.

It was the vast majority. So this was where I saw something was wrong in the data. And I constructed a story that led people through that same narrative arc that we just talked about with the intention. The objective was to get people to figure out a plan by the end of the meeting—how we can start improving that—that was my end result in mind, but the way to draw people into why they should do something, that's where those story mechanics are important tools, yet actually incorporate a story arc into its workflow in any way you are the one piecing that narrative together, and then putting in the visuals you're creating from various tools to support that journey for them step by step. And the outcomes were completely different in those two cases.

Peter Nettesheim (38:50):

Interesting. One thing that I like in what you said is that you hopped on an element of people's self-interest. People aren't going to do something just because you ask them to, it's not how it works, they're going to do it because it is in their best interest. And so when you tap into that, knowing and understanding your audience and what is driving them, well shoot, hang on to that and use that. And I love how you wove that in to the answer in the example that you had. Otherwise, it was just the same thing over and over. Yeah, exactly. Groundhog day. I love it. So interesting. Brent, what do you want?

Brent Dykes (39:29):

I noticed in the chats that there were a lot of questions around infographics and other means for telling data stories. I think you can tell data stories outside of a data presentation, data presentation being probably the one that all three of us prefer the most. But you could tell a data story in an infographic or in a report or even in a dashboard. The key thing is it has to have a linear sequence, and it has to have that narrative arc that we keep talking about. So when you look at an infographic, when you have a static infographic that you're interacting with, you think, okay, I start at the top, and then there's sections that I go through, and maybe you see scrolly-telling as well. Sometimes on media sites, they have a scrolly-telling kind of experience where different visuals will pop up with narrative. And then as you scroll down the page, different things. So there are different ways of delivering a data story. So don't feel like you're limited to just data presentations, but, you know, PowerPoint in most cases for analysts and people who work in analytics is going to be the most common way of doing that besides reports and different things.

Peter Nettesheim (40:36):

Interesting. Bryant, you had something then, and I want to give you a chance here to follow up on that.

Bryant Hoopes (40:42):

Yeah. Thank you. Yeah, the thing about tools are the tools are only as good as the humans that wiled them, or use them. And so I think they're the best answer I can give around the best tools to use for telling a story is whatever is going to help you get to the data and the insight, and then having the framework in which you explain or explore that story. And I recently was able to present kind of my concept of storytelling. And I've got really three layers to a data story. And I talked about one here already. And that's the setup where you're sizing up the impact. You know, you set up your story. The second part is the actual story itself, all of the data and the insights that strengthen it, or the persona or the customer that's going through it.

And then the last part of that, that storytelling for me is always the suggestion. What is it that you're going to do now that you have that information? And so, you know, I know Lea's got her frameworks that she has put together. Brent has his, we all have a different way that we kind of think about how do we build that data story. And ultimately you as an analyst and as an up-and-coming storyteller, will have to find what works for you around which of those frameworks and which pieces will come into play. And we're out there kind of giving you all of the different colors of the paint swatch for you to now go out there and create your own masterpieces. And these are all components that we've had to learn throughout our careers and what works for us and works for my personality.

You know, I've known Brent long enough to say that our personalities are such, that they are always identical, right? We always have the exact same approach in how we do things. And now, I'm obviously kidding, but, you know, the reality is that hopefully having these types of discussions will help inform others around different approaches that we have. And then the final thing I would say is just look for examples. There's been a number of posts I've seen in the chat here of other resources. "Data is beautiful" is kind of a trending topic. There's some really awesome visualizations that are available now. And I find that, you know, cruising all of these different areas and even Reddit areas, you know, subreddits around data visualization, you see examples that you never thought of yourself, and you think, how can I pull that in and use that for this particular story? And so you've just gotta be a sponge for as much information as possible and think about how to apply it in your instance. Right?

Peter Nettesheim (43:35):

One thing I loved what you said there, and it kind of didn't occur to me until you said it, is that there are different frameworks, but find something that's your own. Go to Lea and see what she does, go to Brent, go to Bryant, go to these people, but you gotta make it your own because it's your own personality. I love the idea that there are frameworks out there to guide you. So you're not going off the rails. But I also very much enjoy the fact that you can use these frameworks and take what works for you, because these are the professionals that have built it out. And there are some tried and true things that work that aren't personality related, so stay within the guidelines, but make it your own. I love that concept and what you presented on that. I think that's fantastic. So, all right. I'm just checking on time here. It looks like we're coming up close to 15, 10 minutes left. I think what I'm going to do is, first, Lea, Brent, or Bryant, do you have anything else you'd like to share before I jump into looking at the questions and then asking based on the comments that are there?

Brent Dykes (44:44):

Probably one thing that I would bring up that I don't think we've really touched on it, I'd love to hear the other panelists talk about this too, is that when I started my analytics career, I assumed that when I came in with a logical rational well-researched insight, that logic and rationale will take over and that good decisions will happen. All I had to do is provide the stakeholders with the insights and the logic behind whatever I'm sharing. And I discovered shortly after that, no, actually emotion is a big part of decision making. And one of the ways that I feel that we can bridge this gap that we have in how human beings make decisions based on emotion is through the narrative component of data storytelling. And so by incorporating storytelling into how we share insights, that's how we're going to see people embrace our insights a little bit more readily.

And there's lots of, you know, neuroscience. I have a whole chapter on the psychology of data storytelling in my book. Because as an analyst, I was like, okay, I trust that so-and-so says these things work, but I needed to know why. And so I devoted a whole chapter to that. You understand why this stuff works, but just to give you a sample, what a lot of the neuroscientists found is that when we are approached with data and facts we automatically kick into almost a defensive reaction. Like we don't want to be conned or tricked by the data. And actually psychologists have even seen that when we get a fact that actually disagrees with our viewpoint, it's, it's almost like we're out in the wild and we get a bear, we run into a bear and it's that defense mechanism. 

And they did the scans of the brain and they found the same reaction occurs when we get a fact that conflicts with our current worldview. And so that's hard to overcome. But what they saw is when people come with a story people are less likely to nitpick on the details. They want to see where the story goes. And they've even talked about narrative transportation, where people will actually almost enter into a translated state when they hear a story. And so the more like a story we can make our insights, like, if we could package them up so they become more narrative focused and more like stories, we're going to have greater success in engaging our audiences, being more persuasive with our insights, and also making our insights become more memorable.

Peter Nettesheim (47:27):

Awesome. Lea, you've got something to add onto that, right?

Lea Pica (47:30):

I'm so glad that Brent brought this up. It was pretty much exactly what I was going to say, which was, you know, data does stimulate the brain in some aspects, but I've found that story is the mechanic that stimulates more areas of the brain at once, then pretty much anything else you can do. So I think that's what makes story in our lives. So powerful. So when I'm thinking about the objectives of the business and the objectives of a presentation and whatnot, I try not to look at the business objectives as much as the people fueling those objectives because the business is just a machine fueled by humans with emotions, desires, fears, especially fears. So there are mechanics like loss, aversion that are more inspiring for us to act because we would rather not lose something than gain something. That's actually a more powerful motivator for action. So the questions I ask myself, when I want to help create influence is I look at the people I'm speaking to as humans and I'm asking, what's their goal, their personal career goal for this quarter, what's gonna make this year a success? How are they going to get promoted? What's at the top of their list? What's their biggest obstacle? What's keeping them up at night? Those are ways to humanize their struggles and obstacles. And then you can overlay how your data is going to help them achieve what those goals are or overcome their obstacles.

Peter Nettesheim (49:06):

It's interesting. What I like about where you both went with that is it's not necessarily that you're going to tell them a story where the answer is the thing that they like, but at least if you have their perspective in mind and are understanding where they're going and you can help lead them there, the data and information may be telling a dire story. But if you understand the individuals you're working with, you can get to an end result that helps everyone. It doesn't mean that you're always telling them what they want to hear, but you're at least understanding where they're coming from. Is that what you're saying, Lea?

Lea Pica (49:40):

Exactly. They may not like it, but if your bias is always geared towards what's best for them and the business, I think you're going to win.

Bryant Hoopes (49:50):

One really quick thing add to that—also ensure that your bias is always angled towards what is best for the customer as well, right? You have to please your stakeholders and help for your own career growth, but if you frame it around the customer, it would be very rare for somebody to push back on your insight there. So always frame it with the customer in mind, and then to augment what Brent and Lea said, embrace the emotion. We are emotional beings. You know, even Brent has emotion. He has emotion despite what he's showing here, you know? It's who we are. So bring the emotion in and make it personal when you present and you're going to have much more impact.

Peter Nettesheim (50:42):

Nice. I love that. Awesome. We only have a handful of minutes left. I'm going to jump into some of the questions and see what we have here. First. We talked about infographics a little bit before, and one of the questions is—what is your view on infographics versus data storytelling? Is it the same? Is it different? What's your take on that? Any of you want to take that question?

Bryant Hoopes (51:08):

I'll take that real quick. I think the medium in which you choose to tell your story is dependent on who you're trying to tell it to and what the purpose is. And infographics can be a really good medium in which you're showing a snapshot in time, a particular set of scenarios about the business.

And so I use infographics when I'm doing some sort of a customer persona review where I'm taking a look at a particular customer segment—maybe the loyalty user—and I'm slapping up infographic type graphics because I know that the marketing department's gonna use it. But I can't present that to everybody. So an infographic is a great medium in which that can be shared and printed and used, but it's different than, as Brent has mentioned, how as we do a data storytelling presentation, there's elements of infographics that I'll pull graphically into my data story, but it's a completely different medium in which we're doing it. So just make sure the medium matches the output and the expectation that you're trying to achieve with your audience.

Peter Nettesheim (52:25):

Cool. It sounds to me like infographics aren't data storytelling, but it's a tool you could use. Definitely a story. Okay. Yeah, go ahead. Brett, what were you gonna say?

Brent Dykes (52:34):

I was just going to say the one element that I do take from infographics are the icons and the different images and things. I bring those into my data story because I view them as kind of mental shortcuts where we can use their information clearly and quickly. And even from a consistency basis where I'm always using a certain icon to represent, you know, maybe our mobile app or a particular digital marketing channel or something like that. And so again, it's just about reducing the friction that people have in consuming your data story. And I find that images and icons and different things like that can really help bring your data story to life.

Bryant Hoopes (53:16):

Awesome. Lea, do you want to add anything on top of that? Or should we go to the next question? Keep moving. Okay. We got it. Someone asked about tools. I love how you answered before by saying that tools are just a mechanism. What do you prefer to use when you tell a story? Do you use PowerPoint? Do you use Google slides? Do you use something else? Do you have something kind of proprietary that you use? What are your preferences? I'm just curious.

Lea Pica (53:47):

Sure. So I teach in PowerPoint because I want people to be able to leverage the power of the most ubiquitous, freely available tool, and also show that it's not the problem necessarily. So when I'm constructing most of my data stories, I will put the slides and I'll even chart inside of PowerPoint for a number of reasons. However, there are limitations for certain kinds of visualizations. So I will go to Tableau in order to create when I need a little bit more creative license and flexibility, or if the data set is more complex, but I will still extract that and bring it into PowerPoint because my storytelling style follows that narrative arc, where I might add element by element to a slide and reveal it piece by piece to show—well, first we saw this, but then we were really surprised by this—and you've revealed another element of the chart—and what we found was if we don't do something about it, this might happen. And then you agitate that problem and you create that climax for them. So in terms of presentation tools, the only other one that I would consider using for the actual slide piece is something called beautiful.ai. I've been testing out. It's actually the one that I've seen that enforces design best practices, pretty rigorously, from the paradigm that we, the panelists, have probably learned from. So I would experiment with that tool, but I do really believe it's important to get a basic design foundation in place to understand how to create in any tool.

Peter Nettesheim (55:31):

Awesome. Bryant or Brent, what do you guys use?

Brent Dykes (55:37):

I use almost exclusively PowerPoint, and then where the visual stations are created could be Tableau or Excel. And you could do it in power BI or whatever BI platform that you have. The other presentation tool that I've seen used by many people is Prezi. That's another option. So that's similar to PowerPoint, but more of an immersive kind of experience. Yep. So that be another option.

Bryant Hoopes (56:06):

Yeah, I love the Microsoft office suite, you know, so early on in my career I did a lot of dashboarding and using Excel to manipulate data and do dynamic charts and just kind of having a little bit more of a knack for design and presentation.

Those skills that I learned, you know, 15, 20 years ago have really just given me that foundation to do what I can do now today, where I take those charts and put them into PowerPoint. And you know, I think what I have heard loud and clear from the chat is people are wanting examples. They want to see some of these things. And so you know, I'll commit to writing about this a little bit more and posting on the 33sticks.com blog about, you know, ways that I use little visual hacks and tips and tricks throughout presentations to make them more effective. And then there's books out there. In fact, I'll give a head nod here to Brent that you know, he was actually my supervisor at one point in my career, if you can believe that.

And one of the things, as I first started working with Brent, that he had me do was read a book called Don't Make Me Think, which is an old school web design book, which has actually been something that I'll pull out from time to time. And what's applicable for any type of presentation is that you don't want the people to think that you don't want them to have to interpret on their own. And while that book is maybe dated for web design, the concept is still very much real in that. You know, you want to create your story in a way that they're not having to think and you're guiding them.

Peter Nettesheim (58:02):

Awesome. Yeah, we're running out of time, but I know in a one hour session, it's a little tough to get examples. So I would refer anyone who's watching this to go to these experts, go to their sites, go look at the things that they're, they're blogging about, that they're putting out as far as content. And we can find examples there. Shoot, maybe we have another followup where all we do is examples, but I love that you're committing to provide some of those examples, cause that's always really helpful to people who are just starting out saying, how do I even begin? These examples are a good thing. So, well, we're at the top of the hour. And so our hour is done. 

I just want to thank our panelists here. You guys have been great. I love your experience. I love what you do. I love that you're getting the storytelling concept out there because it really helps get the action out of the information, the data that's created. So thank you for being here and thank you for your comments. It's been great. This has been awesome. So until next time, thank you. But make sure that you check out our next event— I think we're going to do another virtual event in July. But again, thanks for attending and thanks to the panelists for being here.

Previous Video
A Framework for Actionable Insights
A Framework for Actionable Insights

Join experts from Adobe and ObservePoint to learn how to achieve actionable insights from your data.

Next Video
How to Improve Data Quality for Better Customer Experiences
How to Improve Data Quality for Better Customer Experiences

Join Gabriele Sibani and Mike Fong to learn how to use data governance to improve data quality for better c...