The Story Rules Podcast E11: Brent Dykes – Author of seminal book on Data Storytelling (Transcript)

This transcript has been created using a combination of AI transcription tools and (some painstaking) human effort. Please excuse any typos, grammatical mistakes, inaccurate time stamps, or other errors. Specifically, the time stamps would not account for the intro portion of the podcast.

You may share portions of the transcript with due credit. Enjoy!

And, if you find this content useful, it would be great if you could leave a review on your preferred podcast platform.

---

Intro hook:

“I’m in the boardroom, he’s in the middle; (There are) Managers around, listening to my presentation; I get to the slide with this insight or observation that I had found, and he tilts his head, looks at it, and then blurts out “Bullshit!” At that moment I was like, “Oh crap, I have just stepped in it. I am not going to be one of the MBA students hired now. I am in trouble here,” and I was a little bit flustered. I wasn’t expecting that reaction; I was expecting “Oh, that’s interesting, Brent! Tell us more.” or maybe, “We should look into that.” But nope, it was basically flatly denied, and luckily, I had a mentor who jumped in and gave me some cover fire. I escaped the room that day relatively unscathed, maybe my ego (got) bruised a little…but what did die that day was that observation; that potential insight went nowhere. Nobody picked it up and said “Hey, let’s look at this trend!” or anything. Basically, when the Senior VP says “bullshit!” about something, I don’t think there’s another Manager who would pick it up. So, that taught me a very good lesson that I have to find a better way to approach communicating data.”

Welcome to the Story Rules podcast with me, Ravishankar Iyer, where we learn from some of the best storytellers in the world, find their story and unearth the secrets of their craft.

Today we speak with Brent Dykes, author of “Effective Data Storytelling: How to Drive Change with Data, Narrative and Visuals” a seminal book on the topic.

Brent started off his career in Marketing Analytics and is now a speaker and Data Storytelling coach.

In 2016, Brent wrote an article on Data Storytelling in Forbes magazine. That article went viral with 3-400K views. In subsequent conferences and talks, Brent was asked for book recommendations. He realised that there wasn’t anything which covered the stuff that he was talking about.. and decided to write one himself.

I’m so glad he did, because I learnt a lot from his book. In a blog post a few months back, I’d written an effusive review of the book – no wonder, I was excited to have Brent in my podcast.

In this conversation, Brent shares some of his best insights from the book:

• Why do you need to tell a data story: to ensure that your hard-won insights are not ignored
• When do you need to tell a data story: The simple but effective concept of the Story Zone
• How do you tell a data story: the different types of narratives and the one that Brent prefers
• The idea of a Data Trailer as a way of enticing the audience to listen to the entire Data Story

In addition Brent also narrates some fabulous contrasting examples – of people who struggled to make an impact because of poor storytelling, as well as of folks who changed the world through the power of story.

Let’s dive in.

Ravi 1:34
Hi, Brent. Welcome to the Story Rules podcast!

Brent 1:37
Hey, thanks for having me. 

Ravi 1:40
I’ve been working in this area for about five and a half years now, so I have a fair sense of the skill of data storytelling. And I’ve read a fair number of books on the topic, so when I picked up your book, I was like “I think I should know most of what’s written here, and maybe it will top off some concepts.” But I was blown away, Brent! You’ve not written a book on data storytelling, you’ve written THE book on data storytelling. It is fabulous. It’s so fundamental and it goes down to some of the most core aspects of the craft. I think (for me and for) anybody who wants to improve their data storytelling skills, this is now going to be my number one go-to book. Maybe right after Made to Stick. Made to Stick is very high up there for me. But this is really good.

Brent 2:38
You probably saw some references of Made to Stick in that book too, right? I worship that book as well; I really liked that book too.

Ravi 2:48
Yeah!
I want to start off with this question, Brent: you started this journey off with a blog – PowerPoint Ninja — which was great, (it had) tips and tricks on how to become better at PowerPoint. When did you realize that “Hey, what I’m doing is not making slides, but telling stories”?

Brent 3:09
Well, I worked in consulting and analytics consulting. That was my day job, and then the PowerPoint blog was actually when I went to my company that I was working for at the time, and asked if I could start blogging for the corporate website. And they said no; at the time, it was just reserved to – (it) sounds silly now, but at the time, they reserved it to just the CEO and a few other executives. So, I thought, “Okay, well, what else could I do? What else am I passionate about?” and that’s when I started blogging on PowerPoint, because I’d always been skilled at PowerPoint. I always got lots of compliments, and I always seemed to be sharing advice and best practices with other people, so I thought “Well, there’s a topic I can start writing about.”
I’ve always enjoyed the presentation side of things. But then, as I started to look at my day job, I started to see that “Man! People are really struggling with how to communicate facts and data”. We would do analysis for different clients, or I would work with clients who were doing analyses, and sometimes (where) I saw it for myself that sometimes, I’d have a really good insight that I communicated, but it would fall flat; it wouldn’t go anywhere. I saw many other people struggle with how they communicated data and insights. So, that’s when I started to see that there’s a real opportunity here to communicate our insights much more effectively, and there’s an example of how I failed — I think the very first example that I (talk about) —

Ravi 4:47
I love that! Could you talk about that example? it was a great way to start your book 

Brent 4:52
Yeah! It goes back to when I was an MBA student, and I was working at a B2C company primarily that offered a retailer direct catalogue kind of business; they had a very renowned e-commerce team which I was a part of. And I was one of multiple MBA students that were all vying for a position at this company, and we had to present to the Senior VP of e-commerce, who wasn’t your typical executive. He came from a Special Forces background, he flew helicopters in the military, and had the typical military persona. He left the military, not because he wanted to, but I heard (it was) because his wife was sick of him being in the military. So, he switched to business and got his degree from a very prestigious MBA program, and then was leading this e-commerce team that was very well respected in the industry. But he had a reputation for being very stern, (and he was obviously a) very sharp guy, very intimidating; it wasn’t unheard of for MBA students to come out of presentations with him with tears in their eyes. Anyway, it’s very intimidating and here I was — actually, it wasn’t our final presentation, it was a midpoint presentation (which was) just to check on how we were doing with our projects, where we were, were we headed in the right direction? — And as I was preparing for this presentation, I stumbled across an insight or an observation in some customer feedback data that I had, and it showed that they didn’t feel as strongly about one of our core principles of our e-commerce…(actually) I think it was related to our shipping policy, I can’t remember exactly what it was, but let’s assume it was about our shipping policy. And when I made this observation, I was like, “Oh, wow, this could be meaningful to the business. If our customers don’t feel as strongly about our shipping policy as we feel it should be, then maybe we should be focusing in other areas.” But here’s the key thing: it wasn’t directly related to my project. So, I was debating whether I should include it or not, and I made a decision, “You know what, it’s something of value, I should share it.” So, I included it in my presentation. Fast forward (to) two weeks later — I’m in the boardroom, he’s in the middle; (There are) Managers around, listening to my presentation; I get to the slide with this insight or observation that I had found, and he tilts his head, looks at it, and then blurts out “Bullshit!” At that moment I was like, “Oh crap, I have just stepped in it. I am not going to be one of the MBA students hired now. I am in trouble here,” and I was a little bit flustered. I wasn’t expecting that reaction; I was expecting “Oh, that’s interesting, Brent! Tell us more.” or maybe, “We should look into that.” But nope, it was basically flatly denied, and luckily, I had a mentor who jumped in and gave me some cover fire. I escaped the room that day relatively unscathed, maybe my ego (got) bruised a little…but what did die that day was that observation; that potential insight went nowhere. Nobody picked it up and said “Hey, let’s look at this trend!” or anything. Basically, when the Senior VP says “bullshit!” about something, I don’t think there’s another Manager who would pick it up. So, that taught me a very good lesson that I have to find a better way to approach communicating data. Did I fully support that observation? Did I present it, or communicate it, in the right way? I would say, no, I let it down. And partly, that has led me to my journey, or to the discovery of data storytelling and how we can take the combination of data, narrative, and visuals, and bring them together to really tell powerful data stories. And I really believe that it makes a big difference. (With) me working in the data world, and I’m sure lots of your listeners also work with data a lot (too), we make these observations, or we find these insights, and it can take a lot of work to just get to that point of finding an insight. (We have to get) all of the data, we have to capture all of the processing (and) preparation of that data, then the analysis. And then, in the final last few yards to get to the finish line, we stumble and we don’t communicate it effectively. So, that’s what I’m trying to address with data storytelling; I’m trying to say, “Let’s try and preserve some of these insights because they have value, that is just not being communicated clearly. People just don’t understand them, they can’t fully appreciate why they are valuable.”

Ravi 9:47
This is a great example, and I’m sure a lot of folks go through this (situation) where they have the data, they have the insights, (but) they’ve not done that hard work of having a narrative. And I really liked the three pronged (approach), (or) the three pillars of The Data, The Narrative and The Visuals; often, we take the data part for granted. The other part that I liked in the book was that you do talk about situations where we have the data, but we don’t have a clear narrative; but often, what happens is that we do have a narrative, but we don’t have the underlying data. And that happens a lot too, especially (with) folks who have a clear point of view but they don’t really bother to (back it up with data). Can you share some examples of that and how is it that you can probably avoid going down that path?

Brent 10:37
Yeah, that is a danger, because if we have a story in mind, or if we have an agenda, or we have a position that we want to take on something, we want to show that our…say we’re a marketing manager, and we want to show that our campaign was successful. So, the danger there when we try to tell a data story from that approach is, what are we going to do? We’re going to cherry pick the data points that support why our campaign was successful, and either inadvertently or consciously, we (might) avoid data points that show that maybe some of the aspects of the campaign weren’t that successful, or (that show that) maybe we didn’t achieve all of the goals and objectives, and any data that doesn’t support the point we’re trying to make gets ignored. I think in those cases, we may produce something that looks like a data story, but in the way that I define as true data storytelling, it’s not going to really be a true data story. Because, I think at the end of the day, obviously, it’s hard to remove the bias that we have. We have these biases, we have a hunch, we have a hypothesis, but the key difference is that as we go into the data, if we take it as a hypothesis, and then the data shows us that no, actually, our hypothesis is not correct or isn’t valid, then we’re going to be open minded; we’re going to listen to the data; the data is going to teach us something, it’s going to show something, we’re going to learn something, and then maybe the story that we wanted to create doesn’t get created; we create a different story, the data shows us something else and so we have to go into the exploratory side of finding insights and exploring the data with an open mind, and seeing what it reveals and whether it can help us to prove the business (…). Not coming in with an agenda – that’s the crucial thing.

Ravi 12:39
And this is so common, right? So, Brent, do you have a hack for this? (Or maybe) is there a process that we can use that ensures we have that little bit of a devil’s advocate position for our own position as well?

Brent 12:55
Yeah, I mean it’s the scientific method, right? It basically comes down to that. We have a hunch, we have a hypothesis, we test the data to see if our opinion or our conclusion is correct. And the danger is that we can (fall prey to) confirmation bias, (and unknowingly convince ourselves) that we were right and that our campaign was successful. And so, I think (by) going in with it to test it, you’re trying to disprove the hypothesis. So, you’re trying to go in and (see if) there is any data that can invalidate the hypothesis. Once you’ve done that and you’ve said “No, everything looks great, it’s looking very positive,” then you can move forward with confidence and tell that data story, and know that you’ve exhausted the possibility that you were wrong. 

Ravi 13:54
I think that’s a great point – actively try and disprove your own hypothesis, that’s a good way to do it.

Ravi

One thought that keeps coming to me, (since) we’re talking about the importance of narrative in data storytelling, and, I think, both of us do believe that it’s probably the most important part of your data story. But so far, if you look at this still fledgling field of data storytelling, why do so many people conflate data storytelling with data visualization? A lot of the books that have been here so far are all to do with charting and good charts, and the right visuals. Why do you think that is happening?

Brent 14:35
I think the danger (is that) having worked in tech for a long time, we jump on to buzzwords, (or) jargon terms, and I think that a lot of vendors have jumped on to data storytelling, as a way of (saying) “Hey, you can build a dashboard and you can tell an amazing story to your users.” So, I think a lot of that’s been reinforced, both by the media and by vendors, and perhaps people have felt more comfortable with the visual side of storytelling. And that’s something that, as I was looking to write my book, I was fearful (of). I mean, I could see the power of data storytelling, but I was very concerned that jargon and buzzwords were going to take it over, and then data storytelling, like so many other buzzwords, would be relegated to the scrap heap of terms that are now dismissed or have had their fashionable flare up, and then they disappear. Like ‘big data’, right? I think, outside of the analytic space people might still use the term ‘big data’, but if you’re working in analytics today, (you’ll notice that) nobody says ‘big data’, it’s just ‘data’. It’s gone away and nobody talks about ‘big data’ anymore. That was my fear about data storytelling. That so many people will just use it as a synonym for data visualization. And that’s why I felt like I needed to write my book to kind of clarify that “No, visualization is a means to an end, it helps us to take very complex information or data and share it in a way that other people can understand it and interpret it. But at the end of the day, if we’re looking at what’s more important, the visualizations or the story? the narrative is the most important part. The visuals are there to support the narrative, not the other way around.”

Ravi 16:45
Absolutely. I never thought of it from this point of view that there are folks who have visualization products, but there is no ‘narrative product’ as such, so there’s no one to kind of push the case of narrative being more important. That’s a very interesting take. Coming to narratives itself, Brent. It gave me a bit of clarity, looking at the different narrative frameworks that you have talked about in the book. You have divided it very broadly into these two camps: there’s the journalist, or the writing narrative; and then there is the more the traditional narrative framework, which has been used in mythology, books, and movies, etc. So, here’s one point where I had a question or a slightly different point of view – I’d love to get your point of view – I believe both have their way, I agree to what you say in the book, that the journalistic method which is the inverted pyramid where you start with the big news or give up the big reveal first and then get into more details later is something that is not fun because you’re giving up the main reveal at the beginning. But, if you actually build up your story and give a hook (at the start), and then give some details and talk about a problem and then slowly come up to the big reveal towards the end, then throughout your story, you’ve got your audience’s attention, which is a great way to do it. What I would like to ask is, are there situations where you feel that the inverted pyramid method, or if you go to some of these consulting firms, (like how) McKinsey has the pyramid principle, do those kinds of frameworks work in some situations?

Brent 18:40
Yeah, I wouldn’t say they don’t work in certain segments. I work in business and I see those used all the time and they can be very effective. The key thing there is that they’re based on more of a logical appeal than an emotional appeal. And when I looked at it, I saw a lot of people saying, “Oh, story, story, story” and I was like wait a second, they’re kind of talking about it more from a news story perspective. And even within news, if you look at it, there’s narrative journalism, and then there’s the regular journalism, so there’s ‘60 Minutes’ (inaudible). Like the investigative kind of (reports), you read one of those and it’s more like a story, it isn’t written with the inverted pyramid.
Now, the key thing that I found is – and I was coming to my conclusions around this – is that, probably in business today, (there is) a lot of this survival mechanism, because we have so much information. There’s probably a lot of bad information sharing, where it’s overloading people. And so, I think it’s almost like a coping mechanism for a lot of executives and leaders to just say, “Give me the facts!”, because they don’t have time to kind of walk through the analystsanalysis, (or go down) the road of discovery that they go through to find everything in on all the minutiae now; “At the end of the day, I’ve got 15 minutes, or 10 minutes, tell me what I need to know. What’s the decision you want me to make? or what’s the most important?” and so it’s kind of a survival mechanism. But what happens when we take that executive summary approach, whether it’s the McKinsey principle or the inverted pyramid, (is that) we’re basically killing the story. I talked about it being an ‘anti-story’ and I wasn’t the first person that coined that, other people have called the inverted pyramid the ‘anti-story’, because it has no arc. The arc is gone. There’s no build up, there’s no hook, there’s no climax, there’s no resolution or any of that stuff. So basically, we don’t get any of the benefits of emotional connection that we (would otherwise have) gotten; human beings are storytelling creatures, we try to take facts and data and put it into a narrative, we try to create a narrative around the information. And so, with storytelling what we’re doing is, we’re guiding them through that process of understanding the information and putting it into some meaning for them. Whereas with the executive summary, (it’s as though we’re saying) “Here’s the main data point, here’s the main takeaway, form the narrative for yourself.” And that’s where you can get executives who may take very different approaches, because they may not have the full context; they may not have all of the connecting details that make sense, and can then inform that the decision that they need to make. And so, I kind of see it as there are situations where we will use executive summaries. I am not saying that everything needs to be a data story.

Ravi 21:51
What could be a situation (where one will use executive summaries)?

Brent 21:54
So in my book, I call it the Story Zone, and basically there’s a typical consulting, four-by-four matrix, more or less. I mean I kind of skew it a little bit but, but so, imagine, this is so everybody on the call on the podcast can interpret it. So basically, along one axis you have the value of the insight, is it a high or a low value insight? That’s one determination. And then the other scale that I have is whether it’s a hard insight or an easy insight to process or understand. So, what do I mean by a hard insight? Well, if I’m coming with an insight that says that the program that we just launched six months ago has been a failure, and we need to stop it, and (suppose) you were the creator of that or (you were) the team that I’m presenting this to that have invested hours and money and all kinds of effort to do, that’s going to be a hard insight for you to swallow. Or, I come back with something that’s counterintuitive to the team. I could come back with an insight that’s going to be very costly. And so, in these situations, the status quo is always going to feel much easier to go with. It’s when we have to drive some kind of change and we know that there’s going to be either some emotional or other kind of resistance to that, in those situations where we have a hard insight to process or understand, and we also have high value – and I would say even medium value, that we would want to build a data story in those cases.
Now, let me give you an example of where you wouldn’t need to invest a lot of time in building a data story: what if I come back and said, “Hey, that campaign that you guys just rolled out or that new product launch that you guys just did, it was an amazing success! Let me walk you through how it was a success.” And everybody’s going “Yeah, I knew it was going to be successful, you know, oh, pat on the back…” Do I need to do that, the full data story? No, I don’t. I can just give them the facts and there’s no objections, it’s all just feeling good. Or maybe it’s low value, do I want to invest in an insight that’s hard, but at end of the day the payoff is very small? In that case it’s like, “meh.” Sometimes when I do workshops and when I’m training people, I’ve had some data analysts come back to me and say, “Man, this data storytelling seems like a lot of work”, and it is! I will say it does require an investment of time, it’s not as easy as just building another dashboard, or just pulling together a report, and then just throwing it over the wall, and whether that report or that dashboard survives or not, or how it’s received, doesn’t matter. In this case, if you have an insight that is meaningful, that is maybe harder to process or to accept but we know it’s valuable, then we need to invest that extra time. And that’s what I try to convey to people that, “Don’t worry, you don’t have to create a data story for everything. But in those moments where it is required, and you’re going to run into some resistance, you’re going to run into some questioning or some doubts, that’s when you need a data story to help push people over the crest of the hill to actually take action, and to take the right action.”

Ravi 25:46
This concept, Brent, of the data Story Zone — and I’ll share the visuals in the podcast show notes – really clarified a lot for me and I was really curious to know, and quite amazed to think of, what your thinking process would have been (when you came up with it). How did that idea of “Hey, there is something which is a Story Zone.” come to you? Was it when (you were) reading something, or when you were out on a jog somewhere…How did it come to you?

Brent 26:19
It was a while ago, so I’ll try and reflect back (on it). But it was probably just because people would push back on me a little bit, like “Do we always need to tell a data story?” I think it was just that kind of push back and I would say “You know what? No, you don’t. You don’t always have to tell a data story”, and that might come as a shock to some people who think like “What? Aren’t you going to advocate for data storytelling?” and I’m like, “Yeah, but I’m pragmatic. I realize that it’s going to take effort, it’s going to take a lot of work to sometimes create a really powerful data story. And is that appropriate in every situation? It’s a tool in your toolbox. It’s a powerful tool and it does require some effort. And so, use it for those moments when you need to use it. And then you can do other approaches when it’s not needed,” and that, for me, is the pragmatic or practical approach to how we communicate data. And it is very important because, there’s lots of experts out there that will say you have to do this or that, but what I’m trying to say is: let’s be pragmatic. I was in your shoes, the listener’s shoes; I know how it is, you’re busy, there’s lots of different demands pulling in different directions, you’re not going to have time to always create these beautiful visualizations and craft these amazing stories. So really, it’s about those moments when you need to do it, and I’ll give you an example…I was actually meeting some customers when I was working at another company, and we were talking, we had kind of a lunch at a round table. And one of the individuals that was at my table, he was Senior Analyst and he said, “You know, I never really put much credence into data storytelling, until I saw how much of a difference it made.” And he said that there was a scenario where there’s two analytics managers, (who were) probably (at a) similar level in the company, (with) similar backgrounds, (both of whom were) obviously skilled with data and would have no problem finding insights. And they had two different proposals: one guy’s proposal would be a multi-million dollar ask to pursue; The other guy had six figures, it was a smaller ask. One of them told the data story – it was actually the guy who was asking for the millions of dollars to fund his project; and the other guy who was just looking for six figures, did not use storytelling. And then the guy said, “I couldn’t believe it, the guy who was asking for more money for his project (got approved). Simply because he was able to tell a data story and form that narrative, and use visuals effectively.” His project was approved, and the other guy, he said that to this day he is still struggling to get funding for his project, because he hasn’t figured out how to tell the story of why the company should invest in this project. So, for him that was a revelation, that “I need to do this, in those moments when I need to employ these tactics because they work.”

Ravi 29:46
That might be a good time to actually talk about the narrative arc that you talk about in the book. What are the components of it, and how does the story flow in this framework?

Brent 29:58
Yeah, so, when I was researching for the book, or even as I started to delve into data storytelling, I think the most common kind of storytelling model out there comes from Aristotle, which is the three-act play. A lot of people summarize it as, ‘your story has to have a beginning, middle, and end’. And I will be honest, I hate that! It’s not helpful. I’m sitting in presentations and listening to other people talking about data storytelling, and they check the box by just saying, ‘you just have to have a beginning, middle and end.’ And I’m like, “What the heck does that mean?” It’s completely useless because I can look at a report, and it has an opening paragraph, it has some paragraphs in the middle and has a concluding paragraph. It’s not a story, but it matches that! So anyway, to me, this wasn’t this helpful. And then on the other side of the spectrum, I ran into Joseph Campbell’s ‘The Hero’s Journey’, and that’s a very fascinating model. If you’re not familiar with it, basically, George Lucas, used that model for the first Star Wars movie. So, Luke Skywalker went through all the steps of the hero’s journey. There are around 17 steps, I think, in the original version, some other people have simplified it to 12; and it’s all very focused on this journey of the main hero or character. Now, I’ve seen other people talk about that as like a really effective model and I’ve even had people come up to me and say, “Have you heard of the Hero’s Journey, Brent? because you’re really into data storytelling.” And my reaction is the opposite. It is too complex, it’s great for writing fictional books, it would totally be a template that you could (use to) build your characters and think about “Okay, this is what I need to do with this character,” and it would work really well. But, in a business context, for me, I just didn’t find it that practical or useful. So, where I landed was with the Gustav Freytag model, the Freytag’s Pyramid. And I modified it to data storytelling. Just taking his model, and thinking “Okay, in terms of data, what are we doing here? How do we connect these?” If you’d like, I can kind of give an overview of how (I went about this).

Ravi 32:37
Yeah, I’d love to have you talk about the modified model that you use.

Brent 32:41
Yeah, if we go back to the original Freytag model, it basically starts with the Exposition. With an introduction to the characters, setting the scene; And then there’s, I don’t think it was in Gustav Freytag’s model, but it’s a modification on his model, but we talked about having an Inciting Incident in the story; and I use Harry Potter a lot when I’m explaining this Freytag’s Pyramid, because we have this poor, orphan boy living a miserable life under the stairs, with his adoptive relatives who hate him, and mistreat them, and then one day they go to the zoo and he speaks to a snake. And that’s the inciting incident, that’s when he realizes, “Oh my gosh, there’s something going on. I’m not just a normal boy.” And then we go through the next phase which is the Rising Action, where Harry Potter gets rescued and brought to Hogwarts, and then he gets assigned to a house, he does Quidditch, and then he battles Voldemort. And then, that’s the climax. And then after that, you have Falling Action. So, after that main battle scene, where we’ve learned that Dumbledore had defeated Voldemort…Oh, sorry! (These are) spoilers if you haven’t read it. I figured if you haven’t read Harry Potter by now, you’re probably never going to get to it. So, everything’s kind of resolved and then you have the resolution at the end.
Now, how do we approach that with data storytelling? That’s pretty similar, we start with what I call the ‘setting and hook’. And so, we’re introducing people to the data set that we’re looking at, what’s the status quo? How do the trends typically go? What are the expected results that we typically see? And then we have what I call ‘the hook’, which I also call a “hmm” observation. It’s an observation that makes us go “Hmm, that’s interesting. Something’s going on with the data.” Now we notice a spike in a particular metric that’s unexpected, or a drop in a metric, or something changes in the data that would catch our attention, and would catch the audience’s attention. That’s the hook that begins the story, so we provide just enough context to understand the hook, and then we go into what I call, instead of rising action, I call it ‘Rising Insight’. So, it’s like an onion and we start to peel that onion and dig into ‘Why did this happen?’, ‘Why did this metric spike?’ or ‘Why did this metric go down? What’s contributing to it?’ and we start to reveal to the audience what’s going on. And then, we build up to our ‘Aha!’ moment which is our big reveal, our big takeaway. And at the end of the day, if they don’t remember anything else, as long as they remember that ‘Aha!’ moment, then we’ve succeeded.
Now, the key thing is, how many rising insights you have can depend on the story, or the complexity of what you’re analysing; It can be very short, it can be very long. Obviously, as long as there’s progression towards the ‘Aha!’ moment, that’s the main thing. And then, after we reveal the insight, we’re not done. How do we drive action? That’s a big focus of my book; how do we drive change? How do we drive action? How do we influence decisions? Well, we also have to help with forming a decision around what we do with that insight. So, we’ve identified a problem, or we’ve identified an opportunity. What are our options? How do we tackle that? And when I talk to analysts, I say “Your job is not done. You found the insight, but then your decision makers need help with making a decision. There’s more analysis that goes into the options, so you need to analyse option A, option B, C, and then make a recommendation.” In some cases, it’s appropriate to make a recommendation to the stakeholders, because they trust you, they believe that you’ve looked at the numbers, and they place value on your recommendation. It doesn’t mean that they’ll follow it, but I think it’s helpful to provide a recommendation.
And then we talk about the next step. So, we provide (them with) “Here’s our recommendation, or here are the options, and then here are the next steps that we need to take from here.” And so, through this process we’re doing a couple of things: One is we’re helping the audience learn more about whatever we analyse, whether it’s customer data (or something else), or we’re helping to better understand our customers; we’re helping them to understand a process; we’re helping them to understand the market; whatever it is that we’re (trying to do), they’re growing, they’re learning. The second thing is hopefully we’re putting them in a better position to take action on the insights and make a decision and drive change. And so, that’s how I look at it as a model for how we can tell data stories. Now, the interesting thing is, as I’ve shared before, I’ve had pushback saying “Brent, I love the data storytelling arc that you just shared, but,” it goes back to our point before, where we have executives who say “just tell me the numbers”. And they’re programmed to operate that way. In some cases, there’s limited time, so they can’t spend a lot of time with each team that’s presenting them with information. So, the thing I came up with is a slight modification, I call it the ‘Data Trailer.’ The data trailer is a modification of the model to accommodate how a lot of executives look at their data today. Just to recap, you have your setting-in hook, you have your Rising Insights, you have your ‘Aha!’ moment, and then you have your solution in the next steps. So, what we do for the data trailer is, we take the hook with just a little bit of context – it’s hard to present a hook without some context – and then we take the ‘Aha!’ moment. We don’t give them all the Rising Insights; we don’t give them the solution. We don’t talk about that, we just say, “Something’s going on here, and if we don’t fix it or if we stall, we’re going to miss this opportunity.” and then we present it as a data trailer. It’s an abbreviated (version), and I would say it’s the worst movie trailer out there because you’re actually giving away the climax of the movie. But what it does, is if you get an executive who says, “Well, tell me more,” because (they’ll be curious about) how you are connecting it. “Why are you saying we’re going to lose $2 million in the next quarter?” Then now, you have permission to tell them the rest of the story. That’s my workaround for those scenarios where people may be sceptical. I don’t know if a data story will always work for my audience, if they’ll be receptive to it; In those cases, I’d say “Well, try a data trailer first”, and that could mean that maybe they say, “You know what? that’s interesting. Thank you for that information but it doesn’t go anywhere. And then it’s like, “okay, okay.”

Ravi 40:23
I love these approaches, Brent, because one, you’re saying is a data trailer – (for) people (who) talk about time being a constraint; The second, the counter you have for people who talk about time being a constraint is to say, “Hey, you don’t have to do it for every data presentation, look for those in the Story Zone.” And when something is in the Story Zone, I go back to your original story, where you had an insight which was powerful, which was hopefully valuable, and it was hard. And if you have an insight which is valuable and hard, and if you don’t put in the effort to make it into a data story, then it can suffer the fate that that insight did, which is (it gets) consigned to the heap and then nothing happens. And if you don’t want that to happen, then you get the time from the Executive, like “Sorry, but this deserves half an hour of your time.” and then you drive the trailer, and I think that a great way to (go about it).

Brent 41:20
And in part, maybe a subtle detail on that is if you’re the Manager or the Analyst, or whoever’s presenting this, and you feel passionate about something, “This is valuable. This is important,” I think that over time as you pick and choose when to tell data stories, you’re going to be building your credibility, the ethos; You’re going to have that as a storyteller. There are going to be moments where people wouldn’t typically give you the time of the day, but because you’ve built up credibility over time, and you’re saying, “You know what? You need to hear the story, you need to hear this data story because, I believe it’s important. Trust me”, and then the executive says, “Well, you haven’t led me astray before, why would you now? So, I need to hear this story because you believe in it”, so that’s kind of like being unsaid in that but I think that that could also be a big factor.

Ravi 42:21
Great points, Brent. The other point that I really loved about the book is the collection of real-life stories that you use to talk about the power of data storytelling, and I love the detail which you have gone into in describing these stories so maybe you can talk about the amazing, fascinating contrasting stories of Dr Semmelweis versus Florence Nightingale and Dr John Snow.

Brent 42:49
Right. Yeah, I guess I’m kind of a history buff as well. I was listening to a podcast and they were talking about Semmelweis, and I was like, “Oh my gosh, this guy had a problem with data storytelling!” So, for those of you who aren’t familiar with Ignaz Semmelweis, I’ll give you the quick version: He was a Hungarian doctor in the 1800s, and he was working in Vienna at a maternity hospital, and he was assigned as the new administrator. They had a problem at this hospital at the time, again, when he was there it was the 1840s. They didn’t know about germ theory, so there was no Louis Pasteur, germ theory hadn’t been discovered yet, and so, they really didn’t know why a lot of women at this maternity hospital were dying of something called ‘child bed fever’, and they had two clinics – they had a clinic that would train student midwives, and they had another clinic that would train student doctors. Now the shocking thing to the university was, the doctor clinic actually had a mortality rate more than double that of the midwives, and they couldn’t understand at all why the doctors’ clinic was killing more of these women. And so, they tried everything; they looked at (the temperature and thought) maybe the temperatures in the rooms are different. They had this miasma theory which is, back in that time they thought the foul smells would be an indicator of disease; kind of a typical correlation not causation mistake. But anyways, they looked at the birthing techniques that the midwives used, they looked at overcrowding in the rooms and every time they checked all of these different hypotheses, they couldn’t prove anything. Until one day, when Ignaz was, I think, he was out of town…I don’t know if he was on vacation or if he was out of town; but a fellow doctor was performing an autopsy on one of these women that had died. And he was doing that with other student doctors, when one of the student doctors accidentally cut his hand while he was in the middle of this autopsy. And sure enough, that wound got infected, and a few days later that doctor died, and Ignaz heard the news and rushed back to the hospital. He had the tough job of performing an autopsy on this doctor that had died, but that’s when the ‘Aha!’ moment for him came, because as he was preparing the autopsy, he found that the pathology of how this doctor died was very similar to how these women were dying. And then, he started to connect the dots and he was like, “Wait a second, maybe…”, and they had a standard practice at this hospital to perform autopsies on the dead bodies, most of them (were on) these women that died of child bed fever – in the morning, and then the doctors would do their rounds, examinations and deliveries the rest of the day. But he was like, “Wait a second, maybe there’s particles on our hands that are being transferred to these women and making them sick and killing them.” And so, he immediately introduced a hand washing policy after they performed these autopsies, where they wash their hands in a chlorine lime solution. And obviously, the chlorine would actually kill the germs. So immediately they saw a dramatic drop in the number of mortalities they were having. And over the course of the next 18 months, they were able to maintain relatively low levels of mortality rates for the clinic which was unheard of, up until that point. For five years, they went through a very high mortality rate. And in a couple of months, they actually had zero deaths in this clinic. But, at the end of the 18 months, Ignaz was let go. What happened was, that his superiors could not embrace the idea that the doctors’ hands were killing these women. They said it might be one factor, but it’s not the sole factor, and they were looking at 15 other different causes. So, nobody would agree with Ignaz and he was unfortunately mistreated. He was treated very poorly, he was ostracized from the medical community, he couldn’t get another job in Vienna, and he had to return to Budapest. And then for the next 10 years, he waited for his ideas to be embraced and accepted, but they never were; and he published a scathing journal article, attacking the leading obstetricians, and calling them ignoramuses and murderers, and he went unhinged. But when I look at what he did and what he didn’t do, is he didn’t tell a story with his data. And you might say, “Well, that’s not fair. He didn’t have Excel, he didn’t have PowerPoint, he didn’t have Tableau, how can he tell a data story? How can he visualize it?” Well, here’s the thing. At the same time period, there are two excellent examples, even in the same medical field!
We have Florence Nightingale, who is well known for introducing modern nursing, but she was also a statistician. When she got back from the Crimean War, she had been enlisted into this movement to help the British soldiers, because during peacetime, they actually had a higher mortality rate than the general public, simply because their barracks and their hospitals, the military hospitals, were unsanitary. And so, a lot of soldiers were getting sick and dying in peacetime just because of unsanitary conditions. And so, she got behind this; she did a lot of analysis, she created her own visualizations to something that’s sometimes called the Nightingale Rose Chart or the Coxcomb, (which is) basically a polar area chart, it’s hard to kind of explain, but yeah, she created it.

Ravi 49:17
And, that’s crazy, right? For a nurse to say “Hey, we need to visualize this stuff and tell a story”, that’s crazy that she thought of (doing that).

Brent 49:27
Yeah! And she even noted that she felt like words wouldn’t get through to people, it was only when people could see the data that they could make a decision. So, her impact in that case was that she was able to influence changes in the British military, where they improved the conditions of their barracks, improved the conditions of the hospitals, and she even did analysis and said, “You know, approximately, I think we’ve saved 1000+ lives by (implementing) these changes”, so, she had success where Ignaz didn’t.
The other one was Dr. John Snow – not from Game of Thrones; he was a very famous doctor actually, he basically anaesthetized Queen Elizabeth at the time for two of her pregnancies. Anyways, he was a big proponent of cholera being a waterborne disease. And at the time, again it goes back to the miasma theory of how in London, they had very strong smells because all the sewage drained into the streets and into the Thames, which was their source of drinking water. And you have over 2 million people crammed into very confined space, so it stunk. And again, a lot of the disease and everything was correlation not causation; but he believed that the cholera outbreaks they would occasionally get in London were based on waterborne transmission, which wasn’t the accepted hypothesis at the time. So, there was an outbreak in the Soho neighbourhood in 1844 or 45, I forget. But he basically rushed there to investigate what was going on, and then he started to isolate the problem around one water pump. And what he did is, he created a map of the neighbourhood. He showed a dotted line, where that pump was the walking distance for most of the locals, and then he put a black bar where people were dying of cholera, and it was highly concentrated around this Broad Street pump, and they later discovered that was there was a lady whose child was suffering from cholera, and she had been cleaning the diapers, and it was seeping into the water supply of this pump. The baby died and then, sure enough, the cholera spread to other people. But basically, what Dr. John Snow did, is he created a map that visualized this, shared it with different council members in the local community, and they removed the handle of this water pump so that nobody could continue to draw water from it and get sick. So, he was able to effect change. Now, unfortunately he died of a stroke before he was able to see the change that he had caused. But in 1866, when they had another cholera outbreak, the government said to boil the water. And I believe a lot of credit for that goes back to Dr. Snow saying that water was the cause, even though they didn’t know germ theory at this time. So, if we compare Nightingale and Dr Snow to Semmelweis, (in the case of) Semmelweis, it wasn’t that he didn’t have data. He had 60+ tables of data, showing different interesting things, but he never visualized it. He never told the story, and I think that’s where he failed and that’s what we have to learn.

Ravi 53:17
I love how in the book, Brent, you’ve actually shown one of his tables; and one just cannot make anything out from the table, even though it’s there, the data is right in front of you, but on the next page (is) a simple line graph (that you have included) and you can actually see (the visualized data) and, oh my God! It is fascinating.

Brent 53:35
Whenever you can take a historical example like that one, just (representing it is helpful). One of the challenges when you write a book on data storytelling, is that there are many really great data stories out there, told at different companies, but the challenge is that nobody wants to reveal that they had a problem with their customers. And there’s these amazing data stories but it’s very hard to get real world examples, so sometimes, you have to go a little bit into history and, obviously, all of these characters are now dead, and hopefully they won’t mind us learning lessons from their mistakes or what they did right.

Ravi 54:16
These also kind of show how important this skill can be, right? It’s a life and death issue sometimes. Although they do come more under the visualization part of the data storytelling skills, but I think that the narrative part and everything else also comes into play.

The other thing that I really liked about what you have done in the book is the strength of your research sources. You mentioned in the book that if you have a clear narrative, it should be backed by strong data, and your claims and your points are backed by really solid research papers; you’ve got a Kahnemann there, and you’ve got really strong sources. So that’s one very cool thing and I want to talk about your process of finding those. In addition to that, the other element that I really loved reading in the book, is a very eclectic, or very thoughtful collection of quotes on stories and on data and data storytelling, by a very wide variety of people. So, I’d like to get into a little bit of the process of writing this book and I know that you took a little bit more time on this book, so how would you actually (find these quotes?) Would you always have a lens over your eyes that (would suggest) “That quote is interesting,” or “Oh, that source is interesting” and then you would put them into some sort of a folder in your laptop somewhere, or what were those tactical processes that you’d use?

Brent 55:43
Yeah, with the quotes, I think, sometimes you’re just out in the wild and you read a quote and you’re just like, “Oh my gosh, that’s perfect. I love the angle of that.” And I would say sometimes, when I started going down the path for the readers or for the audience, I have a lot of quotes in each section, I basically have quotes that introduce each one. And, it was a great idea when I started but then sometimes (I’d realize) “I don’t have a quote for this one”, and so then I would have to go out and find a quote (to use). I had a Word document that I kept and whenever I find good quotes, I would keep them (in that) and I continue to do that today. I would collect everything that I had, so I definitely had a great library of quotes. But then there were some sections where (I’d ask myself) “Do I have a quote for this? No, I don’t, I need to go find a quote.” But that was fine, that was the minority, probably 80% of the time I had quotes for every section. Once, I even used the same quote twice in the book without realizing, and then right before we were about to publish, I changed it because I caught that I’ve reused the same quote in two places. I must have really liked that one.

Ravi 56:55
It must have resonated with you. What about the research papers, Brent? Was it that you did a lot of additional reading for the book, or this was kind of stuff that you would have been reading anyways?

Brent 57:05
It was both; Obviously, I have a focus on data storytelling so I’m going to always be like a squirrel, turning my ears to where (I hear anything related to it, like) “Oh wait, they said something about data storytelling” or something related to human perception, or other (avenues). There’s a bunch of supporting areas that relate to data storytelling. It was an interesting process, because I had some research that, as I was about to put it into the book, I would check on it, just to make sure that it’s valid. And, in a couple of situations, I had well renowned authors who use this data research in their books, and the easy way to go about it with this is to say “Oh, so-and-so said this”, and (then) I run with it; but I found that in a number of cases, the observations that they had were not backed up by data. I was going to include them in the book and it didn’t make it into the book, but on my blog – Effective Data Storytelling – I have an article there, where I have a number of misconceptions that I found related to data visualization, and also about how we retain information, (etc.) But there’s no supporting research to really validate those (findings). I thought it was very interesting when I was writing, because so many people are quoting this as a given, but it was not proven to be true. They go into stuff like saying that we process visuals 60,000 times faster than text, I think. A lot of people attribute that to 3M – the corporation that had it in a brochure in 1997, and somebody even offered a reward saying that if you can actually find the source of this research, I’ll pay you some money. And nobody’s claimed it; The furthest back that they could go was to 1982, when it was in an ad. So, it’s a very unreliable source.
There are also other ones that the other myths that I talked about, that 90% of information that is transmitted to the brain is visual. That’s just a quote from one author who didn’t provide any link to any research to support it. Maybe it is real, but I can’t use that in my book. Another one that’s very common is that 65% of people are visual learners. That one is, again, not validated. There’s the one that says information retention differs by delivery approach, that we only retain 10% of oral, 35% of visual, and if we do visual and oral, then that’s 65%. And that was quoted in a very famous book, but when you go back to look at the original research, (you see that) it’s garbage. It’s not substantiated. There are other ones that said people remember 10% of what they read, 20% of what they see, 30% of what they hear and it’s like this hierarchy, that is not proven to be real based on any research. And then the last one I talked about is presentations that have visual aids are 43% more persuasive. That goes back to an actual working paper by a number of professors, but when you actually look at the report, it contradicts itself; it’s flawed. So, it’s interesting, and that was problematic for me because as I’m writing my book, I’m counting on certain statistics that I want to tell in my book. And I’ve seen (them be) touted by other experts in data visualization and data storytelling, and I’m like, “No, I can’t use it in my book because it’s not backed up by real research and data”, so that was the thing that delayed maybe some of the time it took to produce the book, because I had to verify my sources. I talked about in my book, data being the foundation of your stories, if you have a weak or erroneous data in your foundation then, you know, that kind of impacts your whole story.

Ravi 1:02:18
I love the diligence that you’ve put into this, Brent, because yeah it is very tempting to say “Oh yeah, this famous guy has said it”, and part of your mind is claiming “Yes, I agree with that. That sounds true. Let me publish it”

Brent 1:02:35
And the interesting thing is – so I have a whole chapter on the psychology of data storytelling – and one of the things I found is that…I’m forgetting the name of the psychologist, but he noted that when we get data that’s positive, we don’t scrutinize it like we do when we get data that’s negative. If it’s a positive detail, like the example that I saw in an article was, when we’re trying to lose weight, and in the morning when we step on the scale after working out, and we see that the scale says that we weigh two pounds less than we did yesterday, our reaction is not to question that data. No, it’s like, “All right, yeah, I’m good”, and you hit the shower. But what if you weighed two pounds more than you did yesterday? All of a sudden, you’re like, “Wait a second, am I wearing the same clothes? Is this scale level? There’s something wrong, yes, maybe it’s broken.” So, it’s interesting how it’s the same thing when you’re writing a book, and you’re looking for research. When you have research that validates a point you’re trying to make, I think the mistake some authors do is that they don’t investigate that research to see if it’s really strong research. You can’t reproduce the experiment, but at least you can see if it is a reputable source, or if it’s real research, or is it something else? And that’s the danger when you’re writing a book and you’re trying to make key points, and then you find data points that support your key points, do you go the extra effort to just validate whether they are really solid or not?

Ravi 1:04:20
That’s great work, Brent. One challenge that I feel (might occur when) folks want to, read your book and try and apply these concepts, is that there’s a lot of frameworks – I completely go through the same challenge – Do you sometimes feel that it is what it is, that there are a lot of frameworks, yes, and each element or each part of data storytelling needs some set of skills or set of ideas, etc. Or do you feel that there is a need for some sort of a simplified overarching framework, which is of course, the Three Pillars, but which elegantly ties everything together and makes it easier for the audience?

Brent 1:05:11
I like frameworks. Often, when I’ve had readers reach out to me and say they’ve enjoyed my book, I ask them what part of the book they enjoyed. And surprisingly, I’ve seen a diversity of opinions. In my book, I provide a wide swath of different frameworks and principles that you can leverage. Now, which ones you take, you can pick and choose and you can basically say “This one is really helpful for me because I struggle in that area.” Everybody can pick the frameworks that they need, and that help them; and maybe some people already have a framework that works for them in that particular area, so they don’t need my framework. I’m a library of different frameworks. And if you have certain areas where you’re weak, or you need help or templates to follow, then you’re going to have that; but does that mean you need to apply and use all of them? No, I think that people can make the decision on which ones they feel are relevant to them or helpful to them, and I’m not going to say that you have to embrace everything or you can’t have my book, no; It’s more of “Here are a bunch of frameworks, and if you find some useful to you, then leverage those ones.”
And, who knows? Maybe I’ve had people re-read the book, and then they get a different perspective on a different framework that maybe the first time, they didn’t appreciate because they weren’t doing data storytelling, and now that they’re actively doing storytelling, that framework becomes more useful to them because they’re like “Oh, wait a second. This is actually useful. I didn’t totally understand it the first time but now I’m now that I’m hands on with storytelling, this makes a lot more sense to me.”

Ravi 1:07:04
The other interesting thing here, is that people tell data stories in very different contexts, right? So there’s a guy who’s a Marketing Analytics guy, who’s maybe presenting to the CMO. There’s a finance team, who’s maybe presenting to the senior leadership. There’s a consulting firm, which is presenting to a client, and maybe talking to them about whether they should invest in this new project or not; And there are research guys that prepare research reports – that’s also at some level, a data story. I sometimes try and classify different types of data storytellers, so, do you have any broad classification in mind that you could divide data storytellers into two or three big buckets?

Brent 1:07:57
Yeah, I kind of view data storytelling (as being) for everyone, that everybody can use it. You can have executives who can do it, you can have managers who can use it, you can have analysts who can do it; I think, universally, it’s something that can benefit everybody. Now, with organizations that may not be ready to turn everybody into data storytellers, then it may make sense to have dedicated data storytellers; people who are comfortable with the data, have a passion or an ability to communicate effectively with data, to have them start to tell data stories within an organization, and often, there’s a lot of talk about data literacy, and how companies are struggling with reading and using and working with, and even communicating (data). I mean, data storytelling can be considered under the umbrella of literacy. But I think in general, companies are struggling to get people to read, and interpret, and understand visualizations and data. And so, having dedicated data storytellers out there, that are either building data stories on their own, or working with teams to help them form their data stories can be valuable. Especially in the early stages of data storytelling. I almost think of them like personal trainers. They’re helping people who don’t know what workout programs to use, or what diet they should follow, and having somebody who’s very skilled at exercising and is that personal trainer can kind of (help them by saying) “Well, here’s a plan, I think you’re going to want to do three sets of this, and then three sets of this.” And so, they can coach and mentor other people within the organization now. Even just sharing data stories on a regular basis will expose people to data, get them more comfortable with interpreting data, and then maybe eventually they’ll get to a level where they can start to tell their own data stories.
In the analytic space, we talk a lot about data democratization, so that’s like building these tools to put data into the hands of lots of people within organizations. When you do that, you don’t know if it’s always going as planned, and I think people and different companies are at different levels on that journey. But, if people are really getting access to the data and they start to find insights for themselves, then there’s going to be a need for people to be able to tell or communicate those insights effectively. Now, the interesting thing, and this is another thing I mentioned in my book, it’s that if I have a dashboard, and I’m consuming data or analysing – maybe I have an analytics tool, and I’m analysing some data there, and it only affects my role, like maybe I’m managing paid search campaigns for my company. And I find that there’s some kind of opportunity in the data that I can make a change in, and maybe save the company some money from being more efficient with our spend; I can make that change on my own. In most cases, I don’t need to go for approval to anybody, “You want me to build a data story for myself?” No, I don’t need to. The only situations where we need to build data stories is when we have a bigger insight that we can’t act on alone, and we need to communicate it to other people. And, we communicate it to those people for various reasons, because we need their buy in, or I need approval from a manager to work on this, or maybe I need resources or budget, or I need to do need to collaborate with another team and get their buy in. So, there’s all these reasons why we share data stories, but it’s in those cases where we need to communicate an insight to other people. You’re going to see that more and more where, either on a team or between departments, or with the executives, there’s always going to be scenarios where we need to communicate to others.
In terms of whether there are different types of data storytellers, I think it’s more about how comfortable people are (with the craft); how much practice they’ve had once they’ve learned the principles. And obviously if you’re a data scientist, you’re going to have many more tools at your disposal to find insights, than maybe an executive who really is only looking, or maybe is dependent upon the team to do the analysis for him or her. And then, that individual or that executive, she may need to then go and communicate that to the other executive team members. But I see different scenarios, and there has to be a base level of data literacy. As I talked about this in the book, they have to have some comfort with working with data. Now, that’s a scale; you can have people who are data scientists, but you could also have an intern, who’s just comfortable working with data, and that’s fine. And then the next level is obviously, curiosity, I think. That’s also a big part of being a data storyteller, is that you’re curious. And you could look at it from the main analysis level, that’s one level, where I’m going to go in and I’m going to find things; but I think an executive could also be curious. They may not be the one doing the investigation, but they could be leading the investigative efforts and saying, “Oh, that’s it. We got an interesting hook here. What’s contributing to the spike?” Basically, asking the right questions and guiding the team. I could see an executive crafting a very powerful data story because they’re taking their domain expertise, they’re taking their understanding of the strategic priorities for the business, and then they’re combining it with (their guidance and saying) “Okay, dig here, tell me what you find”, and then building the data story. So, I think there’s different scenarios for storytelling.

Ravi 1:14:20
I love that analogy of the personal trainer. It sparked a thought, you’ve said that using data literacy, which will be at different levels in the organization, similarly data storytelling skills will also be at different levels in the organization; do you think there’s a role for some sort of data storytelling team, or a Chief Data Storyteller, so to speak. Who’s got a few folks working with him, and who basically go around the company and help identify these stories which need to be told, and then bring them out?

Brent 1:14:55
Yes and no. It comes from my analytics experience, because there’s two models for doing analytics, there’s the centralized team approach, where, basically, all of the other functional teams need to go through the analytics team to do any kind of deep analysis; Now, the challenge with that model is – I mean there’s benefits and there’s drawbacks – The one thing (about) having a centralized team (is that) they all work together, they’re very familiar, but the challenge is that they may not understand the business. They may not have the domain expertise, and they’re removed from where the action’s happening, essentially. So, the other model that I’ve seen in analytics that works really well, is the Hub and Spoke. You have a central team which is smaller, but then you have spokes, where you have dedicated analysts and each of the teams. And, because they’re in those teams they understand the business context, the business domain for that team, and the challenges they’re facing. And then, they’re able to apply the analytics to that.
So, a couple of questions come to mind: do we need to create yet another team? Maybe. But I would say that would be a short-term solution to get things moving. I would say, eventually, you would have to have those skills within the analytics team, or within the business teams, and then it’s just kind of there. So, I don’t have a problem with having a team set up to spearhead and get things moving, but I think (in the) long term, that these skills just need to be a part of how we work. It’s funny, if you think back to the when the web was first starting, we had a webmaster, how many webmasters are there today? Now there are none, because the Internet and the web has just been embraced by all different teams; and so, I think that, initially…because we’re just beginning, there might be a need for this and some dedicated data storytellers, or a dedicated data storytelling team. But I think long term, I don’t think that makes sense.

Ravi 1:17:16

It’s such a core part of your work. I love that, Brent. Who would you say Brent, are some of the individuals or companies that you look up to the most, who practice the skill of data storytelling really well?

Brent 1:17:33
I think we’re still in the early days of data storytelling, and again it goes back to that problem I raised earlier, that there may be some really good data storytelling going on that we just don’t know about, because it’s happening in a company. Any Fortune 1000 company may have some very good data storytellers, who are doing some amazing things, and those companies know who they are, but they’re also not going to advertise them because they don’t want them to be recruited away. The challenge again becomes, if I went to these companies and said, “Hey, we’ve heard that Kevin and his team are doing an amazing job, would you share some of those data stories?”, and they’ll say, “Well, it’s all proprietary. No, no, we don’t want to share those stories.” So, that’s the challenge on the corporate side.
Now, one of my heroes, and you probably got it from the book, is a guy by the name of Hans Rosling. He’s amazing. You can see his TED talks, they’re (available on YouTube), and unfortunately, he passed away in 2017, but I still look at his work with great admiration. I think, he did an amazing job as a storyteller; very complex data, and so I look to him as a good example of how to tell data stories. 

And there’s some interesting things being done by different organizations, like the Wall Street Journal, or The Economist, or, New York Times, Vox is another one. They do some amazing data visualization/almost data stories. And so, I think those are inspiring. But the danger there is, the tools and the approach that they’re taking with those is very different from what we do in a typical business scenario. They’re using different charting libraries that are just not available to (the rest of us). Yeah, so, the tools that we’re using typically are Excel, Tableau, Power BI, Looker, those are the kinds of tools that people are going to be using for doing their analysis and building data stories.

Ravi 1:20:09
I agree with you. That’s a bummer, in terms of not being able to actually get access to good quality corporate, hardcore data storytelling. On the narrative side, you’d probably find more, right? For example, I really like the way Jeff Bezos writes his narratives, whether it’s the annual shareholder report, or Warren Buffet when he writes a shareholder report, but you’re right that they are not very data intensive, they are more narrative (focused).

(Is there) any other question, Brent, that you feel that I’ve missed, and would like to share.

Brent 1:20:49
It’s been a very good conversation. Nothing’s coming to mind. I think this is a skill that is something that we all need to learn. (One of the things) that kind of showed me how powerful this topic would be, was when I published an article on Forbes and I titled it ‘Data Storytelling: The Essential Data Science Skill Everyone Needs’, and I think it’s had three or four hundred thousand views, and it’s been very well received by people. And when I was thinking about writing a book and when I saw the positive response to the article, I was like “Okay, there’s something here.” And that was my journey. When I present at conferences and stuff, people would come up to me afterwards and say, “Do you have a book, or could you recommend any books?” and that’s where I was like, “Well, specifically about what I’m talking about? No, I don’t think there’s going to be a comprehensive book like that.” And so that, again, reinforced to me that I needed to get out there and publish this book. Now I’m taking the step to kind of go out there and start my own company, and consult with companies around data storytelling. Yeah, it’s exciting. I’m happy to be combining some passion areas: I love narrative, I love data, I love psychology and human perception, and aspects of it. And I also just enjoy presenting, too. I’m actually somebody who enjoys sharing insights and telling stories.

Ravi 1:22:47
That’s amazing. I’m so glad that you took that spark from that article and decided to write this book, because I’ve learned a lot from Effective Data Storytelling: How to Drive Change with Data, Narrative and Visuals, and I’m going to be recommending it on my blog soon. And where is it that people can get in touch with you or get to know more about you?

Brent 1:23:08
Yeah, I think the easiest way is to connect with me on LinkedIn. You can find me there, I’m Brent Dykes. I also have a website for my book, effectivedatastorytelling.com, so you can learn more about my book there. And then on Twitter, I’m on the handle @analyticshero. I love connecting with people who have a similar passion for business communication, and in particular data storytelling. So please, do reach out to me, I look forward to it. Now as I go independent, forging my own path forward, I hope to share more content and work with more individuals. I kind of see this is a movement that needs to occur around combining data with narrative and visuals, so that’s my mission. I love this stuff. So hopefully that comes through, and you can feel that passion in both the book, and in this podcast.

Ravi 1:24:10
That’s amazing, Brent. Best of luck on this new journey, I’m sure you will write a great story, and I look forward to seeing how it plays out.

Brent 1:24:18
Alright, thanks Ravi. I really appreciate it, and thanks for the opportunity to come on this podcast.

And that was Brent Dykes, author, speaker and storytelling coach.

A few things which stayed with me:

1. The Story Zone: Not everything needs to be told as a Data Story
2. The contrasting stories of Dr. Semmelweiss vs. Florence Nightingale and Dr. John Snow – which indicate the criticality of this skill
3. The need to ensure that you have Relevant, Comprehensive and Credible data – and double check your hypotheses.

If you find this content valuable, please rate and review this podcast on iTunes, Spotify, Google Podcasts, or wherever you listen to them. It’ll help others like you discover these insights!

This podcast was hosted by me, Ravishankar Iyer. Audio editing by Kartik Rajan. Transcript editing by Amisha Jha and all-round support by Sanket Aalegaonkar.

Until next time, may the force of good stories be with you