Thomas: Oh, it is episode 15 of season two of the Presentation Boss podcast. It is an exciting one today. It is another guest interview, and if there is such thing as a data storytelling hero, we have one here today. Um, Kate. Kate, I think you might be even more of a fan girl than I am.

Kate: Yes, definitely. Brent Dykes is our guest today, and in a previous episode, if you listened to Data Journey episode, that was basically directly from his book, Brent wrote, effective Data Storytelling. Uh, we have many data storytelling books that we have, you know, taken from and learned from, and upskilled from over the years.

And the reason that I like his books so much is the way that he articulates things that you kind of explicitly understand and kind of a feeling that you have. And he just manages to put words around it really nicely. His analogies are really good. Um, just the way he, I, I guess, communicates is, oh, I’m gonna say effective.

Thomas: What it says on the title.

Kate: Yep. What it says on the tin. Yep, yep, yep. So, um, it was a fantastic conversation with him Again. He used some new analogies and, um, yeah, just a, just a good conversation.

Thomas: Yeah, really, really good conversation. I currently have stolen Kate’s copy of the book or you know, the business’s copy of the book. I dunno how that works. Um, and it is unreal reading it again now ’cause it’s been, it’s been a while, um, reading it and just how much we are on exactly the same page about best practise, about how to go, about data storytelling, about how, um, AI is clearly changing the world of data and analytics and We had a little chat in there about what has changed since this book came out in only 2020.

Kate: And I just wanna say beforehand, we did very much try to make this a conversation about these are all the questions that we really wanted to ask from him and not just a, um, fan girl conversation. But let’s have Brent’s official bio.

Brent Dykes is an internationally recognised expert in data storytelling, an author of effective data storytelling, how to drive change with data, narrative, and visuals. Brent has nearly two decades of experience in analytics working with companies, including Adobe, Microsoft, Nike, and Sony. He’s the founder and chief data storyteller at Analytics Hero, a Forbes contributor, and he was named the Digital Analytics Association’s most influential industry contributor in 2016.

His work today focuses on helping organisations move, beyond simply presenting numbers to communicating insights that influence decisions and drive change.

Thomas: Sounds good. We go on a deep dive. Here is our conversation with Brent Dykes.

Kate: welcome to the Presentation Boss podcast, Brent Dykes.

Brent Dykes: Hello.

Kate: Brent, we have just heard your official bio and I am well on the record as saying that effective data storytelling is perhaps my favourite book on data storytelling, and I have them all. Um, but I don’t know anything about you personally. So what can you tell us about Brent Dykes when you are not at work?

Brent Dykes: Yeah, uh, originally from Canada, uh, moved to the states, got my MBA down here, uh, started my career in marketing. it was de debating between accounting and marketing and, and went the marketing route, which is the, probably the, the road less travelled for a lot of the people that did well in accounting classes.

And, and then I got into marketing analytics and worked for Adobe for many years. And, uh, and then for the, worked for them for 12 years. The company I was working for at the time was acquired, uh, Omniture. And, um, and then I went to a company called Domo, which is a, uh, cloud BI solution. And then worked briefly for an analytics firm called Blast Analytics. And then for the past five years, I’ve been doing training around data storytelling and coaching and, and consulting in, in this area. So, yeah, outside of work, I, you know, love to travel and obviously have a big, uh, you could see the comic books in the background there, so definitely like the, that’s another form of visual storytelling.

So, you know, take some lessons from that. But yeah, that’s kind of me in a nutshell.

Kate: Amazing. So in a, in a nutshell, you said you do consulting and, and, um, coaching. Um, how did you get involved in Data Story and what led you down that track?

Brent Dykes: Yeah, I, I worked in consulting for a long time and, and at Adobe we would have, um, kinda like customer kind of conferences. I was thinking of like, this is back in 2013, I was thinking what would be a really interesting topic? And I started to hear about data storytelling and I was like, wow, that’s, that seems like that would be really, uh, aligned with me.

’cause I, I, you know, I was always complimented on my presentations and my presentation skills and I really was a big proponent of storytelling and, and visualisation and all that. And, and at that time, you know, I probably hadn’t fully formed my ideas that I have today, but was an interest in that area.

And so I pitched to the, um, the committee that was deciding which topics that would be breakout sessions for this customer conference for Adobe. And, um, I, I, I they, they took up my idea so I threw myself into the topic, um, did a, did a presentation on it, and it was one of the top five presentations at the conference that year. so I knew I was onto something and it was really well received. And, and then the next year I came back, did the same topic again, expanding on my ideas. And then, and then after that I started presenting at other conferences outside of the, the Adobe, um, world. And, and then I started having people come up to me and, and say, Hey, are there any good books that I can read on this topic?

And, and I was like, not really, there’s not really a book out there. Obviously Cole’s book came out a year or two later. and then that, you know, that then I was as inspired by her book and, and maybe people asking for information on data storytelling. And, and then I felt, you know, like I needed to get my perspective on data storytelling out there.

And that, that kind of started me. And I started writing back in 2016. It took me three years to write the book. I, it took some time to really articulate my ideas and, and come up with the philosophies that I have and frameworks and yeah, that’s kind of how I got into data storytelling.

Thomas: Uh, I think that answers the question. I, so Kate is definitely on record of saying it’s, it’s her favourite data storytelling book. I currently have stolen it ’cause I need to go through some of the tabs and look some things up. I was, I was literally reading yesterday again, back through your, um, love and admiration for Hans Rosling, which I share is probably my first, um, touch point with proper data storytelling or, you know,

Brent Dykes: Yeah.

Thomas: data communication as you would’ve called it back then. What, what was those early days, sort of before Cole’s book came out? What were those early days? Were you just kind of reverse engineering the likes of Hans Rosling, or how did you come up with those philosophies?

Brent Dykes: I mean, I, I started looking at you obviously. I, I took it literally like storytelling. I think a lot of people have emphasised the visualisation part of data storytelling. And then I was concerned actually that I was seeing a lot of just. It was almost like data storytelling was a synonym of data visualisation. And I felt like it was being over rotated too much to the visualisation aspect. And I felt like we’re, we’re neglecting really that storytelling and the story itself, being based on data and evidence and facts. Um, but storytelling is, you know, as I started to research the topic, I started to see, wow, like this native you know, innate skill that we have as human beings to tell stories is really powerful.

And, and so I, you know, in my book I go into the psychology behind data storytelling, I, I really felt like, um, it was important to kind of capture that. And then the other interesting thing that I ran into at the time, and I still, still hasn’t gone away, is I noticed that there are a lot of, like, um, people were using a lot of, um, statistics about data storytelling. I started digging into them, ’cause I, I’m, ’cause I’m very data-driven myself and being an analyst, I wanted to be true to the, the statistics. And I found that a lot of these studies that people were quoting and these facts, you know, from experts were actually not backed up with any data or evidence. so I, I actually had to remove some of these quotes from my book that I wanted to use, you know, and I even had, you know, confirmation bias that, that, hey, you know, this supports this premise that storytelling is important and valuable and everything. But I couldn’t put it in the book because it, it just wasn’t backed up by actual a, a, a verified study or research. And, and if you go to my blog post, one of my first blog posts I did was five, you know, things that I found that I, I was like, these are data points and statistics that a lot of experts would quote in their books, but I could not find any evidence to support them. And so, you know, that that was part, again, part of my journey.

I, I found that there was a lot of misconceptions. There was a lot of, you know, I really love the work of Hans Rosling just at a raw level. His energy, his his, um, passion, you know, and, and, and, and his desire to kind of help and shape minds to understand, you know, some of the health. You know, he, he focused a lot on health related, statistics and, uh, the joy of statistics.

You know, I actually, in my book I cover, um, a short video that he did with the BBC and um, actually licenced that content. So I could actually share that example in my book. That was one of the few things that I had to pay to licence, but it was well worth it. ’cause I, I, you know, I am a huge fan of Rosling’s work and I, I wish he was still around today, but unfortunately he passed away in 2017.

Thomas: Yeah, yeah. As, as we record this, um, the last episode that came out was a breakdown of religion and babies, which is potentially, uh, like favourite, uh, online talk ever. Um, it’s, it’s interesting though, like flipping through your book. It’s very clear, like you say, the mix between data and narrative and visual, it’s really clear.

It’s quite, it’s a how to guide. I keep looking at it here on my desk to make sure it hasn’t escaped. Um, I guess such clarity in the book, and, you know, we’ve both been working in the space of teaching people to communicate data for many years. Why do people still struggle with it?

Brent Dykes: I think there’s a lot of, um, advice and guidance around how to do analysis. There’s lots of guidance and, and advice on how to use certain analytics tools. There’s lots of advice and guidance on how to do visualisation, don’t feel like there’s a lot of really sound, uh, guidance on the narrative and how to form a narrative arc and how to do some of that.

I think there’s a lot of models out there on the storytelling front. So you, you hear the, the, the, you know, uh, Aristotle’s three Act structure. There’s the hero’s journey, there’s, uh, from Joseph Campbell, there’s, um, save the Cat, there’s, um, uh, there’s the, the um, storytelling, you know,

Kate: That’s like 14 steps in the Pixar one.

Brent Dykes: Oh, yeah. Oh, well, actually, I think the, the ho the, the, the hero’s journey is even more complicated than that one. so there’s, and, and there’s lots of different models and, and I think they’re, they’re coming from different places too. You know, there’s some that are coming from the, the world of fiction and drama.

There’s some that are coming from business where maybe it’s more around, uh, you know, sales and cust and customers than, than you have other models that are, you know, even, even some people might include the, scientific method as a, as a means of telling a story, you know? And so there’s, I think there’s a lot of confusion, and I think there’s a lot of different approaches.

And, and, and so it gets really hard to kind of say, how do I do this the right way? And, and maybe there’s even some more, a little bit more subjectivity on the, you know, if, if we’re looking at visualisation and narrative. I, I mean, there’s definitely an art component to both of them, but I would say maybe art is a little bit more on the narrative side, than it is on the visualisation side. Um, so I, I think it’s a mixture of things. I think there’s a lot of confusion out there. There’s a lot of different models, a lot of people are teaching it different ways, and, uh, and there’s just also not as many guides, not as many kind of, um, compasses that we can follow on the narrative side, especially when it comes to data.

Again, you know, we’re trying to, you know, when I took the model that I use, so I, I base mine on Freytag’s Pyramid, uh, which if you’ve studied, you know, literary kind of, you know, it’s kind of this traditional kind of building up to a climax, and, and the, the story arc. And I, but I, even then, I found that his model wouldn’t perfectly translate over to our world of data storytelling.

So I had to modify it, and I, and I came up with my data storytelling arc, you know, which is fairly close to what he had, but I, I did have to make some modifications to make it work for, or interpret it in a way that, that people could then say, oh, okay. I could see how this could work for, know, sharing my findings or my insight with, with a, with an executive or a decision maker.

Kate: So on that, yeah, we, we really talk a lot about bottom up storytelling, you people creating their narrative and, and sharing that up the chain. I guess, do you think that there’s too much pressure on, you know, that layer , communicating upwards when maybe the requests that are coming down from leadership need to be better?

Brent Dykes: I go back to business school. We were trained to do things in a certain way. We have, you know, and, and we’re trained even culturally within organisations, you know, it has to be quick, has to be to the point, you know, we, we have, Mintos, uh, pyramid, you know, where, or Pyramid principle, you know, that comes from Barbara Minto and her work at McKinsey. Uh, and obviously the, the executive summary being a big focus. And, and what I’ve tried to do is not necessarily say that that’s not valuable or that’s not important, but I’ve tried to push back against everything needs to be turned into an executive summary. Everything needs to be snappy and quick. know, and, and I published a Forbes article about this, that, that I feel like, you know, especially with AI now, like AI’s ability to quickly turn anything into an executive summary, you know, traditionally that had been a skill that, you know, like if you think back when you’re creating an executive summary before AI, it, it wasn’t just a simple thing.

It wouldn’t take five minutes. It’s like I’d have to go through and really pick and choose what elements would go into that executive summary to summarise a, a long story or a bunch of research that I’ve done and, and findings that I had. And, and now you can do it like that. but, but I still, I still feel like, you know, there are moments when an executive summary is the best approach, and then there are other moments where people really need to understand this topic or this subject really deeply so they can make the right decision.

It’s if they don’t understand how everything connects in the context around those, looking at three bullet points and making a decision on something could actually lead to the wrong direction or the wrong decision. And, and, and there are those key moments, uh, that, that probably come up every so often where you really do need to have the full context that a story is the best approach we do need.

You know, we can’t just cut it down to a five minute decision. No, this, we need to spend an hour on this. We need to talk about this, we need to talk to other teams. We need to get everybody on board. We need to, you know, figure out what is the best course of action. And sometimes the data story is gonna be the best approach.

Now, I still realise that we work in a world where, you know, there’s, it’s not, it’s not common that you’re, you’re gonna get an hour with a C-level executive. You know, you’re probably gonna get five minutes, 10 minutes. So what do we do in those situations? And then I’ve then proposed, okay, well there’s another short tool that we can use besides an executive summary.

I call it the story, the story, um, teaser. I’ve called it the data trailer, but it’s like the short version of our story. And so we take the, we take the hook, we take the aha moment, we share that. And then all we’re trying to do with that is we’re just trying to see do they want to hear the rest of the story?

You know? And if, and if they make that decision at that moment that, yes, I want to talk more about it, I want to understand this more deeply, I now have permission to tell them the force story. they can make that decision at that point. You know what, no, I’m, I don’t need to hear more. Brent, I’m, I’m good.

Thank you. We’ll move on to the next topic. You know, so I think it’s going back to your original question. I think there is pressure to keep things succinct, to keep things concise. I, even if we tell a story, I think still having that story structure, tells us what should stay in the story and what goes in the appendix. I, I think one of the, the common challenges that we run into when you, whether you’re in a, a, a data intensive team like analytics or, or you know, BI or if you’re in finance or accounting, a tendency for us to inundate audiences with too much detail, too much information. so we, we do need to kind of like. Pair it back to what’s essential. And I think having that story structure, that story to kind of guide us on what should stay in the story and what comes outta the story is still valuable. Even from a, you know, yeah, like maybe we can’t get it down to five minutes, but I can, maybe I can get it down to 10 or 15 minutes with a story structure.

Whereas if I didn’t have that story structure, what would I start doing? I’d probably spend half an hour, 45 minutes, an hour just going through throwing stuff at the wall, you know, seeing if something sticks or not. And then if it does, you know, and, and, and so I think there are, yeah, I think there is pressure from within organisations to be concise and to do what e executives want us to do. I do think there’s room for executives to maybe be a little bit more open-minded and maybe recognise when there’s those moments when, when I need to hear a story on this, not just a, an executive summary.

Thomas: Definitely a decision to be made there. And I mean, you’ve, look, you’ve opened the AI can of worms, so we have to go there now. And it was, it’s amusing to me like reading your book. Um, ’cause what was it, 2020, it came out as something like in the next ten years

Brent Dykes: Yeah.

Thomas: artificial intelligence. And I’m reading this like it happened, it happened.

Um, I guess, um, and in some ways AI is making analytics a lot easier. In some ways it’s probably adding complexity, kind of as you mentioned there, with knowing exactly what to do. guess

Brent Dykes: Yeah.

Thomas: in the age of AI, which parts of data storytelling do you think will remain uniquely human? That we are not gonna be able to outsource to a robot, or at least not yet?

Brent Dykes: I think there’s two parts to the, to that question. One part is like, if we think about all of the things that we can do with AI, what are the things that I want to give to AI? It’s all the things I don’t like doing, repetitive and boring. Okay. So that’s the first thing. I wanna give all of that data cleansing work to an AI tool.

I want to give anything that’s repetitive, you know, time consuming. I don’t get joy from it. I don’t enjoy doing it, so I’m gonna give that to a machine to do so. And, and maybe even there’s, there’s, you could even argue there, maybe the machine could do repetitive things more efficiently and without less error than I could as a human being. Now, on the flip side of that, there’s gonna be things where I enjoy and, and, and I think that state of storytelling is one of those things that human beings enjoy to communicate. And, and so that’s the first thing that I’ll say. I don’t think that we should outsource things that we enjoy and things that we are good at. I mean, obviously I’ve talked about how, you know, innately we are storytellers, so I think there is that component. Maybe some people out there are like, I hate storytelling and I wanna outsource it to a machine. That’s fine. but in terms of the actual things, I mean, I really think it can touch all aspects in all areas of the entire process. I think there’s a build component and there’s a QA kind of audit component to AI, and then if we look at it from a, perspective, like I, I talk about it as macro and micro, right? So if we look, if we, I think there’s been a lot of emphasis on the micro side of data storytelling being automated through AI. Meaning I can create a slide and it’s gonna help me with the headline. It’s gonna pick the chart for me. It’s gonna highlight things, it’s gonna build the chart, you know, it’s gonna do a lot of that manual kind of stuff. And I can generate a slide very quickly to kind of communicate something to an audience. I look at that as the micro side. what I don’t feel is being thought of, and, and I think it is equally important as well, is just how do all of these, I call them scenes. I mean, obviously they’re slides as one form of a scene, but. Data scenes, how do these data scenes come together into a holistic story? And I think that’s not being looked at, that’s not being considered. And I think o often what I, I probably won’t use AI build from scratch. Like too often I’d, I’d want to have one hand on the wheel sort of thing. You know, like I think you’re into self-driving, you might say, oh, I trust the AI to just drive me to the mall and we’re good.

And I’d probably be like, I don’t know, you know, I a I don’t know if it’s gonna take the right, uh, route. I don’t know if it’s gonna be as safe as I want it to be. Uh, you know, maybe it’s a trust factor. But anyways, I think there’s always a component where I want to have a hand on the wheel guiding and, and, and shaping where we go and what route we take. And so it really comes back to more of the feedback. I think I see tremendous kind of like, power in AI to give me feedback on the different components of my storytelling. I’ve, I’ve done that through my own, going back and looking at some of the data stories that I’ve done. I’ve fed it into AI and then told, told it, you know, like, where could this story have been improved?

You know, and I’ve, I’ve uploaded a primer. I’ve got AI primers for both micro and macro. And so it can tell me, you know, oh, you know, you kind of jump from this one scene to this next scene and you didn’t provide enough context or have you thought of adding something, going deeper into this? ’cause depending on your audience, maybe this would be valuable. so I think as a companion, as a way to augment what we do as data storytellers, I see tremendous power. Now obviously there’s gonna be people out there that do want to just it, but then the question I would then push back on, and this is actually a post I’m writing, I was writing before we met, do you know what good looks like? Right? So if you want

Kate: yep. Yep.

Brent Dykes: if you wanna entrust AI to do everything, okay, but it, that doesn’t, that doesn’t surrender your need to be able to evaluate how well it did. Right? You, you have to be able to go in there and, and I think my fear is that a lot of people are not getting trained up on data storytelling or don’t have the prerequisite skills or knowledge and, you know, and now AI can spin up a data story or can spin up, you know, a, a, a data scene and they don’t know what to look for.

They don’t know where to question, they don’t know where to poke and product to really know, Hey, this is, hey, this did a great job. I don’t need to touch anything. Or, oh, wait a second, it got this wrong. Or, or I think this could, maybe if I did it this other way, it would look better. or it, it would come across more clearly to the audience, you know?

And so that’s my fear is that if you don’t know what good looks like. I, I think that’s a real limitation on what you can achieve with ai, you know? But if you do know what good looks like, wow. You know, you’re empowered to really do some amazing things. So you know, from a, and, and maybe this is a bias as, as somebody works on training and, you know, training organisations and stuff, uh, you know, that I still feel like, you know, uh, for a time there I felt like, oh, am I obsolete now?

Am I going away? Can all of my knowledge and information go into a tool and it can fully do everything? And, and I think, no, I think there’s still that judgement . There’s the, you know, there’s the human element to this, especially in storytelling. I think, you know, maybe, maybe building a dashboard could be a little bit more straightforward.

You know, maybe, again, it’s a little bit of bias against dashboards, but I feel like there’s a little bit more clear cut how things could go on that front, on that, uh, aspect. And, but with data storytelling, I think there’s, there’s more factors at play uh, will require us to play a, an still an active role in, in shaping and, and building and, and telling those data stories.

Thomas: I think genuinely that is a fear where people who work with data tend to be interested in tech and so adopters of AI and I I I do get concerned that we’re gonna see just a data set put into a robot and said, Hey, I need to find an insight and communicate to this executive who is a direct type communicator, make it happen.

And then it pumps out something which is probably pretty good. And it kinda like you said with self-driving cars, like my history is in aviation, I stand by autopilot, can probably land a plane better than a pilot can, but would you get in a plane that doesn’t have two people sitting in the front seats?

No, you wouldn’t. Right? Like

Brent Dykes: Yeah.

Thomas: so much of this, this element of, of QA required that know what good looks like.

Kate: Yeah. Um, like my fear is, is not my fear, but, um, the fear out there is very much that AI is going to take over all of these kind of analytical roles. Um, my fear is, is the same that, you know, is my role becoming obsolete? If, if there are no analysts left to train, what happens to my role? Um, so, so for people who are in that role and they’re like, well, that’s my skillset.

What can they do to, to continue being like properly influential? What’s going to distinguish those people?

Brent Dykes: Yeah. I mean, I, I think rather than running from ai, I think you have to run into ai and I, I think that’s where you, when you start using ai, you start to see. The, the cracks, you know, you start to see the gaps, you start to see where, oh, that’s not quite as good as what I could, I could have done. I think it’s rather than like just poo-pooing it or, or ignoring it or, or hiding from it, I think we have to embrace it and we have to push it to the limits and see what it can happen, what, what we can do with it.

And obviously the, the technology is constantly evolving at a, a very fast pace. So, you know, maybe some of the, the early kind of, mistakes or gaps that I saw are now no longer gaps, you know, and, or, or, or the gap is much isn’t as wide as it once was, you know. But I still feel, I mean, I’ve, I’ve always been very, um, I feel like there’s a human part, to what we do in analytics and, you know, and that could be understanding the audience, that could be, um, you know, make, obviously making sure that the data’s sound and reliable and, relevant and all those things.

And, and then, know, I, I just think that there’s a lot of cracks and, and, and not everything is covered the technology. You know, there’s gonna be a role for us now that role’s probably changing and, and what we bring to the table is gonna be different. And, you know, and, and I think there are people out there that be significantly impacted.

I, I know people that were very comfortable with producing automated reports and,

Thomas: Yeah.

Brent Dykes: and I would even, this is even before AI, I was like looking at what they’re doing and I’m like, how can we not automate this, you know, with the existing tools before AI yet they would fight it, you know? And I, I would think that there are, there’s gonna be a segment of, uh, data and analytics professionals that will fight, you know, automation when, when they’re, when their, their contribution should be automated. Um, and, and I, so I think there, you know, I, I like to think of it, one of the analogies I like to talk about is the analytics marathon. did a Forbes article on it. And, and so if we think about this race that we’re going on, um, ultimately we wanna drive action, you know, but that’s at the end of the race, right?

So we go through these steps where we collect the data, well, obviously we have, we start with the data strategy, we collect the data, then we, we visualise the data into rudimentary reports and, and, and different things, dashboards. Then we might analyse the data, you know, when we see anomalies or patterns or trends that we need to understand better. then when we find an insight, we need to communicate that up or down through the organisation. That’s where the data storytelling comes in. ultimately we need to take action on it. And what I feel like, what, what happens with a lot of analytics company or a lot of companies and their analytics teams. Is they go through the, they get through the first few gates very quickly and, and, and regularly, right? So they collect the data, they visualise it in reports and dashboards, and then they kind of, well, they process it, and then they do the reports and dashboards, and then they kind of stop and then they go get the next data set and, and they don’t see it all the way through.

And I, I feel like that’s the last mile. You know, we need to, we need to finish the race. Like imagine, you know, my wife runs lots of marathons and Ironmans and stuff, and the goal of running that race is to get the medal at the end, right? To, to do your, you know, to get the reward of that medal to show that you accomplished that.

And I, I feel like to, for the last several years we’ve kind of lots of races, but never finished them, never really driving action. And, and, and the, and the value that you get from the, from doing all this work really comes at the end that we’re making decisions, we’re taking action on whatever we’re analysing, you know, whether that’s operational data, marketing data, sales data, whatever, you know, like we’re making decisions based on this data and, and optimising what we do and optimising our business performance.

And I feel like too much work is going into the reporting and stopping at reporting and, and assuming that that’s, that’s done enough and that, that the value that we get from all of this effort is sufficient. And so I would say I think we need to shift more into the last mile, you know, into that last mile of getting and driving value.

And let’s, let’s, let’s give, you know, like the, if there’s elements of the collection and the processing of the data and, and maybe the building of dashboards and different things that we can automate and give it to the machines to kind of handle that. and, and again, I’m not saying that there’s not a human component to that as, as well.

I think there is just how much time is spent there? That’s the real question. I think we can shift more of our time to strategic areas that are on that last mile and really driving business value from the data rather than just, here’s another set of numbers, or here’s a new report, or here’s another dashboard. know, on top of the thousand other dashboards you already have, you know, it’s like.

Thomas: Let me quickly tell you about our online community of analysts and people working with data in this brave new world of ai. I think if, I think if our online community had an official book or a book club, uh, the first book we would look at is Effective Data Storytelling by Brent Dykes. We really enjoy his principles.

We talk about them a lot. And, um, inside that group is people who are working with data and trying to build the skills above and beyond, uh, just the analytics and make sure that they are influential and valuable to their business. We have weekly session to get access to Kate and I. Um, I think if you’re listening to this podcast, if you found value in what Brent was talking about today, it’s worth you having a look.

I’ll put a link down in the show notes for the Influential Analyst Academy.

Kate: Let’s, let’s go into dashboards then, because we see this with just about every client. We’ll say, I don’t understand why we can’t get, you know, things happening because we’ve got all these dashboards and everyone’s creating thousands of dashboards. Um, I’ve got like this image in my head now, a marathon where you’ve got like the pack, you know, like there’s that pack of runners and there’s only a few out the front, but everyone’s creating dashboards.

And so then, um, we get with the team, you know, they give us their team to train and, um, what they all want is they just need to know how do they make their dashboard tell a story? And I know that we agree that a dashboard in itself does not tell a story. Um, and so we, we will, we will make some suggestions, like, you know, if if you want it to be more action driven, then you’ve gotta put some sort of, you know, maybe insight in your title rather than just describing the data that’s coming, um, from that dashboard.

And then the pushback is like, oh, but you know, it’s live data. It’s constantly updating. We can’t, we can’t put an insight in the title because it might not be the same tomorrow. So there’s this like, dilemma between making dashboards more meaningful and also wanting it to be live data. Um, so how do, how do we reconcile that?

And I guess I’m probably asking for advice here, like, how do you, how do you make dashboards more useful to people making decisions?

Brent Dykes: Um, so few thoughts on that. One is like, so I did a blog post about the four food groups of, of data communication. So in, in the, I don’t know if this is like, I, I, I was actually grew up part of my life in New Zealand, so I, my, my dad’s, my dad’s a Kiwi. My mom’s Australian. I grew up in Canada.

Thomas: The Holy Trinity right there.

Brent Dykes: yeah, yeah. But, um, one of the, one of the things, um, when growing up in Canada, in North America, we have this concept of like the, you know, the, the, the, the, the, what do you call it? The food plate. You know, you have your, your grains and your vegetables and your protein, you know, and so you gotta eat from dairy, you know, and those, the different groups, well, I did the same with data communication.

I said, okay, one food group is monitoring, right? So monitoring the performance of the business are, you know, is there something broken? Is something up or down, you know, just monitoring how things are going, that’s perfect for a dashboard, you know, and, and that’s really zeroing in on uh, what are those important KPIs that we need to focus on those key metrics, uh, you know, and just monitor how things are going. So monitoring is one type of communication and dashboards are really good. Now, there’s a another type of dashboard that’s an exploratory dashboard. it might be where the analytics team, you know, wants to provide a, a business team a, a more exploratory version that they can kind of go in different path filter, drill down, you know, and, and answer business questions for themselves, maybe on those ad hoc questions that they occasionally would come to the business team or to the analytics team for now, they could give them, okay, well here’s a bunch of data sets and, and you now have a tool where you can kind of go in and explore the data. So exploration is another great tool for dashboards. One of the problems though, is sometimes what I see is those get merged together. So they try to do monitoring with exploration,

Kate: Yep.

Brent Dykes: it, it, and then it doesn’t serve either. It’s too complex for the people who only want to use it for monitoring and then for the exploration, they’ve got a bunch of other, you know, in there that’s not really, doesn’t really help them from an exploration perspective.

So, uh, the third one is reporting, right? And so reporting are those qbr, the, usually in PowerPoint, they’re usually more of like a, a monthly quarterly cadence. but again, it’s, it’s, you know, a bunch of reports, tables, information packaged up, standardised, reported, you know, on a, whatever cadence they’re assigned to do.

And, and, and they’re, it’s in the family, but it’s, it’s not that daily kind of cadence of monitoring where things are at. And then the fourth one is explanation, right? So explaining the data, that’s the storytelling angle. Um, so those are the, you know, that’s one challenge that I see with dashboards is they kind of, you can have different types of dashboards that serve different data communication needs. And I think I’m, I’ve tried to shed light on this, that. Um, I think people are misusing dashboards in different ways. And, and, and when it comes to storytelling, you know, we also hear that now some of the ways in which I’ve recommended, um, to different clients. One really good way that I saw that I, I really liked was, um, with clients when I was working for Domo, we would often build the dashboards and we, instead of having like a chart title in each of the reports in the dashboard, we would have a question. So we would ask the question we want people to get from that report. And, and the good thing about a question is, it, you know, how, who are the top performers in our, in, in the sales region? You know? And so they go into the report looking, oh, okay, so here are the, you know, obviously they can use it different ways, but, but giving them the main way to look at the data can be very helpful. and then the other thing is, is that I found that there’s traditional reporting and then there’s storytelling. And then in the middle I have something I call narrative reporting. And so narrative reports are where you add in interpretation and context on top of, of a, a standard report. So a standard report would go out and you might have four sales regions, so north, south, east, west, and it would provide the data for each of those traditional report, you know, all of the metrics, blah, you know, once a month. Well, a narrative report would then take that same report and it would layer in interpretation of some of the key anomalies that you see giving context. So, why did this spike up? Or why did this dip? Oh, we looked into it. Now is it a de, is it telling the whole story? No, you know, and it could spark a story.

Could somebody could say, well, well, why is the south region underperforming on this one metric? Can you tell me why? Now we have an invitation to then go do a deeper analysis, we find something, and then we can tell the story around that deeper analysis, that deeper dive. that’s kind of my, I don’t feel like a lot of traditional reporting does a, a good job at storytelling.

And that includes dashboards. So a step up from that would be to go to narrative reporting. Now, the problem with dashboards is they’re constantly refreshing. They’re constantly updating. Again, that’s, that’s another problem with storytelling. Storytelling requires snapshots of the data, right? Like, what happened in this moment? And, and, and so, you know, if it’s, if it’s like a CCTV camera that’s constantly, you know, a video, you know, the video’s constantly going through and recording, know, it’s hard to say, oh, wait, stop it there, you know, and, and that’s what you need to do with this story. You take a snapshot, take a picture, you know, you take a couple other pictures of what’s going on at the same time, and then you, you craft that into a story. Um, you know, a con, a live stream. It’s really hard to, you know, and that’s what a dashboard often is. It’s a live stream of your business performance. Uh, and, and it’s very hard to tell stories with dashboards and yeah. So I’m, I’m aligned with you guys. I don’t feel data,

Kate: Yeah.

Brent Dykes: I try to educate people that data, you can’t tell data stories with dashboards despite what vendors have posi, you know, and that’s, that’s part of the problem.

I feel a lot of vendors have latched onto the term data storytelling. And, know, I’m a marketer at heart, you know, I know, I know how it is. When you have a popular term, you wanna, you wanna, you know, make, make sure your product, you know, associated with that. And so a lot of, uh, dashboard and bi vendors have latched onto data storytelling without really understanding what the term means and why product does not deliver data storytelling.

Kate: Huh. I actually wrote a blog post that was titled, A Dashboard is Not a Story despite what the Tableau Marketing Team tells you. That’s Yeah, exactly that. Um, yeah. Oh, that is so helpful. And again, um, I, I, I come back to it like, one of the reasons I’m, I, you know, I, I like your work is because you articulate something that I like, have muddled in my head.

Um, like, just as you were saying that, I was like, oh, of course, these categories of dashboards make a lot of sense that that is, yeah. Really helpful.

Thomas: I was, I was gonna say exactly the same thing. It’s it, we are speaking the same language here, but I think, uh, a bunch of it is certainly out of the book. And I guess I wanna, I wanna, um, move into like is there something that you believed or thought about data storytelling, Brent, back when you wrote effective data storytelling that you’ve maybe since changed your mind about, or has just become outdated?

Brent Dykes: Um, I mean, one of the things that comes to mind is like, so I’ve, I’ve always emphasised the three pillars of data storytelling, so data, narrative, and visuals. I would say I, I’ve, I’ve, I’ve always kind of been more like, Hey guys, we’ve gotta make sure we’re balanced here. Not just visualisation, but data and narrative.

And I would say one of the things that, that’s really, you know, I, I had somebody come up to me after a presentation say, what about like audio stories? And, you know, like podcasts, right? If you think about a podcast, sometimes they can tell a very rich data story no visuals. Why is that? Why is that?

Well, it’s because data and narrative are really the two key pillars, right? I, I often talk about data being the foundation of a data story, right? If you, if you don’t have data, you’re just telling a fictional story, you’re just making up stuff, right? So it’s, it’s not evidence-based. And so that’s a core, core element of a data story. Then the narrative, you know, like how you’re actually stitching it all together and weaving it all together into, into this compelling arc that our brains respond to. and, and even from a sense-making perspective, you know, like what is the brain, the human brain making sense of the world around it.

And so when we package up something into a narrative, the, the, the brain is just like, oh, thank you. You know, like we’ve done a lot of the heavy lifting for the brain. So it’s just like, you can just go, oh, thank you. I’ll just plug that in. That’s great that you’ve connected all the dots for me. You’ve helped me to kind of quickly absorb that information and it’s in a, in a memorable and engaging package. And so I would say, I think the one thing I, I’ve kind of come to the realisation is that the visuals and, and I still say that visuals are important because often data is very complex and it is hard to kind of like fathom, right? And so when I talk about a podcast, having, you know, they’re usually dropping data points, you know, they’re kind of saying, but we’re not seeing the trends.

We’re not seeing the patterns, we’re not seeing the anomaly. You know, they talk about it. Um, and, and so visuals, I see that, see visuals as serving the needs of the data and narrative. enable the data and narrative to tell the story. And so they’re, you know, I kind of see visuals as more subservient than maybe I did when I was writing the book, whereas I saw them as, as an equal partner. And, and again, I don’t want to diminish the importance of visualisation. ’cause often data is very complex and, and sometimes we can only see things by visualising the data. So I, I don’t wanna, again, I don’t wanna diminish it, but I think I have changed my position a little bit more open-minded, you know, but I’m also now very, adamant that the data and narrative are, know, it’s imperative that we understand the role of those two pillars.

And, and obviously the third pillar, visuals are there to support them.

Thomas: It’s a bit funny, isn’t it? Like, I dunno if this is the same in your world, but every now and then we get the comment of like, just help us know what the magical correct chart choice is to make this make

Brent Dykes: Hmm.

Thomas: sense. And it’s like we

Brent Dykes: Right.

Thomas: are, we are on step like 10 here. We need to go back to like, what are you trying to do?

Who’s the audience? What’s the message? What’s the, what’s the insight

Brent Dykes: Yeah.

Thomas: All of that. Um, yeah, so it’s, it’s a bit funny, isn’t it? We’re here on a data storytelling episode of a podcast and sort of saying maybe the visuals aren’t that important, which kind of echoes what, what we believe in and teach. Yeah. interesting. Interesting.

Brent Dykes: mean, I, I often compare, think, I think what people think they’re, they’re kinda looking for that silver bullet chart, right? What’s that silver bullet? And it’s like, it can, it can change by many factors. Like what are you trying to communicate the, the features of the data, right? If you have a very, let’s say it’s very skewed in one side, maybe, maybe that chart that you would initially choose doesn’t work.

’cause it’s so skewed to one side, you know? So it’s, it’s, it, it’s a, it’s a, it’s a combination of different factors. And so I often say, when I’m teaching about this, I’ll say, it’s kind of like, imagine you’re moving into a, a new house, right? And you’ve already, in your mind, you’ve already envisioned how you’re gonna lay out the couch.

Let’s take the living room, right? You’re gonna put the couch here, you are gonna put the TV there, you are gonna, the bookcase will go there and the rug will go here, right? And then you move into that room. First, first day you’re in that room to enjoy your new living room. then all of a sudden you’ve noticed that from the window, it’s now glaring onto the TV screen. like, it’s not, the TV can’t go here. ’cause in the, you know, the, the glare of the sun on that screen, I can’t, uh, you know, I want to, maybe I want to keep the windows open, or I don’t have, you know, like I can’t block the sun or whatever. So then I decide, you know what, the TV’s gonna go over here, means then the couch has to go over here.

The rug has to, you know, the book, everything changes. So I think sometimes it’s a matter of, when you go into the data, you might have a good idea of how you wanna visualise it, but you have to keep an open mind that maybe first version of that doesn’t visualise the data as well as it could. And then you, and then, and then it’s about knowing what other, the, what are the other options that you have to visualise the information. And you might need to iterate a little bit until you find, oh, now, now I’ve got it, you know, this, you know, I was gonna use group chart or group bars, but, but a dumbbell chart works much better at telling, you know, showing this pattern or this, this, you know, part of the data I wanna feature. And, you know, or slope graph or something, whatever.

You know, they, there’s so many different, what I, what I find is there’s certain, like, families, right? Where you run into a situation where you’re comparing things or you’re, you know, and then it’s just knowing what those other options are. ’cause sometimes the first choice you have is not the one that’s, that does it, you know, it doesn’t click, it doesn’t, doesn’t communicate as, as well as maybe one of these others does.

Thomas: So much.

Brent Dykes: yeah.

Kate: Part science, part art,

Brent Dykes: Yeah,

Kate: like interior design. Yeah. Great analogy.

Brent Dykes: Yeah.

Kate: All right. So Brent, um, you have influenced Yeah. How we communicate and how, um, we teach. So I would love to know what is up the chain from that? What is a book or a resource or something that has influenced the way that you communicate?

Brent Dykes: Um, yeah, I, I think one book, I mean, I’ve tried to, I, I need to do more reading than I have. I’ve kind of slowed down on the reading. I’ve been distracted, but, but one of the books that I, I really, I, I like going outside of the world of visualisation, outside of the world of, data. And one book that I would recommend to people is, um, a book by Gary Klein. Um, so he’s a, he’s an author. He’s somebody I quote in my book. He gave me a really good definition of what an insight is, and he’s written a lot of book. He’s a psychologist and he writes about, um, insights and, and the, the book that I would start with is, uh, Seeing What Others Don’t. And so that’s, uh, he goes into different of insights and where they come from.

And I just found, you know, and I believe that data storytelling is really based on sharing insights and, and his definition for those of you who haven’t heard what his definition is. So when I was writing my book, I had gotten feedback from some of the reviewers saying, you mentioned the word insight in your book many times, but you haven’t really defined it.

And so I was like, oh, okay. Yeah, that’s great feedback. Thank you. Went to chapter one and then I was like, starting to type out, okay, wait a second. What is an insight? You know, and I struggled to really come up with a really good way to kind of articulate what an insight was or is. um, and then I started looking at dictionary definitions and, and they were kind of like close, but just not quite as strong as I, as I really wanted or needed. then somebody pointed me to Gary Klein’s definition, and he defines it as an unexpected shift in the way we understand something. And for me, that was such a concise little definition and really reflected what I felt is an insight. So we have some kind of understanding of the world. We have some assumptions.

We, let’s take it from a marketing perspective. believe our customers buy our product ’cause we are the low cost leader. And then all of a sudden we do, we get some feedback from a survey showing that no price isn’t the main reason why they buy our product, it’s because of our integrations. You know, and then now we have this insight where we, we thought that they bought our product ’cause of pricing, ’cause of our low price. And, and our marketing went that way. All of our engineering on our product went that way to kind of keep our product low cost. So now that changes our marketing strategy.

That changes where we invest in terms like, rather than just sticking to a minimal level of integration, maybe we expand it even more. Let’s play to our strength so that having that one shift in our, you know, unexpected shift in the way we understand something huge. And, and I feel like a lot of data stories are about sharing those insights, you know, and, and, and maybe there’s, you know, it can be a surprise and, and maybe there’s even resistance to that new paradigm, you know?

And that’s often the case and that’s often where I say I, I can’t guarantee that a data story will overcome every, you know, part of resistance. But you’re sure as heck not gonna get, you’re not gonna punch through that resistance without a data story if, if, you know, if, if you, um, want to. So anyways, that’s one example of a book that I, I found very, um, influential.

Thomas: Look, Seeing What Others Don’t. I think that’s a new one. It’s not on either of our bookshelves. So good recommendation. I, uh, I have a quick one, which is if you could just ban one piece of presentation, communication data, storytelling advice tomorrow from the universe, what would it be?

Brent Dykes: Keep your slides to X number of slides.

Thomas: Yep.

Brent Dykes: So I, the reason behind that is tell me how many slides I need to have. Just tell me how much time I have, let me control how many slides I need to have. ’cause otherwise, what happens is, I see this when I work with clients, is they have these kind of restrictions on no more than 10 slides, or no more than 20 slides.

And what happens is they take their 60 slides, contract it down into those 10 slides, and now you’ve got this horrible mess of information that, rather than conveying the information, just saying you’ve got, you know, you’ve gotta present in 10 minutes, rather than saying, you know, you’ve got 10 slides.

Um, just tell me what I, how much time I have. And, and, and I might have 20 slides to present in those 10 minutes, but

Thomas: Yeah.

Brent Dykes: I’m doing it with the best interest of communicating effectively. So that, that’s my, and I, I’ve, I don’t run into that all the time, but I have had some, you know, when I’ve speak, spoken at conferences and stuff, and then people, I use a lot of like quick transitions and things.

You know, I try to break it down into chunks, right? And, and so I share little chunks at a time, it’s very kind of quick. and then people look at the length of my decks and then they say, oh, you gotta, you gotta cut this in half. And I’m like, but you told me the time.

This works for that time. It’s not, yeah. Anyways, that’s my, that’s my one pet peeve.

Thomas: Uh, no, no. Uh, the, that, that sounds like it came out of my mouth. Yeah. Crunching 60 slides into 10.

Kate: Do you feel vindicated?

Thomas: yep. Yep. Brent said. So, uh, look, if people like, uh, the way you think, Brent, where can they find you?

Brent Dykes: Uh, you can find me on LinkedIn. Uh, you can also find me at, um, effective data storytelling.com. Uh, and, uh, yeah, those are probably the two main areas, um, to find me. And then if you like what I talk about, sign up for my newsletter. Um, yeah, those are the ways, the best ways to follow me and then buy my book, obviously, if you like.

If you wanna, if you haven’t read my book, check out my book.

Kate: Yes, yes, that is, um, always the one that I recommend when I’m on any, on a podcast or anything as well. All right, amazing. This has been a really fantastic conversation. I feel like I’ve learned, I feel a little bit vindicated with some of the things you’ve said. Um, you know, it’s, it’s the best combination.

So, Brent, thank you so much for coming and chatting to us, um, from Utah, the other side of the world, from us.

Brent Dykes: Yeah. Well thanks Kate and Thomas and great to, great to chat and love talking about data storytelling. So yeah, thank you.

Thomas: Thank you for being with us here today. Head to presentation boss.com.au/podcast where I have put all the links, uh, down there from this show.

Kate: And of course, if you have anyone else, if you know someone that you would love to hear from on this show, please email us podcast@presentationboss.com.au.

Thomas: I’m sure there’s some excellent people out there. And if you found value in today’s episode and what Brent had to talk about, do subscribe to the Presentation Boss podcast here in your player of choice and recommend this episode to a friend. Have a great week.​