[Radar Recap] The Art of Data Storytelling: Driving Impact with Analytics with Brent Dykes, Lea Pica and Andy Cotgreave

r audience. So this is an essential thing to learn about for your career success. And we've got three illustrious speakers to teach you their storytelling secrets. First up, Brent Dykes is the chief data storyteller at Analytics Hero. He's also the author of Effective Data Storytelling. We also have Leah Pika, who runs Story Driven Data. She's the author of Present Beyond Measure, as well as hosting the podcast of the same name. Our very nice plug there. Good timing. It's convenient to have a copy there. I dunno what just happened. That's so weird.

(02:29):

And thirdly, we've got Andy Reve. He's a senior data evangelist at Tableau, and he's also the author of the Big Book of Dashboards. I can do it too, Brent, if you've not got a copy of your book. Excellent. Yeah, so you've got some reading to audience. Alright, so all three of our speakers have got decades of experience training companies in data communication skills, and I'm pretty keen to hear all their tips on how to make use of data storytelling. I hope you are too. Okay. So let's hear what they have to say now to begin with might wonder why do you need to create a story for your data instead of just letting the data speak for itself? So Andy, do you want to take this one first?

Andy Cotgreave (03:19):

Yeah, well, I think the fundamental thing is data doesn't speak for itself, right? It's just like a neutral atom of information. And then the second you aggregate, it becomes a subjective decision you have made when you aggregate it. Even if I average somebody asked, am I a city or a United fan? Well, I'm a Manchester City fan. For those that don't like that, I'm sorry, but that's my truth. If I was to average the wages of everybody who worked at Manchester City, I can tell three incredibly different stories. If I choose a median, a mode, or a mean average, and that's at the most fundamental level of data, doesn't speak for itself, I get to make that choice. They're all valid choices, but they're completely different stories. So that's the power we all have as data analysts is the second we take those atoms of information we are curating and editing the insights we found to convey information to drive decisions. So that's my nutshell answer.

Richie Cotton (04:25):

Okay. I like the idea that you're making choices about what you include in your analysis. Brent, do you have anything to add to that?

Brent Dykes (04:32):

Yeah, I would add to that saying that often when we look at our numbers, they speak to us. We look at a dashboard, we look at a report, we look at some information, it speaks to us, and we assume that it's going to speak to other people equally well because spent hours in the data, we've analyzed it, we've see the patterns in the things amidst the noise, and then when we share that out, we assume, oh, it's going to speak to them. It speaks to us, but it doesn't. And so that's really that process. We need to go from really putting it into an explanatory format so that it can really speak to others.

Richie Cotton (05:07):

I felt it's a very common communication problem where you think, oh, this is really, really obvious, and then nobody else gets what you're on about. So yeah, I like the idea that it's going to help communication.

Andy Cotgreave (05:19):

When I was first building dashboards, one of the most humiliating but important things left, things I learned is to sit down with your target users and get them to use the dashboard, but sit behind them and don't tell them how to use it. And it's a very humbling experience because as Brad says, you think you've designed something perfect and then you give it to real people and they just have no, oh, I'm like, why did you not use the filters I put on this dashboard? And they're like, what filters? I was like, what? I'm like, they're there, but people don't see them. Right? So yeah, do those tests.

Richie Cotton (05:56):

Absolutely. That's great advice there. Leah, do you have anything to add on why storytelling's important?

Lea Pica (06:03):

Well, stories are important because we're human beings. So they show research, especially the work of Dan and Chip Heath In Made To Stick. He shows the what actually happens when you share information via just numbers and figures versus tying it to a story. And the results don't speak for themselves. The story speaks for itself, where the story data is always remembered for longer and more comprehensively than just the figures. And what I love about that story is you'll probably remember that now better because I told that in a story. So we just made it really meta.

Richie Cotton (06:47):

Yeah, I like that idea. Yeah, it is going to help you help the audience remember things, which is probably what you want if you want them to. Excellent. Alright, so hopefully everyone's persuaded that you actually need to make use of data story, Ellen, but I'm sometimes not sure. Okay. I've done some analysis on a sales pipeline or something, and then how am I going to turn that into a murder mystery novel? So it's not obvious. Can you tell me what sort of a plot a data story might have? Let's make this the order a bit. Brent, do you want to go first this time?

Brent Dykes (07:25):

Yeah, so I mean there's a specific approach that we take with a story. There's this narrative arc. You may have heard of freight text pyramid, or you may have heard of stories, have a beginning, middle, and end. Basically there's this arc that we're building up to, in my book I talk about it, we start with a setting. You have some kind of hook that gets people interested in what you have to say. So that might be where, Hey, I noticed that there was this metric that decreased or spiked up, and that would capture the interest of your audience. So that would be that hook. And then you basically build up to your climax of your data story where you have, in my book, I talk about it being an aha moment. That's your main insight that you're sharing with the audience, and hopefully it's going to mean something to them.

(08:14):

And then we have what I call rising points. We basically share bits and pieces, the critical pieces of our analysis that really unpack what's going on with this problem or opportunity that we've identified, and then building up to that aha moment. And then we're not done at that point. Once we get to the climax of the story, we don't end on the climax. We actually then say, okay, what are we going to do about this? What are the actions? What are the recommendations? What are the options that this audience can consider? And so really having that story arc, that adds tension, that adds there's interest, there's intrigue. We get all of that rich benefit from telling a story that perhaps a reporting structure wouldn't really necessarily generate the same intrigue or interest.

Richie Cotton (09:03):

I think the first point about how you've got to have a hook at the start to get people to care is going to be really important. Because actually if you don't do that, then the rest is going to be ignored. The other point about having a data climax somewhere near the end seems interesting. I'm wondering what that would look like. I dunno, Andy or Leah, what's going to be this key moment? Leah, do you want to go?

Lea Pica (09:26):

I'd love to do this from a musical theater background, and I'm obsessed with movies, so this is my advice. You make the villain of the story dastardly. This is what I feel is missing from a lot of our data presentations where we don't actually blow it up as much. What is the problem that your customers are facing? How can you show the data that is going to help them? So make the villain dastardly that climax looks like 99% of people that we pay for to land on this landing page are leaving. We're paying for 99% of people who leave. This is make it dastardly. And then put the downside of that, the stakes that you have to lose as a result. And then put it in terms that the audience actually understands and cares about. Are you losing money, market share, brand favorability, customer satisfaction, customer retention? What are you losing? And blow it up. People don't care if you go in and you're like, yeah, our P value of ES was not, no,

Richie Cotton (10:41):

It's a really harsh impression.

Andy Cotgreave (10:44):

That's my business reality. Leah,

Lea Pica (10:48):

Would you watch a movie where people are like, oh no one's not that bad. He's an okay guy. No. So let's bring that to how can we bring that to the actual challenge that our customers or our stakeholders are facing? Thank you. Just get that out.

Richie Cotton (11:10):

That felt therapeutic. I like the idea of blowing stuff up. It

Lea Pica (11:12):

Was cathartic. It was

Richie Cotton (11:16):

Alright. So Andy, it seems like you took a little bit of issue with the impression there. So it seems like you have to tailor your story to different audiences. Do you want to talk us through how you might do that?

Andy Cotgreave (11:26):

Yeah, just to add to what Leah and Brent have said, I'm just going to paste a link to a YouTube video from Kurt Ger who's a fantastic author about the story arcs of data. It is, just think about that arc. There are seven different story types. Think about what stories that exist in our storytelling. Somebody at Ben Jones who runs data literacy.com, he did a great talk a few years ago about applying the those seven story narratives to data. So thinking about that narrative is really important. The thing about tailoring though is not everything has to be a data story. Sometimes I find in the dashboarding world, people think, well, the dashboard has to tell a story. So you have to, and it's like, well, does it, or is that about story finding rather than storytelling? So first we need to make those distinctions. And then in terms of different audiences, I mean the thing is, I think in a presentation you have to know which audience you are targeting and design for those people. And if you have 10 audiences in 10 different types of people in one audience, it can become a challenge, which is probably beyond the scope of the ability I've got to answer right here. But yeah.

Richie Cotton (12:43):

Alright. So I like that idea of distinguishing between gear upfront, exploratory data analysis where you're trying to just see what's going on in the data and the explanatory stuff that happens later on when you tell your story. Okay, Brent, do you want to talk about different tailoring things for different audiences?

Brent Dykes (13:01):

Yeah, yeah. One of the examples that I use in my book is I talk about Robinhood. And so if you think of all the different ways that Hollywood has shared that story of Robinhood, there's different ways in which they've done it. So you have the animated version with Disney, you have Kevin Costner as Robinhood with his mullet, you have Russell Crow, you have Errol Flynn with a musical. You have all these different kinds of men and tights, you have all these different versions of the same story, but they're told to different audiences. So it's very, one of the things I talk about is that even before we get into the narrative structure and all that, who is the audience? It's going to change the way we deliver the story. Because if you're addressing a technical audience, that's going to be different than a business audience.

(13:49):

If you're sharing your story with an audience that is very data literate and data savvy, that's going to be different than maybe an audience that isn't S data literate. And then the other example would be what if you're presenting to the team that implemented the process that you are now nitpicking on, that's going to be a different approach than maybe an audience that has nothing to do with implementing or setting up or devising that process. So you can be much more open and candid with them. So all of those adjustments will impact your story and how you tell your story. So definitely you always need to think of your audience first before you build your data story. And then obviously we've talked a little bit about the structure and stuff that's going to come later, but if you don't understand your audience, it's going to be hard to tell them a story that's going to really resonate with them.

(14:39):

And sometimes you have mixed audiences, that's a real challenge. You have people who maybe come from the technical side of the business side, maybe have executives, stakeholders who are the decision makers, and then we have other influencers and different things. So in some cases, you might have to prioritize who you're going to really target. You might have to sacrifice a couple of the audiences in the room because it's like, okay, I can't tell 10 different versions at the same time. To all everybody in the audience, I have to maybe say, you know what? The CMO or the CTO are the priority and I need to make sure that she gets it and that this resonates with her. Maybe some of the other people in the room, I'm going to have to follow up with them and give them maybe a different version that helps explain it a little bit better for them. But those are some of the choices, the hard choices that we have to make as data storytellers.

Richie Cotton (15:34):

I find the idea of sacrificing audience members strangely appealing. We'll bring that back to a future radar. But no, seriously, it just seemed important to think about who are the key decision makers or the people that really need to understand your story in a room if you've got lots of people there. Alright, so let's move on to talking about presenting data. This is a common place where you have to do storytelling. So first of all, what are the most common mistakes people make when they present data? Andy, do you want your request on this?

Andy Cotgreave (16:10):

Yeah. I was sat in a meeting about seven years ago. It was 2018, our CMO was presenting some data to an audience. And finally after 20 years of career, I was like, this is awful. She's saying, as you can see, blah, blah, blah, this, that and the other. And I'm sitting there going, nobody knows what she's talking about. The data's sort of on the screen, but I don't know. And I'm like, and I've been in meetings like this my entire career. So instead of listening to that meeting, I then wrote down all the things that people do wrong when they're presenting data and how to fix it. And that should turn into a book. But yeah, Lee has been leading on this for years as well. Her is brilliant, I think. Right? I'm going to go three mistakes. Three mistakes, yes. First of all, if you copy and paste a dashboard which has been designed for exploration on a small screen onto a PowerPoint slide and say, as you can see, blah, blah, blah, that's a failure.

(17:12):

The dashboard is designed for a small screen, that information not designed for a big screen, don't do that. Second mistake is people think, oh, I'm presenting on a 200 inch laptop or a keynote screen. So it's a big screen. I don't have to adjust anything, a big screen. But if I'm sat in the 10th row of a meeting room or in a keynote room or even on a virtual meeting, actually that screen, the real estate in front of my field of vision is smaller than the real estate of my own laptop, my own cell phone. So it's not a big screen. So you have to adjust fonts accordingly. You can feel my passion, Isaac here. This is something that drives nuts.

Richie Cotton (17:49):

And as I get deeper into middle age, my eyes get tired. Yeah, I'm definitely with you there. True.

Andy Cotgreave (17:55):

When I was 18, when I was 25 or 23 in my first job, somebody trained me in presentations and they were about my age 53. And they said, the smallest font on your screen should be the age, should be the size of the age of the oldest person in the audience. So the smallest box.

(18:13):

And I was like, I'm 23, I don't care now I'm 53. I'm like, oh yeah, he was totally right. And then the final thing is pointing. If you stand at a lectern and point like this and go, as you can see this, that and the other, the only person who knows what you are pointing at is you have to physically bring attention with PowerPoint like arrows or good clickers to what you're pointing at. Otherwise, everyone in your audience is just seeing you wave a magic wand without knowing what they're going at. I'll stop there and just take a moment to relax, but please be better at presenting data. Everybody,

Richie Cotton (18:46):

I feel like. Thank you. Those brilliant examples. I'm sure Leah and Brent, you must have examples of terrible presentations as well. Leah, do you want to talk us through an example?

Lea Pica (18:56):

I do. I just want to name that this feels very therapeutic for us. Yes.

Andy Cotgreave (19:01):

I just want to thank you. I don't feel like I'm getting a closure to these issues. Some people change.

Lea Pica (19:06):

Really? That's good. Yeah, hopefully they're not watching. So a second, all of that. One of the mistakes that I want to name is technical people presenting very technical details to non-technical people. So oftentimes what I find is we are trying to show to demonstrate just how hard our work was, how hard we worked on it, how long it took us, all the details, all the CP value comment earlier, we'll go back to that one. And we just sort of talk over our audience. What they're trying to do is they don't care. They don't care about any of that. What they care about is did your analysis lead to something they can take action on and why should they act on it? Because what is at stake if they don't take action? So for me, present all of your technical details to your boss. Have a dance party. Celebrate how hard you worked, but keep what matters to the stakeholders. Keep that for them and really pretend that you're trying to explain this to a 12-year-old because that might be their level of savvy in terms of the kind of information you're working with.

Richie Cotton (20:27):

I like that even you might go, okay, I'm speaking to someone who's a grown, but let's go down bit just to make sure they actually have, can actually have something to take away. And yeah, you're on the point that you want some kind of action to happen afterwards. So there's got to be a call to action within your presentation that seems, Brent, do you have any examples of presentation mistakes? What do you need to get off your chest?

Brent Dykes (20:54):

What I need to get off my chest when I speak at a conference? So you have your dedicated breakout session you'll speak at, and then usually I'll poke my head into other breakout sessions and then I'll see just these very complex diagrams of data pipelines and how all this stuff, and they're data dumps, they're really, you're basically dumping all of the information all at once. And it speaks to what we said we talked about at the beginning, it's that curse of knowledge. The people that are presenting, they totally get it. They know what they know, but it's hard to not know what. And so that's the challenge that a lot of times we run into as presenters, we don't take the time to say, wait a second, I'm obviously biased. I already know this problem. I've spent hours, days, weeks, analyzing this. It's really that step to say, what is going to really communicate strongly to people that haven't spent as much time as me in the data, don't know it as well.

(21:58):

And so I think a key step in that that often is neglected or overlooked is maybe taking time and saying, Hey, I'm going to be presenting this at an upcoming meeting or conference or whatever, and get a perspective from somebody who maybe isn't in the data like you are, and see if you're able to communicate your points clearly to them. I think if many people just did that, just got feedback from somebody before they presented and then they would realize, oh, oh wow, I didn't realize that you didn't know this assumption or I didn't know that this wasn't coming through. Clearly that curse of knowledge really trips us up. That's a big problem.

Richie Cotton (22:44):

Yeah, definitely seen a lot of slides where I'm like, yeah, I'm not going to be able to understand this in any time soon because of course in a presentation people got to understand it in more or less real time they, they're not going to spend 20 minutes thinking about stuff. Yeah,

Andy Cotgreave (22:57):

Rick, I think the thing, when I'm talking to people about this, I'm like, why are you making this presentation? You are making the presentation to try and share insights to drive change. So if you actually care about doing that, then make it work. Don't just stand up and blah blah, blah, blah. Leah, fantastic, fantastic impression of every, most meetings, blah, blah, blah it before, I don't care. Just don't tell me, don't bother. If it's not that important, just give me five minutes back. If it is important, bring the energy in it that you need to make that change. And it's hard. We've got a question about how to practice and get better at this and it takes practice and it takes dedication and it takes repeating and thinking about what you're going to talk about.

Richie Cotton (23:46):

Absolutely. Alright, so we talked about presentations. The other case where you're going to want to use a story is in a report. So can you talk me through how you might change a report in order to include a story? Does the content need to change? Does the structure need to change? What would you do differently to include a story who's not talked in a while? Brent, do you want to go?

Brent Dykes (24:12):

Yeah, no, I'm super passionate about this, so I'm glad you gave me first dibs on this question. So one of the things I talk about is if we take a standard report, and I'll just give you a basic example. You've got four sales regions, so you've got north, south, east, west. So how would you structure your report? Well, I'd have a slide on the north. I'd have a slide on the south. I'd have a slide on the east and the west. That's the reporting structure. We are kind of method methodical and we basically are systematic. We kind of go through a set of things in an order and we're not going to skip over one of the regions because trying to be comprehensive, we're trying to cover everything in a data story. It's completely different. What we're doing is we're basically saying, Hey, nothing's weird in the north region, nothing's weird in the south of the east region.

(25:02):

There's something going on in the west region, so I'm not going to talk about those other regions. I'm just going to dig into what's the problem in the west region and I'm going to show you that we're under training our sales team, or basically they're skipping on the trainings and that's why the west region is underperforming. And then we go through telling a story. So the reporting structure is very different. It's kind of very organized and structured in a meaningful way, very comprehensive. Whereas a data story is really diving into a specific opportunity or problem and really explaining it. Now there's kind of a step in between those where I call it narrative reporting, where we're taking that standard report that we have north, south, east, west, and we're adding in a little bit of context. We're adding in some interpretation of maybe some key observations that we have, but still it's still reporting, right?

(26:00):

Because the structure hasn't changed. We're just adding in layering in the observations and the context and different things. But that story, the structure of a story is going to be that narrative arc we're going to take people through. And so the simple way of thinking about this is report structure is really about what the narrative report that new in the middle is kind of what plus, right? Because we're adding the context and the interpretation. And then if we go over to the data story, that's the why plus and why do I add the plus to the why while we're digging into what's causing the problem in the west, but we're also offering solutions. We're offering options to help fix that problem that we're seeing in the west region. So that's kind of how I look at how we can make reports more like stories by doing narrative reporting. But at the end of the day, unless we're willing to fundamentally change the underlying structure, reports can never become stories if they don't change their structure.

Richie Cotton (27:00):

Yeah, it just seemed like the order of the content is going to really affect how people read it and how people listen to us. It's like scientific papers where everyone sort of reads the abstract and then they jump to the conclusion. I have to read it out of order, not written the way most people want to read it. Okay. Do either of you do have anything to add on how you might change the structure of reports? I felt fairly comprehensive, but I dunno whether you want to say something

Lea Pica (27:28):

I cannot. That is a tough act to follow. I cannot improve on that answer. Alright.

Richie Cotton (27:34):

Okay. In that case,

Andy Cotgreave (27:35):

I'll just add one thing, just one quick tip. If you look at the background behind Brent, he's got covers from Marvel comics. Marvel comics are grid-based. All comics are grid-based narratives. And certainly in the western culture we read from top left to bottom, that is a really an applicable thing. You can turn specifically when you're making a dashboardy report, it's like what is your flow? So go from overview and create that flow there. Comic strips, there you go. I've just given you all an excuse to go and read some comics and use those as inspiration. You used them as inspiration for a lot of dashboards. They're very helpful.

Brent Dykes (28:18):

So Andy, I'm actually going to be publishing a blog post on what lessons can we learn from comic books and apply it to data storytelling. So just giving everybody a heads up that's coming

Lea Pica (28:31):

Collab.

Andy Cotgreave (28:32):

Yeah, let me know if you want to proofread or some thoughts because we've got many.

Brent Dykes (28:37):

Yeah. Awesome.

Richie Cotton (28:39):

Alright, we're going to go to audience questions in five minutes. So for everyone in the audience, please do write all your questions. There

Andy Cotgreave (28:46):

Are some great stories and comments on this feed. By the way, I'm reading

Lea Pica (28:49):

As many as I can.

Andy Cotgreave (28:51):

Yeah, really good stuff.

Richie Cotton (28:53):

Excellent. Alright, so we ought to talk about data visualization because having some images is going to make your story a bit more pretty, a bit more engaging. So who can talk me through great data visualization? What do you need to do in the context of data storytelling? Go on Leia.

Lea Pica (29:14):

Well, I think something important is to know the audience and then if your audience are not made of data visualization experts don't put Mary meccas and spider graphs and highly complex alluvials and graphs that they're going to spend their time trying to decipher and figure out how to read instead of just immediately intuiting and understanding what you are trying to explain to them. So the visualization you choose absolutely matters. So it's really worth getting to know the different kinds of data and messages that you're trying to convey with that data. And the real key is if you plop it in front of them and you make an observation about the data because you want to not let them figure out the story, you are there to tell them. The story is that if you make an observation about the data and they near instantly understand and come to the same conclusion, you've chosen an appropriate data visualization for your story. That's my rule of thumb.

Richie Cotton (30:25):

I like that. Yeah, no difficult statistical plots if the audience doesn't care about that sort of thing or doesn't have the Yeah,

Lea Pica (30:32):

No.

Richie Cotton (30:34):

Okay. Data VI feels very close to dashboarding. Andy, do you have an opinion on this?

Andy Cotgreave (30:41):

Yeah, I think generally a lot of them, whether it's storytelling or dashboarding generally the advice is generally the same. I find most people in their data viz career go through the similar path and Brett and Leah, if you did the same as me at the start, you don't really know what you're doing. So you just press the bar chart or pie chart button in Excel. Then you learn a little bit about a data visualization and how to use these tools. So everything you do is like radar charts and Merrimack goes and all the complex charts because you think you're really good. And then you go through the trough of disappointment in that nobody has a clue what you're looking at. And the wiser you get, the more and more you just draw bar charts because they might be boring, but they are effective. We are trying to convey information as simply as possible, and as I've done this for years, there is a great joy in crafting a bar chart that hits that goal that Larry just said, can I make a bar chart deliver the insight?

(31:36):

And one other example of that is if I want you to focus on one particular bar in that bar chart, that's the one L color, a dark blue, everything else will be gray, right? Simple things like that make beauty, but in simplicity. Oh yes, right. But don't take it. It is possible to tell a story very effectively. We were the very complex chart. This is a challenge. If you want to see the master of somebody doing this, go and watch hand rosling's Ted Talks. He told amazing stories with complicated animated scatterplots that have color, size, shape and everything. And yet he brought it to life because he recognized I have to explain what this is, I have to explain what this is, what this stock means, what this color means. Now we can understand complexity, so it is possible to do it all, but I think if anybody listening, if they're early in their career, really keep it simple, really keep it simple and build up with the skill over time.

Richie Cotton (32:38):

Absolutely.

Brent Dykes (32:39):

I put a link to Han Rosling's famous, one of his famous presentations in the chat.

Richie Cotton (32:47):

Yeah, Han Roly. He was definitely a master as well. Alright, before we wrap up, we have to talk about generative ai. And I suppose this can be a bit controversial. So I know I saw you on LinkedIn, you were recently posting about complaining about using generative AI to create presentations. Do you want to explain what you think the problem is and then I think we probably going to have an argument.

Lea Pica (33:09):

Sure. I want everyone to know I'm a generally agreeable person. I only complain on panels about things, so I just want to make that very clear. So generative AI is a tool kind of like PowerPoint that is best when we have a certain skillset and intention that we bring to it. And it's a little bit of the wild west. So one of the questions I get a lot from workshop participants is this all irrelevant? Is this entire skillset just going to go away because I can just go to AI and tell it to make me a presentation with a bunch of slides and do it for me. So I went, and AI is really an exercise in good communication and clarity. It's actually great for human interaction in some way. You have to be so explicit. But no matter what I did, no matter how much explicit direction I gave for creating slides with simplicity and gestalt principles and structure, it just kept creating these crazy conglomerations of 3D rounded spike pointed bars with rainbow gradients.

(34:26):

And my son and I had a blast trying to get it to do what we wanted and it just kept getting worse. And really the last straw for us was the final slide. They put the slide in a room with a bunch of fake people with thermostats on the wall and I'm like, I didn't ask for any of this. So where I think we need to know where it's going to be useful. So some of the places I am enjoying using AI for presenting, I use a framework called StoryBrand for a lot of marketing sort of storytelling. And it knows the StoryBrand framework and you can help train it on that a bit. But I like to kind of have it structure a presentation, give me an outline of what it would do with a certain objective that I have. I train it on a kind of audience avatar, I'm presenting to a typical C-level who I have to convince and what their challenges are, what they're looking for, and then kind of upload my outline into it and have it evaluate whether it feels compelling or not. So those are some of the ways I'm exploring, so know what to use for it and beware when it comes to the visual aspect. So Rand complete.

Richie Cotton (35:41):

Okay. Absolutely. Yeah, you've got to be very, very careful in terms of making sure that the output quality is decent. Andy, I'd be amazed if you're not doing anything with

Andy Cotgreave (35:51):

Ai. I mean, I know we need to go onto questions, so I don't think there's much I can add to what Leah has said. There are things it can do amazingly, but if you try and do exploratory data analysis or storytelling with generative ai, it is at the moment laughably bad for actual production level output and it's the antithesis of flow. One of the great things about using Tableau over 12 years is that I get in a flow state, I'm just playing with data and using my skills in the exact thing. With gen ai, you type something, you get anxiety because is it going to create usable output and then you get bored, it takes forever. So then you contact switch and go and look at TikTok and then that's what you're doing. You're going through anxiety to boredom endlessly. And it's like, well that's not how I want to do my day job. It can be brilliant, it's got a long way to go.

Richie Cotton (36:45):

Alright. Okay.

Brent Dykes (36:48):

I would just quickly add, I think really data storytelling is going to be augmented. When we look at data storytelling, the role that AI is going to play, it's going to augment our abilities and skills. I don't see data storytelling being automated, fully automated,

Andy Cotgreave (37:03):

No chance.

Richie Cotton (37:06):

Yeah. So that for everyone listening, you got to keep learning stuff still

Lea Pica (37:12):

Security.

Richie Cotton (37:13):

Yep, absolutely. Alright, so let's go to audience questions. We a little under 10 minutes to go. So this question, oh, so the top question is, will we have a certificate of attendance? Short answer. No, you don't. Alright, so better question comes from Barrett, say, what's the biggest pitfall people make when they present data to an audience that isn't familiar with your subject area expertise? How much should you patronize the audience on some of the terms? So I'm not sure about patronizing, but how much do you need to explain jargon if people don't understand what you're talking, if you don't understand the subject, who wants to take this?

Andy Cotgreave (37:52):

I feel like we sort of answered that already. My three points about how to present data, Leah and Brent talked about that as well, probably don't need to repeat those. And then I think as Brent said, which audience are you talking to? If you've got six audiences, which ones do you have to sacrifice? And then I think it's fair to make assumptions that not a lot of people know the technical terms you're talking about. So err on the side of caution.

Brent Dykes (38:19):

What do you benefit from using jargon? Jargon, right? It's a shortcut for a technical audience. If it's not a technical audience, then the jargon is actually problematic. People don't know it. So I would say avoid jargon unless there's an advantage of using it, speaking to a technical audience that quickly gets the shortcuts that you're using.

Lea Pica (38:44):

Absolutely. If I can add to that, oh sorry.

Richie Cotton (38:47):

No go.

Lea Pica (38:49):

If I can add to that. So two things. One of the biggest mistakes I see is practitioners or presenters using a lot of acronyms for technical jargon. So not only are they using the jargon, but they're plastering it with these abbreviations and people are like the ACL and the Z, YX, what? So if you're going to use acronyms, especially jargon, make sure that the audience understands what they are, you define them in the beginning and then you can repeat it later. And then the other thing, and I know Brent is going to back me up on this one, huge fan of analogies, if you have to explain something technical in order for the audience to understand the story. So if the technical part is necessary to understand the story, analogies are mini stories. They're a comparison between a concept to something relatable in the real world that people can make a connection and then understand something that's conceptual and inaccessible and it makes it accessible to them. So I think there's great examples in our books for that as well.

Andy Cotgreave (39:59):

Can I just add one more example? Sorry. Something particularly giving technical presentations. If you're teaching people how to use Tableau or Excel or whatever, you've got the content you can deliver in the talk, but then we have blog posts, we have LinkedIn. So really good practice is I can deliver content at this level in this 40 minutes, but here's a link to all the context and further reading. So all the things you can't get in the talk, you can have a link live when that talks, go when that talk's ready to go. And that works for technical talks, but also for important presentations you might do with data in an organization.

Richie Cotton (40:35):

Excellent. Alright, we've got two more questions that I want to get to and we've got four minutes left, so let's go. This one. I like this question. How do you tell stories that may be uncomfortable to tell or upset stakeholders? How do you upset your boss without causing a problem? It's a tricky one.

Lea Pica (40:56):

Can I?

Richie Cotton (40:57):

Yeah, go for it. I think no one else wants to answer this

Lea Pica (41:01):

Really quick. It's about a mindset shift and a reframing of what bad news is because every story has bad news in it or else it's not a story. So if something negative doesn't happen in a story, there's no story. There'd be nothing to watch. There has to be a conflict or challenge of some kind. So when you reframe news they don't like as just the next challenge or villain to overcome, but you back that up with a plan of action to show how you plan on turning things around. Now you've created a potential hero story out of what feels like bad news. And the other thing I'll say is it's very important to just acknowledge that upfront that that might be challenging news to hear. It's the truth. And what you do as an analyst is you are in service to the truth, but you are acknowledging that it didn't go the way they hoped. So you quickly, very quickly have that plan of action as a backup. That's worked well for me.

Brent Dykes (42:05):

I like that. I would add that empathy is really a key component that we need to have here. Empathy for the audience and really understanding politics and different things. You need to be careful entering into a situation where you're going to embarrass a team or a stakeholder or something like that. That's never an intention, but we could inadvertently do that and then that kind of defeats the purpose. So we have to have empathy for the audience. We have to think about how we position things. Sometimes it may be doing what's right for the customer and making, it's not about placing blame anywhere. It's about how do we correct this so we can help the customers do something better or achieve what they're trying to do. And so the blame isn't on the people, it's on the process. Maybe that is broken that we need to fix for the betterment of the customer.

(43:01):

And maybe that's the shift we focus. Or I was talking to an analyst yesterday and she talked about how she worked with different cultures and in one particular culture she had to be very careful how she positioned things. And so she started, maybe there's like three or four bad things that she had to bring up, but there was one good thing. And so with this one culture, she would start with the one good thing and say, Hey, this is a positive. Now let's get into some of these other negative things. And so that's part of that empathy and just being emotionally intelligent with how we present our things rather than just like, oh, the data says, the data says and too bad and you're screwed and we're all screwed. And no, we're not going to do that. We're going to try and be as empathetic to the audience and emotionally intelligent as we can be.

Lea Pica (43:55):

I have to add that Brent I had was coaching start a presentation, practice session with me and he goes, okay, here's how I'm going to start guys. I got the bad news, the bad news. I was like, no, no, stop.

Richie Cotton (44:13):

We lost everything. Okay, we got 30 seconds left. I'm sorry we didn't get to all the audience questions. So many great questions from the audience. We'll have to get the speakers catch up with you later. Okay, so 10 seconds each. Final advice on how to tell stories better. Brent, do you want to go first?

Brent Dykes (44:35):

Yeah, I think the key thing for me, one simple lesson is remember, and it kind of highlights what Andy was talking about, there's a difference between exploratory kind of approach that we do in dashboards, we do in reports sometimes. And what we're trying to do in data storytelling, which is more that focus on explanatory, right? We're trying to highlight, we're trying to simplify, we're trying to really laser focus in on a key thing that we're trying to communicate and think about every slide when I don't really use the word slide, I say data scene in your data story has to have a purpose. What's the takeaway? What's the one thing? Don't try and pack in multiple takeaways in one slide or one data scene. Each

Richie Cotton (45:17):

Data scene. Alright, I'm going to have to cut you off there. That's one 10 seconds. That's fine. That's fine. Andy, 10 seconds.

Andy Cotgreave (45:25):

Alright, if I'm in the presentation you are giving and I can't read the fonts on your slide, I will deliver all my wrath onto you. So make sure the fonts are readable.

Richie Cotton (45:33):

There you go. Excellent. Thank you Leah.

Lea Pica (45:36):

Okay, read a story, watch a story, watch a movie, and notice they all have common things. Drama, anticipation, suspense, a clear, satisfying conclusion. How can you bring that to your presentations?

Richie Cotton (45:50):

A clear, satisfying conclusion. I like that. That concludes our session. Thank you everyone.

Andy Cotgreave (45:57):

Good luck everybody. Thank you. Thank

Lea Pica (45:58):

You. Best of luck. Thanks to you guys,

Richie Cotton (46:00):

Everyone in the audience, head to the next session. It's going to be great. Yeah, thank you to the three speakers again and I hope everyone in the audience learn something. Alright, see you in future sessions later today. Bye.

Lea Pica (46:14):

Namaste.