#298 Data Storytelling Skills to Increase Your Impact with Kat Greenbrook, Author of The Data Storyteller's Handbook

Hi, I'm Maggie, VP of Curriculum at datacamp and I know firsthand how overwhelming it can be to keep up with the rapid changes in data and AI. But staying ahead doesn't have to be hard. That's why our curriculum team has curated the best learning content to help you and your team stay ahead. Covering everything from data and AI literacy for busy executives to cutting edge generative AI for developers, and everything in between. Whether you're upskilling yourself or your team, we make learning practical, engaging and flexible. It just works. Power your data and AI transformation journey. On datacamp.com you could write a really interesting data story for yourself, but if you're giving it to people who aren't interested in that topic or that data, it's not going to be classed by them as a good data story. We're not just sharing, hey, I'm building this model, we're sharing. This is the problem that the business have that this model is going to solve and the flow on effect from that is that I'm going to create this positive impact that's going to be good for the business.

Welcome to Data Framed, this is Ritchie. Today we're revisiting one of my favorite topics. If you've ever tried to explain something with data and your audience started dozing off, then you need data storytelling. Humans have spent thousands of years communicating with stories because it's a really effective way to keep your audience interested and get your point across. I'm a great believer that learning to tell data stories is one of the highest impact skills you can learn. Kat Green book is the author of the best selling Data Storytellers Handbook. Her book is full of frameworks for categorizing data stories and providing structure and incorporating data stories into your workflow. Today we're going to learn about these frameworks and get tips on how to tell a great data story. Beyond her book, Kat also runs the Rogue Penguin data storytelling training and business communication consultancy. Let's hear Kat's story. Hi Kat, welcome to the show.

Thank you for having me. I'm looking forward to it. Cool.

So to begin with, what's the greatest data story you've ever heard?

That is a tough question and I'm going to start slightly. I know, no pressure. I'm not going to share the greatest data story because that's way too much pressure. But I will share some data storytellers whose work I think is amazing. I really admire the work of Daisy Chung. She does some really beautiful scrollytelling. She's done a piece that pops her mind on fentanyl and she gets into the chemistry of. Seems very technical, but the way that she writes it is just so accessible and approachable. I think it goes with her beautiful illustrations to make it more engaging. And I. I love reading her pieces. I really admire the work of John Byrne Murdoch. I don't know if you've come across his work before. He works for the Financial Times. I came across his work during COVID So he was posting on Twitter at the time and sharing all these graphs, but also weaving the context and the meaning behind them in the way that he told them. And at that time, it was very, very powerful because it was what people needed to understand that kind of information. So greatly respect him as a data storyteller. I think a lot of the data storytelling that seems to make it into the public domain is coming from these media outlets or news outlets, because that generally has been historically the domain. And now that some of these journalists are upskilling in data, they're combining what they know about people and narrative and story with data, and those are producing some really, really beautiful data storytelling pieces. I think a lot of the more traditional people that work in bi, they work for companies and they're not necessarily allowed to share a lot of the work that they do. So their data storytelling doesn't necessarily make it into public domain. And so what we see and what we read about are being produced by these data journalists.

That's interesting that you've got data journalists and business intelligence analysts doing, I guess, similar things or maybe different styles of data storytelling. You only kind of get to see one side of things. It's a bit more public. So you mentioned a term, scrollytelling. I've not come across that before. Tell me what this is and how you might make use of it.

Yeah, scrollytelling is just making use of scrolling on websites. And so you start from the top and you scroll down, and as you scroll down, the story unfolds, whether that's in a visual sense, so you can have interactive things moving and popping up as you scroll along, or just the narrative advances. So it's just a fancy term for scroll down.

That seems straightforward and very useful because I guess a lot of stories are consumed on the web and you want to have an experience that goes alongside that rather than, I guess, flipping pages in a book. Do you have any more sense of what makes these things great? You mentioned a few data journalists there. What's sort of the essence that makes the data stories great?

Yeah, there's quite a few things that can make a good data story. And a lot of that depends on who your audience is. And so what are they interested in? It's gotta be relevant. So you could write a really interesting data story for yourself, but if you're giving it to people who aren't interested in that topic or that data, it's not going to be classed by them as a good data story. Right. And so you've got to think about how it's going to resonate with your audience. A lot of data stories that go viral are the ones that are quite topical for what's going on within the political landscape or within the social landscape, whatever it is. They generally are something that the audience can relate to and the audience care about. And also they have a good structure to how the information is presented. So they're not just throwing a whole lot of data at people and expecting them to kind of piece it together. They've done all that hard work for the audience. And so the audience can actually just enjoy absorbing the information because it's already presented in a way that they can.

So I like the idea of doing a bit more in the sort of content creation side and then making it easier for the audience. So I'd love to get into some of the details around this. I know in your book you have this idea of it's the problem goal, action, impact framework, which is to help you do this. Talk me through what is this framework and how do you use it?

Yeah, so this is the very, very basics of how a business should kind of operate. You have a problem that is trying to be solved, and problems and opportunity go hand in hand. So could also have an opportunity that you're trying to grasp from that you set a goal, whether that goal is to fix the problem or take advantage of the opportunity. In order to achieve that goal, though, you have to do something, you have to do some sort of action, and from that action, you would hope that as well as achieving the goal, you create some sort of positive business impact. Because that's what we're all trying to create in the jobs that we do. We want to leave a positive impact in the organization that we're working with. So that's the basic structure of how you work. Everything you do in your job is an action of some sort that is hopefully going to achieve a goal and to solve a problem or create an opportunity, and that action is going to drive a positive impact. But unfortunately, most of the time we don't know anything more really than the action that we're doing, because we've been told to do it. So just taking the time to think about, okay, I'm doing this piece of work. What is my goal for this piece of work? What problem within the business am I solving? What action am I creating further upstream that is going to produce a positive impact from the work that I do? And having this understanding of the problem, the goal, the action, the impact, the hierarchy in which that we are existing, it can help when we share the work that we do, so we are able to influence better. We are able to share more of the context around the work that we do. So we're not just sharing, hey, I'm building this model, we're sharing. This is the problem that the business have that this model is going to solve. And the flow on effect from that is that I'm going to create this positive impact that's going to be good for the business. But we just don't really take the time A lot of in the situations that we work with, we don't understand those problems and goals and impact of the work that we do. We just understand the action that we take.

Yeah, I can certainly see how I've built this model. It's not going to make a lot of sense to many of your colleagues or certainly people you're not working with directly if you don't provide the extra context. So I like that idea of like, well, figure out what problem you're trying to solve even before you're doing any work. And then at the end of it, you're going to try and figure out what's the impact of my work. In fact, it sounds like this goes beyond data storytelling even to. It's like, it's probably a good idea to think about a lot of these things for any kind of work you're doing. Do you have a concrete example of how you might put this into action?

I mean, we could go with that kind of model example. Let's say you're an analyst and you've worked on a model that's going to analyze customer churn, or you've done some analysis around why customers are leaving the organization. And you might think, okay, I'm just going to share the visualization results of that, the output of my model. But you can frame it using the problem and the goal and your action and the impact you hope to create in a way that's actually going to engage people. Think of it as kind of like an elevator pitch for the work that you do. So you could say something like, our organization has struggled with increased customer churn. For at least 18 months. But I'm analyzing these customers to understand why they're leaving us. And as a result of my work, I hope to add value to our organization through the use of better retention strategies. So just including some of those downstream and upstream things, problem and impact, you can just weave a little bit more context around this very, very sometimes technical and specific action that you're working on, and more people will be able to understand and realize the value of the work that you do.

That makes a lot of sense. So it's like what happens kind of upstream and downstream from it, it's just all around the problem is just given that sort of the business cushioning, I guess, when do you bring it into your workflow? Is this something you think about, like after you've done your analysis? It's like, well, how do I now present this to a group? Or is it something you need to think about before you even start analyzing data?

It can be both. So you can think about this kind of thing before you do the analysis. So you can use data storytelling to help share the results of something that you've done, or you can use data storytelling to help influence an action that you are about to do. So data stories revolve around business actions as a way of sharing them or influencing them. And so having an understanding of your problem, goal, action, impact, before you go to do the analysis can help you influence that buy in. So it can help you actually get that piece of analysis or whatever action it is off the ground. And if you're sharing the results of something you've already done, you can use the problem, goal, action, impact to help weave that context into the way that you explain the importance of the work that you've done.

I like that. And so it feels like this is fairly flexible in terms of when you make use of this, sometimes you're going to want to think about the whole story beforehand, sometimes afterwards. Sometimes I guess it comes up in the middle. So it seems like it's sort of like a parallel task to the actual analysis.

Yeah, I think this is just about understanding the world in which you're working in. So that could be a business world, that could be a scientific field, whatever world that you're finding yourself in. Having an understanding of what problem you're solving and what influence and what impact you're trying to create is a good thing, regardless of whether you're doing data storytelling or not.

Now I'd like to talk a bit about tailoring your story to different audiences because it seems like if you're speaking to a business person or speak to a scientific person or the technical person is going to be with a very different story. So do you have any examples of how you might want to tailor your data stories to different audiences?

It can be very specific to some audiences. So you could have the same output from piece of work that you've done and you would have to produce two different data stories to meet the needs of two different audience groups. Going back to the example we talked about analyzing Customer Churn, you could say that your initial audience might be your boss, might be your manager, and you could show them what you found in terms of this is what's happening. But you need to think about their motivations, what they care about and probably for someone in that position, they're going to want to know, did you go through very accurate data analysis process in terms of I don't want anything being shared from this team unless it's really accurate because it's going to come back and bite your boss. Right. So your boss is going to be very, very concerned about if anything goes outside their team. It needs to be have gone through a very robust analysis process. So they are going to be focused on that. But if you were sharing that insight with say the marketing team, they will want to know what happened. Yes. But they will also have a focus on the future, what's going to happen in the future. So they probably don't hear about how accurate your analysis trust you. They trust the position that you have within the company. You're employed because you're good at that. That's not a concern for them. They don't care about how you've done the analysis. They just care about what it means for them going forward and how they can use it to make their jobs easier. So it all comes down to audience motivations and what they care about. Because you really have to think about that in terms of what you communicate. If you were to go and communicate and very in depth playbook about how you've done the analysis to a marketing team, you're going to bore the hell out of them. So it's just being very, very aware of what it is that your audience care about and also how much technical detail they need. So your boss is probably going to be very interested in the technical detail. Your colleagues are going to be interested in the technical detail if you're a data analyst. But anybody probably outside the data team is not going to need those technical details. You're going to drown them in information that they're not interested in. So you have to be aware of what interests your audience.

I've said the story hits home quite hard. I remember, like, earlier in my career, I'd moved from working in a scientific organization, like talking with other scientists, going into a lot of technical detail, to going to a commercial organization, speaking with customers. And, yeah, I started presenting about data cleaning processes to some executive. Then it was like, at the end, it was like, oh, yeah, and this is how you're going to save £1.2 million. And he's like, yeah, you could have led with that.

Yeah, totally.

Didn't so much care about data cleaning. Okay, so it sounds, from what you were saying that this sort of tailoring, it's not just about changing individual sentences or phrasings. You have to completely restructure your story or like the entire content, depending on who you're speaking to. Is that right? Do you have to start from scratch every time?

You don't have to. Not every time. So if you have very different audiences, yes. I would recommend doing two different forms of communication, but sometimes it can just be switching terminology. So you could be, if you're talking more public storytelling, instead of saying spending on our people, you could say something like investing in our people. And just that switch from spending to investing has a different way of framing whatever it is that you're talking about. So just sometimes thinking about how emotive words are, and depending on your audience, some words will be very emotive. Like spending can be a very emotive word to some audiences. And reframing that to investing, it's going to resonate differently with them.

I'm curious as to whether your approach to tailoring changes depending on whether you know the audience or not. So imagine like talking to your team or your boss or whatever. You can know them quite well, whereas, like, on this podcast, we don't know all the people who are listening. So what's your approach for tailoring stories depending on, like, whether or not you know the audience?

I think it's definitely easier to create a story if you know your audience well. But you're right in some cases, that's just not possible. And you have to kind of make an educated guess about what their preferences are going to be. Are they going to like a lot of detail or are they going to prefer things in a summarised format? Are they going to make a decision more based on, or are they going to like to pour over all the data? You're going to have to ask yourself some of those questions and just kind of think about okay, my audience is kind of sitting over here. I think they're going to prefer a more summarized format. Generally, if you go away from the technical detail into a more public audience, they're going to prefer a lot more context and far less jargon. So you have to think about how much technical detail does your audience prefer? And if they are working in a technical space, then maybe that's okay. But even some technical spaces don't understand certain jargon terms that maybe you're very familiar with and you don't even think about anymore. Sometimes we fall under what's called a curse of knowledge, where you forget what it's like to not know the information. And so therefore, you just communicate at your current level of knowledge and forget that other people don't have that. So I like to, when I'm going to a general audience, whether it's. If you're communicating to a wider group of people in your company, start with the person who has just joined the company, so today they've just joined the company, you're presenting to them, and work your way into the detail rather than starting with the detail, because if you start with the detail, you're just going to lose them.

You mentioned the idea of explaining technical details. I think this is something where a lot of people tend to get tripped up. Like, you've got a lot of numbers, technical methodology. How do you go about explaining those depending for different audiences?

Yeah, I think you have to, first of all ask yourself whether your audience needs to understand those technical details, because most of the time they don't. Especially when you're talking about methodology and technical details about how you did something that is not really important for the majority of people on the receiving end. Generally, people care about the results and what it means for them rather than any of the detail in terms of how it's been done. So I think that's the first question you have to ask in terms of communicating technical information is whether or not your audience needs to understand it in the first place. And then just kind of what I said before about adding more context and removing that jargon, just really, really being aware of your own curse of knowledge. Everybody has it. It's not a bad thing, necessarily, but it's just being aware that it exists and trying to think about reframing what you know to a more beginner's perspective.

Okay, so really just including, I guess, the minimal viable amount of technical terminology that you can get away with, does that sound a little bit.

Yeah, I mean, depending on what your communication is. So it could be a report, it could be a presentation. You can always include that technical detail if people ask for it. So reports are set up so that technical detail can live in the appendix and presentations are done. And you could say at the end of it, if you get any questions, I can send that detail through to you. So you can still have it there if people ask for it. Just don't start out with it.

I like the idea of using appendices. It's like, probably no one's going to read them, but if there's that one annoying person asking the questions, then, yeah, you've got the evidence to back up your arguments. Okay, that seems pretty cool. All right. And just while we're on tailoring things, I know in your book you've got different sort of archetypes of stories. So you got. I think you mentioned the one about, like, a character data story. And what about a time data story? Just talk me through what are these different types and when you might use them.

Yeah, so these are the two main types in which you would actually contrast data measurements. So data stories are built around how you can contrast your data measurements. So a time data story uses time as a way of contrasting these measurements. So it looks at, okay, last year we had this expenditure, but this year we have this expenditure. So we're using the difference in time to contrast that data measurement and character data stories. Look at how these data measurements are different across different characters. And a character in a data story is not like a fictional character from a book. A character in a data story is generally whatever the data is describing. And so it can be a company, it can be a product, be an animal, it could be a person. So a character data story. Character data story contrast looks at, okay, our business did this this year, but our competitor did this. And so we're using different characters to contrast our data measurements. And contrast in storytelling in general is very important. Stories are built around contrast, and data contrast is just another way in which we can add that contrast into our storytelling.

So moving on from tailoring to structure, I think there are a lot of kind of standard story structures. So there's the hero's journey and all these kind of things. I'm not sure which ones of these are applicable to a business analytics scenario. What sort of structures do you think work for data stories?

Yeah, basically. I mean, any story structure can work for a data story. I think familiar with the hero's journey. I think it's a bit too involved for a data story, especially the data stories that are used within Business as just a way of general communication. They don't have to be that detailed. They don't have to be that hard to do. I like to use the and, but therefore narrative structure, which is basically a form of the three act structure, just to make sure that you include those three acts in terms of how you present information. So making sure that you have your act one there to set the scene to provide that context, you have your Act 2 there. Generally it can go into what is the problem we're solving, what are we banging up against, what are we finding hard, what's our struggle? And then Act 3, this is what we're going to do about it. This is our conclusion, our resolution, whatever it is. But those three acts, hitting those points creates a full story. It's not just a case of sharing random numbers without a structure. Having a story structure in place is the backbone of your data storytelling. So, yeah, there are really complex structures out there. The hero's journey, one of them. I just don't think that amount of detail is necessary for business communication.

Okay, so I like that sort of simple idea. So let me see, there was scene setting, and then some sort of tension or problem and then some sort of hopefully happy ending, I guess. In Act 3. Can you talk me through like an example of how the three acts work for a particular use case?

I can give you an example of how it works just in general storytelling. And I can give you an example of how we've. In data storytelling. I've got an example here for general storytelling and I'll just read it off. It's an and, but, therefore. So they use the words and, but, and therefore to define those three acts and see if you can guess the movie. It's based on a movie plotline. So I've made it quite simple. It starts with, Ariel is a young mermaid princess who dreams of exploring the human world. And after she rescues Prince Eric from a shipwreck, she falls deeply in love with him. But her father, King Triton, forbids any contact with humans. Desperate to be with Eric, Ariel makes a deal with the sea witch Ursula, who grants her legs in exchange for her voice. Therefore, Ariel embarks on a quest for love and her life in the human world.

All right. Yes. Ariel, the Little mermaid. I like that cartoon. When I was a kid, it was like, ginger protagonist. It's good stuff. Go on. And what's the Deidre card?

So those are the three acts. And so the word and helps join together statements in the first act to provide that context so that all the Statements and is the most common agreement word. They're all these statements that you join together in the first act. They're all in agreement. Agreement. They're all helping to set the scene, introduce the character. Barthes introduces the second act because it's a contrast word. So it's providing that tension between Act 1 and Act 2. So when King Triton forbids any contact with human, that's the tension of the story and therefore helps resolve the tension that was introduced in Act 2. So we're introducing Act 3 with the word therefore and continuing the story on in a way that's going to conclude it. And we can do the same with our data stories just by using these three words. And I'll give you a very, very simple example, but you could say last year we had 20,000 customers and held a large proportion of the market share. But this year customer numbers dropped 25% because of increased competitor activity. Therefore, we need to think about how we are going to win back our customers. So that's an example of a data story. I don't think it has to be as involved as the Hero's Journey structure. It can be simply these three acts, but making sure that we have something for each act will help us to tell a more balanced story.

Okay. And this sounds like it's going to feed naturally into that understanding what your problem is. The previous PGAI framework we.

Yeah, so having having an idea of the problem that your action is actually solving will definitely help in terms of how you put that story together because it can weave through some of that context for your audience.

Do you always want these three acts or do you ever want to like break your story down into smaller building blocks?

So I always start with the three acts and the more general you can make it, the better because then you can start to go deeper if your audience needs a little bit more detail. And so I call these nested stories or stories within stories. And so if you have a problem statement for your second act and you find your audience is probably going to need a little bit more information about that problem, more than just maybe a sentence or two, then you can actually create another abt structure, another story within a story that just supports or goes into more detail about that problem. And so this is how you can flesh out your high level stories but still keep them engaging so you're not just throwing a whole lot of information at people. Okay.

Stories within stories. I can imagine this could get quite complicated, some sort of inception type situation, lots of layers. But yeah, I guess at some point like three sentences worth of story isn't going to be enough for some people when they want more detail.

Yeah, you'll be surprised though, because I think it's actually enough for a lot of people. There'll be times when you obviously have to create more detail, but for the majority of situations in a business context, you can get away with just that high level story.

Okay, yes, but if you're like presenting to the whole company about like whatever your project is, most of the company only really wants to have a very, like, they only care a little bit. So just having a high level overview is enough. All right. I guess the other thing with stories is how you make sure that the story is an honest representation of the data because it's very easy to kind of massage the results to fit your story rather than the other way around. So how do you stay honest?

Yeah, and this is a good question because storytelling, I think, gets a bad name in terms of people don't trust it because they do think of it as something that is made up, something that people use to manipulate other people. And that's absolutely true. People do make up stories. And people do use the engagement of stories to manipulate other people. There's no denying that. But I think they use it because it is so powerful. And if people can start to use it to present facts, present honest data, then it can be just as powerful in resonating and influencing them. But there does come ethical decisions and things that you have to make along the way. If you call yourself a data storyteller. So you have to make sure that your source is reputable. You have to check your own biases. So are there conflicts of interest that maybe you have in terms of how you interpret the data and how you communicate that data? It helps to understand any assumptions that you're making. So assumptions are fine to make. I think when doing data analysis, sometimes we have to make assumptions, but just understanding why you've made them. So looking at correlations versus causation, data accuracy, have you made assumptions around that that maybe just writing down will help you actually readdress and think, okay, is that right? Or even just asking another person, do you think this is right for me to have made this assumption? And also when you are writing a story, ask yourself, what is the counter story and is that true? Because data can tell different stories based on the context in which it's examined. So you could be looking at a specific number, and if you compare that number to a higher number, it seems lower. But if you compare it to a lower Number, it seems higher. So you can kind of shift how people respond to that particular number that you're sharing, whether it's in a good or bad way sometimes. So you really have to think about, yes, I'm pushing this particular narrative, but is another narrative also true that could contradict this one? And you have to be comfortable with data holding multiple truths. And how you communicate that can sometimes be quite difficult because you have to communicate those multiple truths. Oh, man.

It was starting to sound quite simple. Like, you just need to make sure that you've got these checks in place, make sure your analysis is kind of correct and that your conclusions are valid, and then just try and communicate this as accurately as you can. But then the idea that you've got multiple truths within data, so you might have to communicate things in a different way. Actually, first of all, do you have an example of this? And then how do you go about communicating these different sort of truths?

Yeah, I think the different sort of truths, I don't have that concrete example in my head right now, but it could be just in the perception your audience has of particular data. And so you could have, let's say, the dairy industry. You could have people from the vegan side of things and who see the truth of the dairy industry in a very negative way in terms of environmental damage. You can see another camp of people looking at the dairy industry and thinking, thus applying food to the world. This is a very positive thing. Both of those things can be true. And it's finding a balance between. You don't want to necessarily just show one side of that story because that's not the whole truth. You're just picking a sliver. And I mean, that's a pretty general problem. But you could have more detailed situations where if you just picked one side, it's not a true representation of that particular thing, whatever it is that you're looking at. So I just think the world is not as black and white as people think it is. There's a whole lot of gray in there. And as data storytellers, we have to get comfortable in that gray space.

Okay. Yeah, I can certainly see how anything that's going to be, like, remotely political, that's going to require some sensitivity in terms of, like, how you balance your story beyond, like, just pure throwing in numbers. And I suppose that even applies to, like, business context. There's always some kind of business politicking going on. It's like marketing attribution. So it's something people, like, with different teams argue about, like, who gets responsibility for bringing in customers. Okay. So, yeah, I can see how if the story sounds like it's gonna cause conflict between teams. Yeah, you've gotta be a bit sensitive there in terms of your framing. Does that sound about right?

Yeah. And I think it's not necessarily about trying to be political in the way that you communicate and you're not trying to please everyone, but you don't want to sensationalize necessarily things when you only share one side of the story.

Okay, all right. So I guess just watch out for your phrasing and maybe, I guess get someone to check it before you start that.

Yeah, I mean, running whatever it is that you're communicating with by other people, just to get a sense check is always a good idea.

Yeah, that seems like a wonderful idea. I guess. More generally, do you have advice on how to make use of data stories at work? I guess in meetings, that's like a common place of sharing information around data. How do you tell a data story in a meeting?

Generally, when you have a team meeting, maybe it's a weekly team meeting, and you kind of all gather around a table and you may have an agenda, you may not, but they generally led by the manager of the team. And sometimes people will kind of talk about what they're doing, sometimes they won't. Sometimes a lot of people will sit there and be silent for the whole meeting. I used to hate those team meetings, but they are a great opportunity for people to start sharing data stories within a team. So I do encourage a lot of teams to have a couple of people in each meeting for different weeks, but they each bring along a really simple data story and they just read it out to the team and it can be a couple of sentences and it can be about the work that they're doing. It could be about a situation that they've come across within the business, it could be anything. And this is just to give people the opportunity to practise, because sometimes we get so caught up with the work that we're doing, and if that work is very technical, you don't necessarily get a chance to practise your narrative writing and your narrative storytelling. And so giving people that opportunity, and it really doesn't have to take long, especially when you're writing these and. But therefore, they are just a couple of sentences. And if you're not comfortable presenting in a meeting, you just have to stand up or read out a few sentences. And it can be really daunting to start with, but it's giving everyone in that meeting, not just the people doing the storytelling. But everyone an opportunity to build their narrative intuition. Because the more you start writing stories, the more you start hearing stories and you start to recognize patterns and the framework that's being used. It'll just make you better and better at it yourself.

Yeah. So you mentioned team meetings. I have to say I find the boring bit of team meetings where you go through like all the kind of quarterly targets and it's like, hey, we're at 53% of the way to target is like, great. Is that good or bad? You're not quite sure. Having a bit of narrative around that has got to spice things up, I think, a little bit. So it's going to make your weekly team meeting a bit more informative and a bit more engaging, hopefully. Okay, so how about other forms of communication? Because I think a lot of business communication tends to be written, so things like emails, Slack messages, all that kind of thing. Can you do data stories within these sort of short form written communications?

Yeah, absolutely. So whenever I'm putting together a data story, it always starts out in the written form before it gets any design added to it. Because if you start adding the design before you understand the message, it becomes very difficult to make that design good. So all of my data stories start off in their written form first and they can be sent just like that. So you can write a data story, email it to someone, or put it on Slack, whatever it is. But that's the most basic form of a data story. When you get to start thinking about how you're going to tell that data story, sometimes that's just written form and gets emailed, but most of the time that incorporates some sort of data visualization, whether that's in a form of a presentation or a report, whatever is best suited for the audience because that's really what you've got to consider. How does my audience prefer to receive this information? And if that's in an email, that's great, but if that's in a presentation, then you have to think about, okay, well, how am I going to put that presentation together? What visuals am I going to design that are going to help me tell this story? Any design work needs to come secondary to the story.

And I guess the other extreme of that is something like a dashboard, where it's predominantly visuals. Can you tell us data story with a dashboard or do you need to.

Map adaptation then there's different views on this. I don't think you can use a dashboard to tell stories. I think dashboards are great tools to find stories in or Find the beginning of stories in Dashboards are tools that they present information in a very, very ordered way. So they make it very easy for people to see the latest measurements of the metrics that are important. But they don't give you that context. They don't necessarily change the way in which they show that data to certain audiences to make sure it resonates with them. So they're very, very much designed for people who already understand the meaning and significance of whether a number is high or low. The audience already has that context. Dashboards are a form of data communication, but they're not a form of data storytelling. Just like data storytelling is a form of data communication as well. It's just designed for a different audience. So I think dashboard's definitely a place to find the beginnings of stories, but they're not a place to tell those stories.

Okay, that kind of makes sense. If you've got a lot of charts there, then there's no sort of implicit story. Like I was thinking, like, if you've got, I don't know, your companies, like monthly revenue, hopefully the person who's looking at that is going to understand whether that's a good number or a bad number. You don't necessarily want like text within the dashboard to explain it.

Exactly.

Okay, and does it make a difference if your story is being told like directly to a person or if it's asynchronous? So I mean, we talk about like email or whatever, that's asynchronous communication rather than a face to face chat. Does that affect how you create your story?

Yes, I think so. So if you know that your audience is going to be there, so say you're doing a presentation, you can design your visuals to not necessarily tell your whole story. Because you're there, you're going to be presenting, you're going to be doing that part of the process. Whereas if you know your audience is not going to, you know, you're not going to be there with your audience when they read your story, for instance, a report, you're going to have to include a lot more of that meaning and context into the narrative in which you write down, so your audience can have that understanding. When I design visuals in a presentation, they include very little in the way of words because I'm there to explain it. And I want to walk them through the visual or walk them through the information as I talk through it. But that's obviously very different to the way in which I would design visuals. If I'm not there to do that.

Explanation okay, so I'm kind of convinced data stories are a good idea. You might realize that. But if you are convinced that data stories are a good idea, you want all your colleagues to try this as well. How do you get them to make use of data stories?

First step is actually understanding what it means to tell a story. So make sure you're all on the same page in terms of what you think data storytelling is. And there are different definitions of data storytelling. I've talked about what I believe data storytelling is, but you'll find other people have different ideas about that. And I think this is such a new field that it's okay for there to be multiple definitions of what it means to do this, because we're all trying to figure out and shape the field in terms of where it's going. And it's okay to think differently. But understand what you agree data storytelling is first and then practice on each other. Just talking about sharing in a team meeting. You don't have to go to a team meeting to do this. You can each pick a movie and come in the next day and write a story about the plotline of that movie. That is practicing using this story structure and practicing building that narrative intuition. And over time, it will become more intuitive for you to do this. And so start with something really easy. It doesn't have to involve data, but the same storytelling skills for a data story exist for a story. So it doesn't really matter what context it's in, just practice using that framework and getting comfortable thinking in story.

I do like the idea. Just take a movie you've seen, try and convert it to that and. But therefore structure and yeah, it's good practice for telling you data stories later on. You know, I was trying to avoid talking about generative AI this episode, but I'm getting withdrawal symptoms. So I got to ask. How has generative AI affected people's abilities to tell data stories? Like, how can you use it to help you here?

I think it's very powerful and it's only going to continue to be powerful, especially in this space. I'm not naive to think that it's not going to one day be able to do what I do. I do think some sort of AI will come along and be able to tell the stories of our data. It's a question of whether we let it, but I think that we will. I think it's probably going to happen after it automates or takes over the analytics space first. I think that's kind of the easy pickings almost for AI. To come along and. And start to help us out in terms of how we uncover those insights. I think that's already happening, so there's no reason to think that it couldn't flow on from that and then help us to tell our stories. I don't know if it's going to be able to incorporate all the context we need it to, because obviously it knows what we have fed into it. And so we have to get used to having that understanding ourselves and being able to direct. And that involves having, I think, a little bit more knowledge on how our work is contributing to the business, what problem we're solving, really going back to those basics so we can direct AI in terms of how it can help us communicate the results of what we found or what we've jointly found with analytics. So I definitely think it's going to play a massive part in the whole data space, and that is quite exciting. But it's a case of how much of it we will let go and let it do by itself.

Okay, yeah, so there's a very good point that you're always going to understand, like, the context of your audience and the problem better than the AI. But beyond that, it's very good at writing sentences, so it's going to have some benefit for creating the narrative.

Cool.

So do you have any final advice for how to tell a best estate story?

I think lose the ego. So a lot of the time when we communicate our work, we're super proud of it, and there's nothing wrong with that, but we have to. It's less about. It shouldn't be all about us. It should be about the wider context, the wider problem, the wider impact that we are creating or our work is creating. So having it be less about us and more about the value that we hope it will create, that's how it's going to resonate more. People don't care how we've analyzed the data. They care about what we found. And so we really have to lose the idea of, I have to communicate all the important things that I know in order to be seen as valuable, because that's just not the case. You have to communicate what you have found from using your skills, obviously, but how that is going to impact other people, and having that understanding of the business or the field that you're working with just helps to ground that information.

It's interesting that knowing when to shut up is an important part of telling a data story. Also a problem for podcast hosts as well. All right, wonderful. So just to finish, I'm Always on the lookout for people to follow. So whose work are you most excited about at the moment?

I actually, I've just enrolled in a university paper. I. I have taken a psychology, not psychology, a sociology paper this year because I want to. I kind of. I'm a bit worried about the future of data. I think with what is happening in the world right now, there are massive decisions going on, massive decisions being made that have used data in a manipulated sense or they've just completely disregarded data, and data hasn't been considered in how that decision has been made. And so I'm just a little bit worried in terms of how important data is going to be moving forward. And I think the trust that we once had of data, it's starting to erode a little bit and it's losing some of its social license. So I'm most excited about learning more from the sociology point of view in terms of how societies work, how people allow things like misinformation to spread through societies. I think having an understanding of that as well as the data side of things, because I still think data communication is a very important part of how these decisions should be made. But I also think it can't be the only part. Moving forward, we have to have this understanding of people and how we work as a society and, and how power works and how inequality comes to play. So I think for me, that's a real focus for me this year and I'm super excited about learning more about that.

Okay, I was going to wrap up there, but you've just terrified me.

Sorry.

Everyone's losing trust in data. Okay, so do you have any insights yet from this new sociological learnings, like what do you need to do to bring back that confidence in data?

I am still trying to work that out. So I had my first lecture yesterday and the first paper I'm doing is about social inequality and just learning about these different ways in which we look at society and how some people in society have been left out in terms of. Left behind. All these different terms that I've never been exposed to before. So don't ask me to summarise it in any way. It's very, very early days, but I'm just. I'm very excited to be learning more about that.

There's definitely something to be said for just learning about a completely new field and getting a different perspective on reality. And that's going to just broaden your ability to think critically about things.

Yeah, I think so. One of the things that we did talk about in the lecture was actually figuring out your social position. So where you are in society, you're probably going to be surrounded by a lot of people who are very similar to you. So there's this such thing called social distance, which is not the social distancing we talked about during COVID but social distancing. The bigger the social distance between you and someone else, you know, you're probably not going to have a lot of things in common because you tend to hang out with people who are very similar to you. So just having a realization of where am I sitting, why am I not talking or interacting with this person over here, I think that helps you think about empathy differently. It helps you consider somebody else's perspective or point of view based on their experiences because they're so different to yourself. And this is really important with storytelling in any aspect, because we have to understand our audience. We have to respect the people that are represented in our data. So having and building empathy in some way for different audiences is very important.

That seems like incredibly useful advice. Just think about, like, how different people are from you and start talking to people who are very different from you. Okay, so that's homework for everyone in the audience. Then go and sit with someone new at lunch tomorrow. Yeah. It's going to benefit you and the other person, I think. All right, wonderful with that. We're going to finish, so thank you so much for your time. Kat.

Thank you very much. It was lovely to talk to you.