Effective Data Storytelling for Financial Services

Foreign, foreign, and thank you for joining today's webinar. My name is Rhys, and I'll be your moderator today. We're going to kick off the session in a couple of minutes. We're just waiting so everyone has a chance to join. In the meanwhile, though, we'd love to hear from you. So let us know where you're joining from using the chat or the comments, depending on what platform you're watching on. And yeah, tell us something that you'd like to get out of today's webinar. Uh, if you do have any questions at any point throughout the session today, then please let us know, uh, in the chat, in the comments, and I will be saving them for the Q&A section as well. Um, a few bits of housekeeping before we get started. The session is being recorded, and the recording will be emailed to everyone that has signed up for the event and registered on DataCamp. I will be putting a link to that in the chat and the comments as well. And as mentioned, we are going to have a Q&A as well. So if you have any questions at any point, then please let us know using comments. Uh, brilliant. I'll be back to repeat these messages for any new joiners shortly, but until then, enjoy the backup music.

Foreign. [Music] Foreign. And thank you for joining today's webinar. My name is Rhys, and I'll be your moderator today. We're going to kick off today's session in about a minute or so. We're just waiting so everyone has a chance to join. In the meanwhile, though, we'd love to hear from you. So let us know where you're joining from using the comments. And yeah, tell us something that you'd like to get out of today's webinar. If you have any questions at any point throughout the session, then please let us know using the comments as well. And don't forget to get signed up for the event. So if you get registered, we will send you the recording, and we will send you all of the resources that we shared during the session as well today. So I put a link to that in the chat, in the comments. I'll be posting it again shortly. Uh, but if you can't find it, head over to our LinkedIn, and it's a pretty easy find. You'll see the sessions live, and you can check the comments there. Brilliant. I'll be back to repeat these messages shortly, but until then, enjoy the background music.

Foreign. Hello everyone, and thank you for joining today's webinar. My name is Rhys, and I'll be your moderator today. We're going to kick off today's session almost immediately, in a couple of seconds or so. We've just been waiting so everyone has had a chance to join. Let us know what you're joining from using the questions, and, uh, sorry, not the questions, the comments and the chat. And yeah, tell us something that you'd like to get out of today's webinar. Um, if you haven't already, make sure you're registered for the event, and we will send you the recording and all the resources that we share as well. And if you have any questions at any point throughout the session, then please let us know using the chat or the comments, depending on what platform you're watching on. Brilliant. I think that's everything from me. So now I'll hand you over to your host for today's session, Richie. Richie, please take it away.

Welcome to the webinar, everyone. This is Richie. Now, I spend a lot of time interviewing data experts, both in these webinars and on the Data Brain podcast. And one of my favorite questions to ask is, what are the most important soft skills needed by data professionals? And almost everyone says communication skills. So this is something that every manager agrees on. You need to get good at communicating. And when you dig deeper into communication, a lot of good communication, particularly for knowledge sharing, is about storytelling. Because humans have been storytelling for thousands of years, and the fact that the technique is still around means it works. And all this is a long-winded way of saying that storytelling is a fundamental life skill, never mind in the context of work. And today, we're covering data storytelling for financial services. And this sort of more niche, uh, part of storytelling provides some really interesting constraints on the kind of stories you can tell. Because you've got technical language around data and finance, and then you've got some very particular audiences in financial services, and they have their own expectations and culture. So, uh, teaching us today is, uh, Nikki Cross, the Director of Data Science Solutions at Nova Credit. So Nikki has an incredibly fascinating job because while she runs a data science team, she's also responsible for communicating what her team is doing to clients. That is to say, uh, telling stories about data to financial services customers is her forte. And so I'm going to, uh, interview her and find out what she has to say. So, uh, first of all, uh, welcome to the webinar, Nikki. Glad to have you here.

Pleasure to be here. Thank you so much.

So, I'd like to dive in with a bit of motivation. So, uh, just to begin with, uh, can you tell me, uh, why is, uh, communication so important? Why can't we just let the data speak for itself?

Yeah, I think for a lot of folks, people think that, you know, we've got all this data, like, we'll share that, and people will kind of figure it out on their own. And data without a story, it just isn't enough, right? So we've got to be able, as data practitioners, to provide those insights that are targeted to the need and to the audience, to their specific use case. I think there's a, there's this concept of like a data, information, knowledge, wisdom pyramid. How do we keep kind of leveling up the information that we have in a world where there are infinite numbers available to us, infinite things that are all true, but most of them very, very irrelevant to the folks that we're talking to, to the work that they're doing? And so that's, that's the beauty of kind of what we do. I say that's our job security in an AI and increasingly, um, advanced world, is that we have to be able to actually take the amount of information, which is more than has ever existed in the history of the planet, we have to take that and we have to make it make sense. And so that ability to really think through what, what we care about, what the story is telling here, and what's going to be relevant to our audiences, to keep them engaged, to allow them to do the work that they need to do using this information, that I think is the art of the work that we do as data practitioners. I think that's the practical outcome, and that's where we have to be able to get those partners to be able to really be effective in our work.

Some great points there. So, first of all, you mentioned this, uh, data, information, knowledge, wisdom pyramid. Uh, I think, uh, we have a cheat sheet on that. Rhys can perhaps, uh, share a link to that in the chat. But you also said that, um, telling stories is about not telling people things that are irrelevant. And that's kind of an interesting negative framing, uh, because you think, well, okay, it's about trying to tell people things that aren't relevant. But, uh, also like not telling stuff and editing yourself down, that's also like a really useful skill. Uh, maybe we can get into that more in a moment. But before we get to how you tell good stories, I'd like to talk about like, what can go wrong first. So, um, can you tell me like, what you think is the biggest mistake in communicating with data?

I had a former boss who had, who's a very, he's a very colorful character. And one of the things that he would complain about when we were doing presentations, we were on the vendor side at that point in time. We would do, or if we were there for somebody else's discussion, things like that. It was the style that he called "show up and throw up." And it was, I'm going to take whatever it is I know, I'm going to give it to you, I'm going to dump it on you. It doesn't matter if you are a big bank, it doesn't matter if you're a three-person startup, it doesn't matter if you have a PhD in data, it doesn't matter if you have a high school diploma. I'm going to tell you the exact same thing in the exact same way, and I'm just going to vomit everything that I know onto you, and then I'm gonna like walk away from this. And it's, it's the absolute kind of worst way to do that because all you're doing is again, you're just giving data, you're giving numbers, you're giving, um, whatever it is that you have brought with you in your case of tools today. And none of that is necessarily relevant, or it's not relevant to your specific audience in the way that it needs to be. And so that's a thing that I always kind of keep with me. It's a, it's a joke that I laugh about sometimes when I'm at a conference or in a presentation, and you're like, "Aha, we have found ourselves in a show up and throw up situation." So you're giving this audience information, but you're not telling the story. You're not customizing what you're talking about. And again, these things could be, they can't, they're probably factually correct, but they're not getting to the right end result. And so when we think about it, like those times that you've been in a presentation, or you can talk to somebody, you've had this screen full of numbers, and you find yourself, your eyes, because you're a data person, and that's what you do, right? You go to these numbers, and you're trying to figure out the pattern that they're talking about. Um, you realize they haven't given you units. So now you're trying to figure out exactly what they're looking at, what they're referencing here. Something's mislabeled. So I get distracted by the fact that this was the label from two slides ago, when they copied and pasted the slide, they didn't update it. All of those sorts of things take you out of that moment, and they take you out of the thing that that presenter should be guiding you towards, right? And so all of your, your tools, your Excel, your presentations, any of those things where it should be in support of that story that you're telling, and it should allow people to walk on that journey, not cause them to become distracted from what you're actually trying to tell them, and the experience that you're having with the data yourself, right? So if I, if I put people, if I put too much up there, if I put too many things there, then those folks are going to get distracted. They're going to start trying to do that data analysis themselves rather than listening to what I, as the presenter, am saying. And so I think that biggest mistake is allowing the data to hijack the story and the explanation you're trying to give about what that data is actually telling us.

Oh man, I have to say, certainly you had to sit through some like 100-slide presentations. You're like, "Well, what is this?" So I do like the idea of, uh, "show up and throw up," even though it's a kind of a revolting phrase. I hope no one does that literally in a meeting. Not a good song. Um, but, uh, so, uh, you talked about, um, uh, not tailoring things to an audience there. So you said like, um, your boss, he was talking about giving the same presentation to big bankers, to a three-person startup. Just in financial services, can you just give me an idea of like, who the different audiences are that you're speaking to?

Yeah, I think there's a couple of different audiences that, at least for, for my role, primarily that I look at. So if I think about, and I use this definitely to tailor, kind of the message that I'm giving as well, because I want to make sure I do the things that matter to them, right? So part of the audience, uh, scope that I have are other data practitioners. So for those folks in the work that we currently do, I'm generally justifying, I'm explaining a model, I'm helping them understand what we've built, what we're leveraging. And for them, it's a lot of, "Do I believe the data behind this?" Right? So I'm probably going to have things like KS and Genie statistics, I'm going to have lift charts, I'm going to have maybe a heat map of the performance of this model and the work that we're doing there. So with a, with a data audience, I'm often grounding them on the contents, and then I'm going to take those folks and then I'm going to help them believe in the power and the efficacy of the solution that I'm proposing. So I often view that data science team leads and data, those are probably where I'm proving myself and proving the, the value of what we have brought to the table, what we're trying to get them to leverage here. If my primary audience is a business team, so these are the folks who are creating the strategies, who are not just evaluating our work, but they are then going to leverage it in their decisioning. In particular, I work with a lot of lenders currently, and so they're going to be using our data to offer a credit card, a personal loan, one of those products that a consumer will now receive and take on. And so for that team, I'm thinking a lot about dollar signs, I'm thinking about percentages, right? Because what they are going to largely do is say, "I trust the data team has evaluated that this model is going to create differentiation and that there's lift going to be there. I can take that for granted now because that's someone else's to, to figure out. Now that I'm on the business side of things, I need to be thinking about what does that strategy look like?" So that's where I might take that same heat map that I presented to the data science team, but now I'm going to create some carve-outs and say, "If you are to approve this segment of the population that you don't currently approve, if you were to offer an extra thousand dollars of credit limit, for example, things like that." I can start to get to business outcomes for that team to say, "We think we can drive X amount of revenue for you this year because you can leverage our our solutions and our tools successfully in your strategies." And so those are kind of the two main groups that I work with currently: the data practitioners and then the, the business teams. But there are certainly many, many other audiences, right? I do a lot of discussions internally, helping support our own team as they are getting educated in the credit space, in the work that we do. And so there's a bit of a different lens that we use there as well. Something like this, right? Where it's largely going to be, you know, Richie and I interacting, and we've got folks who will be dropping questions in the comments and things like that. Um, I would never, in a five-person meeting, just speak extemporaneously for five minutes and kind of keep going, right? So the level of engagement that you have with the audience is something that's like incredibly important to think about how you customize your message, how you, you do that work, right? Because in my general presentations, I'm going to pause, kind of every 30, 60 seconds, every slide or two, and give people space to react, space to think about things. Whereas in a meeting like this, you know, we're kind of feeding in a one-direction way. And so there's all sorts of different ways to think about that audience span and what they care about, how you engage with them. But it is something that you want to be very, very deliberate and how you consider and how you target the work that you're doing.

That's really interesting. The idea that you have to tailor your story, um, to the size of the audience as well, and not just, like, who, what sort of demographic you're talking to and what they care about. So there's lots of things to think about there. Um, I'm actually a bit curious about your process for how you come up with, um, the story. Because you talked about, well, you need to have some kind of narrative around your data. So suppose you start with your results, and you've got an audience that you know, who you're speaking to, how do you turn that into a story? Like, what's step one there?

Yeah, I think part of it is, it is just being very deliberate about the, "What is the answer I'm trying to get them to?" And in my current role, it's, "I'd like you to use the data that we have available." Um, and then, "Who's the audience?" And kind of, "How do they care about that?" Again, I might be proving this data, I might be, um, convincing people of how they can best leverage that in their, in their decisions. And then you start to walk through having gotten to that answer and knowing what that looks like and what they care about. And I, I always say that that's, you know, the most important thing to be thinking about is what does that audience care about? Because there are many great sales people in the world, I am not one of them. I'm not going to get you to care about something that you, you don't walk in with any interest in, right? So how do I phrase this so that this is the thing that's relevant for you? It's revenue, it's additional audience, things like that. You figure out what that answer is. We start from what we need them to know. Um, so, and again, in my particular world, as a credit reporting agency, there are things that I know everybody needs to be grounded on because the regulators will ask them about that, right? That's a little bit of a secondary audience, that's one degree removed from the folks I'm talking to. But I have to lay a foundation that ticks all of these boxes so that they will immediately, as table stakes, feel comfortable that we have done our due diligence, that we meet the needs there with our data. And then we evolve from there to getting to the steps that they need to understand to be able to do their work. And I use that word "steps" very, very deliberately. Um, I think that there's, in building that story, you have to take these tiny steps as you evolve, uh, where you're going and how to get them to that end answer. So I, I laugh sometimes. I think about, I had this particularly very bad professor in college who I did my undergraduate in mathematics. And so a lot of proofs for those four years of undergrad. And he had this really infuriating habit of doing two whiteboards full of proofs, and then you get to some place interesting, and then he says, "It is trivial," and gets to the end. And you're like, "It is not trivial! It is in no way trivial how you made the leap." That for him was, you know, a tiny gap between here and there. And for the rest of us, I mean, he might have just jumped the Grand Canyon for all we cared, right? Like that, that idea that you have to remember, as this practitioner, you've probably been immersed in this data and in this world for sometimes days, sometimes weeks, in, in my case, my current role, years, right? And that idea of how you get from one point to another, you can't assume that your audience is making those same pieces of movement. And so I will always, always defer. I know, slide presentation or an any sort of a discussion there, take these small steps and look, pay attention to the audience. Look at them. Are folks nodding? Are they kind of like, "Yep, I get it, I get it?" Right? And it's that, "Keep them nodding with you." And I literally say to my clients, you know, "If I do my job right in the next hour, a lot of what you will tell me is, 'Yep, makes sense, let's keep going.'" And that's a sign that they have followed where we are, that we're continuing along that progress. And so I always say, right, like, "Do more slides, do more slides that do one thing well on every single slide." And then use that to walk. You can speed up through those, you can skip the slide if you want. But if you've got four steps that you've taken on a slide, and you lose your audience in that space, it's really difficult to recover them later. It's difficult to kind of handle that in the moment as you then have to take yourself out of storytelling mode and really dive in and try to walk them through an analysis point at that time, which is a very different experience for them, for you. And so I think that, um, figuring out that right version of this. So I'm from the southern US, and we are, I think, by default, kind of storytelling people. And so figuring out that right cadence. A lot of people like to use the hero's journey. That's a really well-defined concept of storytelling, certainly. But figuring out those building blocks that make sense for your data, for your outcome, incredibly, incredibly important. And then take those little bite-sized pieces, again, allowing that audience to be along the journey with you to say, "Yep, I get it, I get it." Is is really, really important to being successful. And you'd rather them thoroughly understand everything then, kind of challenge them in that moment to make a leap that maybe they're not comfortable with, maybe they just haven't gotten there yet.

I have to say, I've definitely been there with that math professor, just going, "Oh, this is trivial," and then lost. Um, maybe it's a thing coming to math professors, but I, I suspect it's, uh, broader than that. And no offense to any math professors, but please don't do that to your students. It's so beautiful. No, I think it's a very real point that, um, it's much better to go a bit too slowly and have people sort of go, "Well, yeah, I know all this, it's too obvious," then the other direction when you just, everyone gets stuck and they lose interest.

Yeah, because I think the thing is, if we think about it from a psychology, and I'll always say, not that psychology isn't a science, but I say like, data science is a balance of the true science and the, the work that you're doing there, and the psychology of bringing an audience along with you, right? You can't separate the data from the people, or you're not going to have a cohesive message and a cohesive experience for them. And when you think about that, right, if someone has saying, "Yep, I get it, I get it," they might think I'm moving a little bit too slowly, but they believe what I'm saying. And if I, at some point in time, make this leap that they're not expecting, especially again, I'm currently on the vendor side, I'm working with a lender or somebody who's in the credit space, that opens up the opportunity for them to say, "Is she hiding something?" And that's the last thing that I want when someone is a partner and we're trying to help them understand how to best use our data. If I go too fast, then it opens up a lot of surface area of what, what aren't we talking about, right? And why aren't we talking about it? Um, and we work in international credit data. There are a million questions that people have about the sorts of work that we do, the bureaus that we partner with, all of these things. And so I have to be able to build that trust. And that's one of the key ways to say, "Yes, I trust that she's thought about this, that she's done the work that she needs to, and I'm not creating that opportunity for someone to, to question, are we not paying attention to other things in the same way that now I wasn't paying attention to your experience of getting from point A to point B in this conversation?"

Ah, really. So a lot of this is about good communication is going to help with trust with whoever you're talking to. And if you know, you don't communicate well, then they don't trust you, and probably that's, uh, everything out the window then.

Yeah, hugely important. And I'll, I'll take one step back just to let us, let the audience on part of the thing that Nova Credit is best known for is is transporting international credit data, right? If someone has moved to the U.S. from Nigeria, for example, we can help them pull their Nigerian credit history and use it with a U.S. lender. Somebody moves from the U.S. to Canada, we can help them do the same thing and then take that information. And so that puts that lender in a position of needing to trust us as a representative of about 20 credit bureaus internationally currently, right? And so that's a big level of trust in a place where they're accustomed to, in most countries, working with two or three credit bureaus that they have known and have vetted over years, decades, right? Now they have to say, "I trust you on behalf of all these places, all this data. I don't understand all of these companies. I'm not familiar with, right?" And so, um, not to have to plug the company that I work for, but to give a little bit of the context on how they have to trust us, because we really are taking a relationship that's so core to being a lender to the business that they've done, to all that they have established in this time, and saying, "We're going to be your ambassador, we're going to be your data translator from these quite literally foreign data stores and pieces, and we're going to make this something that you understand, that you can leverage." And so if they don't trust us in the work that we're doing to get from these other data stores, then there's just no conversation to be had, no business to be done. So absolutely, that trust is incredibly important. Because as a data practitioner, again, there's an infinite amount of data behind the scenes, you've got to be directing them in the ways that they need to go and the things that they need to understand. And if I don't trust that you're finding the right insights, that you're finding the right message, that you're doing all of that due diligence, then I've got to go do that myself. I can't trust that relationship. So absolutely not something that we talk about, not something that I'd really thought about necessarily in depth before we started this chat, but it is incredibly important to the reason that we need to do storytelling and cover those bases. Absolutely.

Now, um, earlier on, you mentioned the idea of the hero's journey. So this is like a standard, uh, movie structure where someone, uh, overcomes some obstacles, defeats the bad guys in the end. And you've got other, sort of, a similar, like, stories like in a romantic comedy, you've got two people who hate each other, eventually fall in love. So I'm wondering, what are these sort of equivalent stories in financial services? Like, what's, what's your hero's journey there?

Um, I, I think for me, and again, I, I'll think about the very specific roles I've had. I've been on both the lender side, I've been on the data vendor side of things. And so again, every audience, every piece of work that we're doing looks a little bit different. But often for the storytelling that I do currently, there's a layer of setting the stage. So that's what I'm going to tell you how many records we have, well, how many of those people went to link when it didn't pay back, what kind of data we had available, right? There's an education and, um, also one step back, a lot of these presentations, while we're doing them live, that presentation is going to be shared to other people within the organization that I'm not going to be part of voiceover. So I have to be able to lay that framework in a way that works functionally for both people who are listening to me talk right now, but also people who are going to look at this with no connection to me whatsoever. And so I need to lay that foundation. I need to help, help people understand the very tactical levels of what we're doing. My next build from there is going to be, "So what?" What have I taken all these numbers on? How do I then build on that and say, "This is how you can use it?" And so that's where we start to think about the KS, the Genies, the lift charts, to show that this model is better than that statistical model, or in addition to the model that you're currently using, this is how you can add the work that we're doing. And so it's taking that, that raw data, and we're now kind of leveling that up to a discussion of how that performs from, again, a very statistical perspective, and an understanding of the efficacy of that work. Now that we're looking at it in a multivariate solution. I take from there, and then that's when I get into the business strategy side of things. So it's like, "Okay, if I've convinced you that we did reasonable things with the data, I'm now convinced you that the, the thing that comes out of it is also very reasonable, you understand, and it's intuitive." So what do you do with that? And that's the peak of the hero's journey for me, is to say, "Now that you've taken this model, and you believe me that it's, it's real and it works, then how can you change your strategies? How can you lend more money? How can you decline people that you might have otherwise not approved?" Um, and so from that point, I can out then kind of get to that point of saying, "This is the revenue impact, this is the number of accounts that you can now book, this is the, the thing with the dollar sign," which ultimately is going to be the entirety of that relationship. So kind of building on those layers of insights, again, it's the, the pyramid that we discussed earlier, right? You've got to get from that data up to the wisdom pinnacle of that pyramid. And that's really how I think and how I do most of my storytelling in those spaces, because I do need to cover that range of explaining the data at a very tactical, this level, but I have to get to, "Why do I care?" And then our industry, right? Why you care is dollars, as accounts, it's some sort of a revenue.

That's really interesting because it sounds like the way you're, um, sort of structuring your stories then, it goes from maybe the more sort of methodological side, like, "This is what we've done," and then that leads to, "This is what you can do now. These are the decisions you can make." And that sort of naturally, um, it, it reminds me of when you have a scientific paper, and it starts off with like, um, "Okay, this is background context," and you go through methods, results, conclusions, and that's your action. Is that always the best structure? Because I know sometimes, um, especially if you're speaking to executives, they're like, "Well, you know what, I don't care about any of the methodology. Just like, you know, give me the high-level results. Tell me, tell me what decision I need to make." So do you ever need to change that structure around?

You do sometimes. And it was a great analogy that you said, similar to a scientific paper, because those are almost going to start with an abstract, right? You're going to tell them upfront, "Here's two paragraphs, three paragraphs of what I'm going to tell you in the 30 pages that now follow." Um, so all of my presentations will also have an executive stakeholder slide, right? So the first thing I'm going to tell you is, "We looked at a hundred thousand records on your kind of portfolio, and we can generate 10 million dollars a year a year in revenue for you." Right? That's the thing that I'm going to lead with. And then I'm going to explain to you the story, and that, that build-up that I, uh, mentioned there. And depending upon the audience, right? Like if I'm working a particular senior person, the data and the answer might look very, very similar, but I'm probably going to skip most of steps one and two, right? Where that might be, uh, seven or ten slides for some of, for some of my audience, that might be two slides for an executive audience. And what we're going to do is, instead of the one example of how you could use this in your strategies, but I might use, if it's a data-specific audience, I might now work through three or four different versions of that business case, and those, um, those different outcomes that we're targeting. Because now, instead of looking this overall, I'm going to really focus on what a strategy that increases your approvals might look like. I'm going to really dive into a strategy where you decline folks you're currently approving and save yourself some losses. And I'm going to work through each of those in a little more detail and ensure that we're, we're really kind of punching into all of those places that make up the business case and all of those use cases. So in that structure, I'm almost flipping the whole thing on its head, right? Because I'm really starting and focusing on the top of that mountain, right? Like, "What this is, so what?" And that business outcome. And I'm using a very, very supportive but very thin version of the data and the, the efficacy build, because it's not the thing that that audience is going to care most about.

Okay. Um, so we talked a bit about, um, tailoring things with different audiences. Um, I'm also curious as to different formats. Because you mentioned presentations, but sometimes you have a written report. What can you talk me through, like, all the different kind of formats of communication that you need to understand, and maybe how you might, uh, tailor what you're doing for different formats?

I would say that for myself personally, probably 75, 80% of what I do, and the way that I tend to think about things, is is going to be presentation style. Um, so my role is to be the bridge between our internal data science team, who's doing a lot of the heavy lifting and the building that work, and then our client teams. So I'm a messenger. That's 75% of my role at this point. So a lot of what I do is in PowerPoint, it's in those sorts of presentations where, again, I have to think a little bit about what's happening live. And so in that live moment, perhaps there's an opportunity where I really just want to put a single sentence on the thing and then talk and make them listen to me as I tell that story. But that's not at all effective because I know that presentation is also a leave-behind, and that's going to need to tell the story within the organization in spaces that I'm not a part of. And so I have to view that as an artifact. And that's a really interesting distinction. Um, when you go to a conference, for example, and someone has a presentation, right? Like it might be a single chart, it might, but it's generally designed to be a background to what they're saying. You should be focusing on that person and listening to the story that they're telling. Very different experience if you need that presentation to also be an artifact. And one of the great compliments I've got in my career was walk through a pretty complicated presentation with one of our lending partners, and our lead on that side was like, "I don't have any follow-ups here because I'm just going to take this, package it, and put it in an email internally. There's nothing else that I need that I think, you know, our teams are going to need to understand the work that we've done." And that's, that's a really great compliment to have received. He was thinking about it from the next steps and kind of a very tactical lens, but it means that for the stakeholders he has, again, those kind of second-degree stakeholders for me, he found that presentation to be both be informative in the moment, but also a great artifact to take along. So that's a couple of different ways to think about like a literal presentation, a PowerPoint or Google Slides. Um, Excel is a little bit of a different story, right? Because it is intended to be far more granular in terms of what you're doing and how you're actually doing that analysis. But I encourage everybody, even if you're passing along an Excel, you still need to do some data storytelling within that sphere. So add a readme, add an overview tab to the front of that, explain what you did in words, explain what you found, what you've looked at, right? Because I think there's this tendency for folks to, someone has asked you for a particular query, you go, you run it, you see that it all kind of makes sense, and then you throw it over that fence. But how much more valuable are you to that engagement? How much more valuable are you to that organization to prove that you have kind of looked at this, you've thought about the data, and here are some things that are really important? And that's one of the key things that I think about when I'm looking to promote somebody, or I'm looking to understand how do they understand the data? It's, are they pulling out the same nuggets and insights that I would? Are they finding things that I wouldn't have because they're a really deep subject matter expert, for example? And so thinking about how you do that storytelling in Excel, that's often going to look like that readme, your overview slide. It's going to look at like color coding certain cells where you find an anomaly, where there's something that breaks the pattern that you're looking at. There's a value that's unexpected. Highlighting those, making those really crisp, so that I'm not having to go look at the max of this column and then decide if I think that's appropriate or not as that subject matter expert. You're pulling out that wisdom for me. It's prepping some of the charts and things that, again, if what I'm looking at is a pattern here, I'm going to see that much more quickly if you've created a line chart for me and you've already embedded that in the Excel, as opposed to just taking that and saying, "Hey, there's some, you know, it's, it's mostly monotonically increasing, it's fine." Um, so those sorts of things allow you to, even in something that isn't designed to be a storytelling tool, you're still telling that story, you're still showing your expertise in that data and using that to advance the discussion. Um, and I think that there are opportunities in some other. Those are the two primary ways I know a lot of folks work. But we also have to think about dashboards, right? Even automated reporting opens up that opportunity to think about how you're doing charts, what you're doing is alerts. Um, those are those are really tactical things that as you're entering your data practitioner career, as you're starting off, things that you might not think about as storytelling because it's, it's alerting, it's things that people have asked you to create. But there's, there's a beauty to the way that you do that. To what are they? We're just having a conversation last week about, "What's the first chart I'm going to show on this dashboard?" And I was like, "We've got this wrong." Because what we're doing is showing me that everything looks okay. What I want the first chart to be is, "What are the things that don't look okay?" Right? Like, I want to take for granted that everything works. What I don't want to take for granted is, "Show me the three things that are red or that the team needs to be looking into." Right? And so by even just rearranging where we're putting tiles in this dashboard, we are telling a story. We're starting with this thing that I want to be more important. And so understanding in any of those formats, what is my hierarchy? What is the? And I often do that thing where I say, um, if that person is going to, we all know everyone's multitasking. I know in my presentation that I might not be the most important thing. Um, everyone's distracted, everybody has a life. If I've got one moment to grab that audience member to say, "This is the most important thing," what is the thing that is, what I want their eyeballs to be on? I want their brain space to be on. And so those are the things that you have to figure out how to tell that story, be it in an executive summary, in your presentation, be it in that readme, an overview in an Excel, or in the way that you structure that dashboard or that report. Always figuring out, if I've only got 10 or 15 seconds of someone's brain power, what's the one thing I need them to know?

So many great points there. I have many questions from that because I want to talk about presentations and Excel and dashboards. Um, I, I can't say which one to go to. Let's go and reverse order. We'll go with dashboards first. All right. So, um, you mentioned that, um, in order to tell a good story, you have to think about the order of things. So, and again, I guess it goes back to like, what you're showing and what you're not showing on there. Is, is there any sort of natural dashboard structures or layouts that you think, uh, will enable you to tell good, good data stories or any particular data stories?

That's a great question. I think that it all depends on sort of the purpose and again, I'll keep coming back to that idea of the audience and the use case. Depends on the purpose of the dashboard. So right now, I'm working on some that are really kind of QC targeted. And so for those, when I think about storytelling, the first thing I want my story to be is, "Is there something wrong?" And so everything else is going to get rearranged based on that. Is there something that we want to highlight that's working, like, not quite as expected? There's a shift in the data that we've seen, things like that. And so we arrange everything there to say, "Here are the exceptions." And then from there, follow volumes, all of those sorts of things, right? The things that if everything is up and running in the way that we would hope it to, that we want to be able to take for granted over time. And so I think that in a world where it is a QC, it is a storytelling dashboard, then like that may be the way that you want to structure that. If we're looking at something that's a little bit more, um, over on the engine, we have a daily report on how much volume we're doing or how many approvals we've had, right? Like those are the things that we're going to start with. And we're going to start with dollar signs, right? So what are the dollars that we have approved? Because ultimately, that's what's driving that lending business, for example. And so it's all different versions. You know, there's, you can justify absolutely anything as being the most important, the most relevant. And it is, because the, the ultimate thing that we want people to care about, the thing that we want our audience to think about, is going to differ based on all of the, the goals that we have there and the, the thing that matters most for the audience that we're looking at. I think that's the, there's not, there's not a set answer, but the answer is, think really deeply about who's going to be looking at this, how they're looking at, how much time they're going to spend on it, and you'll find very different answers for the ideal setup for a daily thing that might go out to your entire company versus a monthly overview that goes out to only data practitioners.

That's interesting. Uh, so again, it's back to tailoring the audience. One thing I'm curious about, um, with dashboards, I suppose this applies to Excel as well, is that people are going to be, uh, looking at this without you being present to actually sort of physically tell the story sometimes. And so, do you have any advice for, uh, how to make sure they're getting the right story, even if you're not there?

Yeah, I think that that voiceover and how you figure out to tell the story within your absence, I think is really important. And so again, in Excel, it's a readme, it's an overview, it's in a way that you've kind of structured that. In a dashboard, it's how you do those alerts. So is there a specific thing that then triggers an email to a certain audience? Is this that people are checking in, and there's something that you can kind of post to the dashboard there? So you find those opportunities within the format that makes the most sense to then be able to engage that audience, to feed that information to them, to say, "Hey, here are some things that you need."

To think about here are some things that might be really important to you. And I think there's there's a bit of a thing. And I'll say, look, one of my big pet peeves, and it's a it's a controversial stance to take as someone who has spent 20 years in financial services as a data practitioner, is that often the problem is too many numbers.

So I think that there are, like, I will forever drive me crazy when someone has a y-axis on a chart that has .00 for every single thing, right? So somewhere in your Excel, you've got something out to two digits. It doesn't matter for this axis, and that's what has default gotten pulled in. And you've created this clutter in your visual because you have just allowed the default to kind of happen. You've got data labels for where you, what you really care about is a trend, and is this going up or down, things like that. And so there's there is a a very specific piece of your storytelling that involves not letting the default things happen in the format that you have chosen, right? How do you remove some of that clutter, the data, be it visual, and any of those tools, so that someone is actually focused on the trend that you want them to think about, and the slope of that line, as opposed to you have put 17 labels for each of the values in those months when I don't necessarily care about that, right?

I think about sometimes we'll do something as an index. So how much better is this than some baseline? I might not care about the values, the absolute values for either of those two series of data. What I care about is, is line A higher or lower than line B, right? And again, you're removing some of that thought from the from the audience in those ways. And you're telling a story by saying, you don't need to know what the loss level was this month. You just need to know if the test was higher or lower than that, and by how much, right? And so that is part of that storytelling by saying, we're going to index this and then use use the trend relative to speaking to tell you this is what you need to care about, is the comparison and not the absolute value. And so there's there are very, very thoughtful ways that folks can kind of go through. And I know that DataCamp has done a number of presentations there as well on how you present some of those things to be able to really start to think through again, what is that, what is ultimately the message that I want? And the number is sometimes not it. Sometimes it is the comparison of the number. It's a relative view of that. It's change over time. And those are really important things and really nuances as someone gets further in their career and more sophisticated in their presentation, their engagement with that data to be able to take it from just a 3.4 to a 4.5, uh, to really be thinking about what is, what is the thing I ultimately want my end user to take away from them.

I do like the point about, um, reducing the amount of numbers in that, particularly like how many digits you have. Definitely seen some presentations where it's like, oh yeah, the answer to this is like 3.14159. Yeah, it's like, no, right? Don't curse that much. No, the answer was three. And was it 12 last month or was it 2700 last month? That's what we care about. Uh, absolutely. I suppose, um, so related to that, do you have any tips on how to make, um, like plots better? Um, like how can you tailor your plot for a particular audience or a particular context?

Yeah, I think, and again, I don't want to step on any toes because there's been a lot of good work that DataCamp and others have done. So there's a lot of like great classes and presentations folks can go on specializing the visual side of that. Um, the thing that I will always say is, as I look at a plot, as I look at any piece of information, one, does every single thing I have here matter? And even something as simple as if you have, you know, the the circles or the squares at each data point versus looking at this as a line and and having the eye look at it more continuously. For example, those nuances are really, really small things that most folks aren't going to spend a lot of time thinking about, but in some, it matters a lot, right? So if we think about, you know, kind of how you structure your labels, where you put those visually, it's all about where the eye goes. And and again, we want to not just take all of the data that's available and kind of dump it into a presentation, but we want to think about a lot of, one of the things that I will do is if I have maybe a couple of different ways I'm looking at data and one's, uh, a year one year and then a year later, maybe I'll make those lines the same color because they're both going to go from January to December, let's say. Uh, make one that's the same color, make one of them solid, one of them dashed, so the eye immediately goes, these two things have something in common, as opposed to this other thing that might also be on that same plot, for example.

Really thinking through what if someone doesn't understand any of the words you've put on this page, and they don't understand the data that underlies it, what would their eyes naturally look at? What would they take from a plot that didn't have any axes or things like that to kind of abstract it in that way? I think is a bit, if I'm doing a very, very intense presentation or, um, sometimes, you know, you do want to think very deeply about a specific plot. That's a great way to to take yourself out of this exact moment and kind of remove the numbers, really, from what you're thinking about and say, what would a kid, was looking at this, right? Like, what would they intuitively think was related here? Um, it looks very different, for example, if you've got bars where they're literally disconnected from each other, right? So these are things that might be categories I'm looking at versus a line where you're implying that there's a continuity of time or of your relationship, right? And so if these are seven different product lines, those should probably be bars. Those are things that aren't inherently connected to each other. They don't have an order other than maybe an alphabetical that you've given them, as opposed to a time series which quite literally like does flow from one to another. And that line structure, that usage, certainly, you know, changes that.

So thinking very intentionally about where you use those data labels, how you use your axes to again, take out all the zeros and the millions, that's fine. Like, people will understand that you mean 2 million on that axis. Um, but looking at literally every digit, every piece of information that you have put on this chart and saying, is this valuable? Because what ultimately you want those visuals, those plots to do, is to be as intuitive to your end user as possible. And so everything you can do to remove clutter, that you can do to help that story be as intuitive as you can. Those little details, no one's going to notice each individual decision, but when you look at the before and after side by side, it's a huge, huge difference in your audience's ability to consume that quickly and accurately. Um, that is a pretty amazing. It sort of sounds obvious, but if you just look at the plot you've drawn and see what your eyes are drawn to, that's going to give you a lot of information about whether this is a good thing or not. Um, and now I'm wondering like, do a lot of people just not look at the plots they've drawn? Because sometimes you're like, I think that, you know, in a world, um, that we live in, right? Like Google Sheets or Excel or whatever tool you're using makes a lot of default decisions for you, right? Like after I've highlighted that data, it's going to guess what the best kind of plot and chart is for me and things like that. And so there are a lot of things where if you're working under a deadline and you're working at volume, um, you know, is it the best use of your time to go back and remove the two extra zeros after the decimal place? Sometimes it isn't, right? Sometimes you're trying to keep things moving forward there. And I think that that's a, I don't want to use the word lazy because that's not it, right? People have priorities and you're working on a lot of things. Everyone's juggling a lot of things there. Um, but it is, it is easy sometimes to let the software be smart and make some decisions for you. And if you come in and think a little bit more deeply about some of those things, again, a little bit of the job security as a data practitioner is that we can make better decisions and we can make more targeted decisions for for that software. And I think that's that's a little bit of where that art does lie in the work that we do. Um, absolutely. I can imagine that, yeah, time constraints and rushing do lead to sure not doing perfect work every time. Um, all right, we've got so many amazing audience questions, but I've got one last question for you before we get to those. So, um, we talked about a few different tools, so particularly about the use of like, uh, PowerPoint and Excel and the the Google equivalent, but one thing we haven't mentioned at all, uh, is, uh, Jupiter notebooks. So here at DataCamp, we're big fans. We've got our own Jupiter notebook product in Workspace. And these are brilliant because they let you tell stories by mixing text with with your analysis results. So I'm wondering, uh, about, um, are these, you, notebooks used at all in financial services? Um, and how might they be used?

Absolutely. Um, so, you know, if you've got people who are active practitioners, who are data scientists, who are in that space building models, doing various analyzes, um, those things are just part of kind of the work that we do every day. And I think it's so important because we've probably all been in that space where someone's asked a question about a colleague who's no longer with the organization or has been promoted into a different space and things like that. So you've got to go figure out because a client, somebody in your organization has pulled up this chart that was created three and a half years ago and said, hey, what did we mean when we did this? I have no idea. So now you are going and you're going to start trying to piece some things together. And maybe you find a, maybe there's a README, an overview, which is fantastic, and it has links to things that you need to know. More often than not, there isn't. And so you've got to dig around and find those things, which often leads you to someone else's code. And I think, you know, I, as we were beginning, I saw that we've got a very, very international audience. So we've probably got folks who've had international travel experience, international engagements, um, and you've been in that place where you find yourself not speaking the language and you're using these cues from people's visuals or signs, things like that. And that's what sometimes going through somebody else's code is, right? You find yourself digging through this code trying to make sense of what data was pulled and how to bring this together based on something that made all the sense in the world to a very smart colleague of yours some time ago because they had been immersed in this date, in this problem, and things like that. And so the thing that I say, like when I've done code reviews with various teams and things like that, I almost care less about what your code does than how you've commented it and how you've used, exactly to your point, the Jupiter notebook and those parallels to tell that story. Because if you are being an effective partner to one, your future self, because it's not always just a colleague, and sometimes when you pick this back up in six months and go, I don't know, I worked on this for three days straight and I knew everything about the data set, I knew everything there was to know about that, and then now I don't remember anything because I've moved on 10 projects from there. So your best favorite of both your future self and everybody else in your organization, right? To to use that to tell that story, right? Like I should be able to look at someone's code and follow that they have done data hygiene work here, and that they've done some data selection, and then from there they've gone to a model and then they've created some performance charts afterwards, without actually looking to see which method, what modeling methodology they use, perhaps. Um, and so being able to do that, that's an interesting thing that was in my undergrad, our first computer science class was actually pseudo code. It was something that didn't have a compiler, and so you couldn't become compiler dependent on does it run? You actually had to think about what was the structure, what were the pieces I was trying to do in this made-up language. And so that idea of what was my thought progression of kind of how I went through this code, how did I, again, code is another way of doing data storytelling, right? And so all of those pieces, explaining the choices you made, because there's a very logical reason that you excluded these 101 records, but when a client comes back and says, why is my total account 101 lower? I need to know kind of the choices that that data practitioner and that that data scientist might have made to be able to explain that as part of my story to my eagle-eyed client who notices that, you know, we had a mismatch or they didn't have the appropriate data and so we had to exclude them from the analysis, for example. That's something that the day you make that decision, it's a great decision, it's an appropriate one, it's a wise choice to have made, but you have to also be able to justify that. And so, um, yeah, and I will, I will say, you know, as you become more senior career, like the beauty of how you comment, how you think through how you structure that code, really is how you teach the people who come after you, right? So as you've become a senior data scientist or a manager, folks are going to look at those things and they're going to learn from you, either good habits or bad habits. They're going to learn either way, what that means, how how you've looked at that, the choices that you've made. And that's part of where we are, right? Like it's that, as I was saying there in the beginning, right? You go from here and then it's trivial to get to there. It's not trivial to anybody else who isn't in that moment. And so you've got to use that format, in this case, you're a Jupiter notebook, to help help everybody else who isn't in your moment in that day understand exactly kind of those choices you've made. And if you do it right, like working with a good client, you're going to have them going, yep, yep, that makes sense, that makes sense, and they follow you all the way through there, even without your leading them. Absolutely. Uh, so I, I've got the notebooks I used and, uh, yeah, certainly I found that future me is always stupid and doesn't remember what I've done. And, uh, yeah, having some good comments, uh, uh, is and a bit of narrative is, uh, very helpful. All right, um, we have many, many great audience questions to go to those in a second. Before that, I just want to say we've got three more, uh, webinars this week. So tomorrow, you've got me talking about, uh, ChatGBT Enterprise. On Thursday, you've got me again doing a code along on an introduction to GBT and Whisper. So we're gonna be using the OpenAI API. And then on Friday, uh, we got, uh, Martina doing, uh, another code along, which can be exploratory data analysis in spreadsheets. That's a good one to get started if you've not done much data analysis before, a nice recession. And with that, let's go to audience questions. Uh, so first question comes from Alex. How do you tackle having two or more different audiences in the same presentation?

Oh, that's an interesting question. So I think you've got to, you've got to think about how you balance those, which is a generic thing to say and not terribly helpful. Um, and often what I'll do is literally address each of those audiences, right? So in a world, you know, in my world, like I know that there are going to be people who are multitasking or checking emails, things like that. I'm going to give them audio and visual cues of when they need to check back in. So there are times where I like might literally say, as that build that I mentioned earlier, as I'm talking about the data and I'm going to be like, all right, are data nerds, like this is, we're gonna be in this for a few minutes and we talk through those kinds of things. And then with a divider slide in my presentation, I will say, and now we're going to transition into the business results. So part of that nuance, and those teams have generally worked together, they know each other, right? I'm going to give them those audio and visual cues of, this is the part that probably you're not going to care about, check your email, catch up on Slack, do what you need to do alongside, but I'm going to remind you when I need you to come back in. And this is one of those, like small tricks that if you've been a consultant for a while, you know, um, if I've got something where I need to re-engage somebody who I suspect is probably multitasking or doing whatever, you always start with using their name, so that you have re-engaged them, and then you ask them the question. So if I know Richie is working on something else and I've been, you know, deep into this, I will then like, kind of make that transition of, so Richie, I really am interested in kind of your thoughts as we transition into this results section. And so I have, rather than asking his name at the end, where he's now missed whatever I said, now he's embarrassed and he's having to say, I'm sorry, I was multitasking, could you repeat that? Just leading with somebody's name, you bait them back in, you warn them. Your co-workers will love it because they appreciate the fact that you make them not look, um, silly or distracted. But even something as simple as that is to say, all right, for the business team, we're gonna jump back in, we're going to look at some of these examples. And so you cue them back in that, and you give them that permission to say, I know that there are parts of this, I might lead when I give the agenda early on, you know, I'm gonna need you to bear with me through these couple of things, we're going to focus here for a bit, and then we'll make that transition. So as part of your storytelling, if you think about your presentation with an agenda slide, with some sort of an index, where you cue for people, acknowledge that there are multiple audiences that you're engaging, and then help them understand where they need to be and where they, um, we're going to have a little more permission in the way to be distracted. I think you, you kind of handhold each of them in a specific way. Feeling very glad I was concentrated that because it's like getting called out. I was multitasking, honest. But no, I do like that tip. Use the person's name at the start of the sentence, not at the end. Uh, just a simple change is going to make a little difference. All right, next question comes from, uh, Ferris, saying, what are the recommended methods for, uh, supporting evidence in the context of data storytelling?

I think we've all been there with like something like data cleaning where no one really cares, but you sort of feel like you ought to include it somewhere, right? So for me, the choice that I generally will make, I believe the appendix is your best friend. Here's the thing. I might have a, um, and Capital One, we were very, very good about saying that this is the way that life should be. Um, you might have a 10-slide presentation that's going somewhere, and it could have literally a 50-page appendix because what you are saying there is, if somebody asks this question, I've got it, right? I've thought about that, I've done my work. There's an again, infinite amount of stuff that's happened behind the curtain, and I can kind of expect these are the sorts of things that people might ask. Um, when they don't, it's there, I've got that stored, I have that information available, but I haven't built it into my story directly. And so I'm not distracting my story. If they're willing to take those things for granted. So, and one of the presentations that I currently give, um, it's interesting to see where an audience might go, but part of what I do is talk about how we differentiate risk, the comparison of two models. If an audience is fairly savvy, they might then start to say, well, what are the sizes of those swap sets? And so in the appendix, I have population size for each of these populations. Not everybody gets there. And so my story is built on the actual losses and the risk of this population and what that means in a business case. But I always have in the appendix as population sizes. So when somebody asks, well, are these two models really, you know, quite orthogonal, are they very well aligned and correlated? I can go to there. But I know that not everybody's going to get there, not everyone's going to think about that in the moment. And so that's my recommendations. There's a lot of data hygiene, there's a lot of work that goes, some audiences are going to take that for granted, some audiences may want to dive in and pick at that. And so I think be very, very strategic and be generous with what you put into that appendix so that you have those resources, you have your proof points, and you're prepared for that, but you don't assume that the story is going to involve a lot of that minutia. That's great advice. Yeah, I do like the use of, uh, appendices. It feels like in video games where you have a main quest and then you've got side quests which are optional. Sure. Um, all right, uh, next question comes from, uh, Abdulaziz, saying, how granular should your visualizations and terminology be when speaking to an audience that may have, uh, senior level workers and entry-level analysts?

Yeah, it's, um, it's a tricky question. And so I think you have to think about what is the purpose of each of those different audiences in that conversation, right? So there are times where I might have a quite senior audience, there are more junior folks there who are intended to be there to kind of listen to learn, to be a part of, uh, the experience, but not necessarily, they're not my primary stakeholders. And so in that case, I'm going to talk to that more senior audience. That's the level I'm going to hold it to, and then I'm happy to address questions afterwards to help fill in some of those gaps, right? But if my primary audience, the primary outcome I'm looking for, is with that more senior group, then that's where I'm going to bring it, and that's where I'm going to target all of my communications. Um, sometimes you may have more senior folks who are in the audience, and your your focus is actually on, I need to explain all the pieces of this data hygiene things like that. That might be a place where I employ that tool, where every few minutes I'm going to have a nugget that I think is actually more relevant to my senior audience. And so I will then say, you know, after we've gone through all of this data hygiene, and I'll say, yeah, so the, the ultimate takeaway from this that I think all of us will care about, give it a little bit of a preference, kind of, you know, remind everybody that we're going to pull them back into this, is that we took seven steps, we removed 101 records from this because of that, and then we've moved on with the rest of the analysis, right? So why don't we took out those 101 records? My entry-level folks might care about, my senior stakeholders really only care the number's 101, let's keep going. And so using those, those tools to pull people in and out as they may need to be, I think is really important. Okay, so, uh, we're on the hour now. Uh, there is still so many more good questions. We're going to continue on for five minutes, uh, if you do need to jump, then please do catch up with the rest of the questions on the recording, uh, because, uh, yeah, there's a lot of great stuff here. Um, all right, and thank you to everybody who do alluding to drop off. I appreciate everybody coming and listening to us for a while. Absolutely. Uh, so, uh, this one comes from, uh, Harinivas, saying, what are the key elements of a compelling data story in the context of financial services?

I think we covered some of this, but you wanna, yeah, yeah. I mean, I think, um, it does depend on the audience, and I know I beat that drum a lot, but I think it's really, really important. Um, ultimately, I think that there's, there is proof and there's foundational buy-in for the data, the methodology, the techniques that you've used, because in my spaces, in financial services, there's a bit of an assumption that everyone is somewhat data savvy, right? Everyone grew up in Excel, everyone grew up doing this work. And so for most of the audiences that I'm engaged with, there is a base level of the work that they've done, how they've kind of come up in the industry, those sorts of things, that even, you know, the degrees that folks have in the colleges that they've gone to. And so you do kind of take that for granted a bit in thinking through how that person is going to approach it. And so in in those spaces, and particularly again, since I'm on the vendor side of things, I need to prove that I've done my work. So there's a proof point kind of level to that. And then you quickly want to get into those business outcomes. And so that looks a little bit different for everybody, loss reduction, increased number of accounts, uh, changes in the the terms that somebody was able to offer to that particular customer. But I think that's more often than not, if I can get to a dollar sign, if I can get to a revenue, that's not unique to financial services necessarily, but the ways that you kind of build that understanding that in other industries, you might not do quite as much, uh, proofing in those ways because folks may want to sort of take that for granted if that's not their background, you do have to think about that in a different way where you expect that everybody in that audience was a practitioner at some point in time in their career. Okay, and the next question comes from Marina, saying, are there any best practices for designing dashboards that cater to both practitioners and stakeholders effectively? We did talk a little bit about industrial board design, but yeah, do you want to, uh, yeah.

I think it's figuring out how you look at both from a very granular and the so what of this dashboard, right? So I think, you know, in some dashboards, you have that functionality to be able to leave comments to say, hey, I checked in on this on this date, and everything looks good. Here's what underlies that. Um, that ability to, and there, there are a bunch of different dashboard tools. I don't make any assumptions about what anybody's using or what capabilities it has, but thinking through the tool that you have been given by your organization and that you're empowered to use, and thinking about what are the ways that I can voice over, right? So if I think about it in a slide presentation, I've got a headline and then I've got data below that. Think about your dashboards in a very similar way. How do you create those headlines? And again, sometimes that's the organization of it, sometimes it's literally that you can create a version of a headline, sometimes it's that you have an automated alert that sits at the top. So think about that. Think about what's the headline? What's the first thing I want people to know? How do I, how do I add those in insights? And then think about your tooling and sort of what that, what that looks like for you, because it will vary a little bit based on what the standards are within your organization, how that's set up, what your specific tool is. But it just seemed like the answer is always coming back to, you need to think about what does your audience care about most, and then make sure that's the first thing they see. That's right, because it's got to be something that I'm personally invested in, that I'm, you know, if you're not, if I'm not giving you my brain power, then it doesn't, doesn't matter, right? If I don't look at it, it doesn't matter what, what somebody's written or what they've, uh, what good work has gone behind scenes. Um, all right, uh, I think we'll do maybe two more questions. Uh, so, uh, this one comes from Aldo, saying, um, sometimes I overcome my code and the results informative, it takes way more time than the coding. Any advice on finding a balance?

It's a tricky one because I, I will almost always say, like, it's almost impossible to over comment code. And I do know that, you know, it does take time, it is something to be thoughtful about. But, um, what's the balance? I think code reviews are a great way. Um, so I know Faiza, who's a former employee of mine, is on here as well. Um, one of the things that our team used to do would be side-by-side coding. So when you're doing those pair coding sessions, when you give that to somebody else, I think that it's really valuable to have those trusted folks within your organization to say, hey, this is not necessarily part of the code overview, whatever, help me be better, right? So who's that trusted partner? Your manager or somebody else, like actually have them kind of go through your code with you and say, yeah, these are the things that I think are probably a little extraneous. And I think that that, that partnership, having the right mentor, having a trusted peer, those things are incredibly valuable. And so we could kind of come up with some rules of thumb and things like that. The reality is, your space, your code is going to look different, and what's needed for the next generation, the next person who picks that up, those are going to look very different. So I would say seek out that person, seek out that mentor, um, that person who you know will say, hey, like, let's just spend an hour and kind of walk through this, help me understand what isn't helpful for you, what might look a little bit different, and really just do that in a live way with with an actual person and focus on just that for a few moments. Um, excellent. Um, and the last question is, uh, so we kind of missed out a bit on talking about different plot types. So, um, there's another one from, uh, Haravina, saying, are there any specific database techniques that work particularly well in financial data storage telling? Now, I know what DataCamp, like, I went through our dashboards, like almost everything we do is like, it's a time series plot or a bar plot, and that's like, almost everything we do from a business perspective. I, I think it really depends upon what your, what you're trying to communicate. So like, I, in particular, right now, and one of the products that we offer, I use a lot of heat maps because I'm comparing two models and how they perform. And so when I'm looking at those two, what's the independent and then the joint performance of two models, heat map is a really effective visual to tool to be able to use to be able to again, draw the eye to an instinctual, ah, this is a diagonal instead of a vertical or a horizontal, oops, wrong hand gestures, about a horizontal or a vertical, uh, view of these things, right? And so I think that there's, you know, when we think about like financial services, there are specific outcomes that people care about that may be different than other industries, but ultimately the advice will always be the same, which is, what is the ultimate story that you're trying to tell? And how do you kind of, you know, lead your folks there? So again, if there weren't axes, if there weren't words on this, what would my eye tell me about this particular plot? That continuity or columns that are disjointed, the heat map and the coloring that we're looking at there, all of those things then start to allow your audience to take certain things for granted, to trust their instincts on that, as opposed to getting into those numbers individually. So I think that's the, it's not necessarily specific to financial services, it is just part of kind of how, um, how you tell a very intuitive data story. Excellent. And I have to say, heat maps are one of those underrated kind of visualizations. Good stuff. Yeah, absolutely. Um, all right, brilliant. Uh, we shall wrap up there. So, uh, thank you so much, Nikki. I felt like we could have carried on talking about this for hours. It's just such an interesting topic. Uh, uh, thank you to Rhys for moderating. Thank you to everyone who asked a question. Sorry, we didn't get to all the audience questions. There you had some great ideas. And thank you to everyone who showed up. Uh, I hope to see you again tomorrow, Thursday, Friday, and in future events. We've got the Radar conference coming up on September the 28th as well. We've got lots of amazing speakers for that. So please do, uh, go to the DataCamp site and register for that. See you all again in future.