Episode 32 | Data visualization with Alli Torban

The Assurance Show
The Assurance Show
Episode 32 | Data visualization with Alli Torban
/

 

Summary

 

Alli is an information design consultant and host of the Data Viz Today podcast.

We talk about a range of data visualization matters, including 3 key mistakes that people make when starting out in dataviz (and solutions to those).

You can reach out to Alli via her website.

 

Links

 

 

Transcript

Narrator: 

Welcome to the assurance show. This podcast is for internal auditors and performance auditors. We discuss risk and data focused ideas that are relevant to assurance professionals. Your hosts are Conor McGarrity and Yusuf Moolla.

Yusuf: 

Today we have Alli Torban on the show. Alli is the host of the Data Viz Today podcast, which is a really popular data visualization podcast, and Alli is an information designer based in Washington, DC. There’s a movement, if you want to call it that, towards better use of data visualization within audit. For audit field work, risk assessments, risk response, but then also for producing visuals that are audience ready. And so really keen to have a discussion about that with you today. But maybe we just start off with talking about your journey into dataviz.

Alli: 

Yeah, thanks so much for having me on. I started as a math major, and then I was an analyst for various government clients, right out of college. And then during my maternity leave, I was trying to figure out, what do I actually like to do? And through my love of maps, I discovered that I really loved data visualization. And so I didn’t really know much about dataviz, so I thought I could start a podcast and interview the best of the best. And I’ve been doing it for over two years now. And I got a job as the date of his designer, and now I’m a full-time information design consultant here in the DC area, but I work with clients all over the world, even Australia. And, that’s kind of where I am today.

Yusuf: 

You talked about what brought you into dataviz. What is it that kept you in the dataviz field?

Alli: 

I really love how each problem is so unique. Each of my clients has a very unique dataset and a unique audience and a unique goal. You can have the exact same dataset and two different people, and the graphic can be completely different. So I just liked the variety of creating visuals and I’m a very visual learner myself. So I really appreciate the service that data visualization provides to the data world.

Yusuf: 

Okay. And you spoke about audience there and that’s something that we learned a lot about from you originally, when we started working together, but, there’s a lot of discussion at the moment around better understanding the audience and not just putting dataviz out there. What’s been your experience with exploring the needs of audiences and how important is that?

Alli: 

Yeah, it’s super important. There’s a few main mistakes that I see people making a lot. And the first one is about audience. When you first start, you have to first define what your measure of success is. You have to make sure both you and your client are pointing towards the same target. And in order to know what your target is, you have to know who your audience is. And in order to do that, you have to ask, who is it for? And then someone might say my manager or HR or something like that. And that’s fine. But then I really like to ask what are three words that you would use to describe your audience? Because that starts revealing a little bit more about who they are. So they might say something like they’re very busy or they’re stressed or they’re non-technical and all these words give you hints about what your visualization is going to have to do. And also how long they’re going to spend on the graphic. That is a super important question that you have to ask, because if they’re going to spend a few seconds because they’re very busy or stressed, or if it’s something I’m going to be looking at this every single day for hours, if it’s an analyst, then that’s something very important you need to know about your audience as well before you even start designing anything. Another very important question is what’s one thing you really want your audience to take away. And a lot of times it’s hard for people to just take it down to one thing. But I like to push it so that they pick one thing so we can aim at that. And then if more things come up, then we can assess later. So really who’s this for, a few words to describe them, how long they’re going to spend with the graphic. And then what’s one thing you want to take away. And then when you have that information, I like to summarize it. And that serves as your success criteria. If you ask those questions, you might get something like, we need a technical analyst to be able to drill down and get XYZ out of this graphic every single day. Or you might have, we need our HR department to see X pattern at a glance in a non jargon-y way. And you can see how having those criteria set beforehand can really change the direction of your visualization.

Conor: 

Coming to that selection of criteria, is that an iterative process that you have to have multiple conversations before you can arrive at that statement of success and what that looks like?

Alli: 

I think so. That has been my experience. It’s very rare that I ask those questions and people know the answers right off the bat. And I’m talking to two people like I was talking to you too. You guys might not agree on who, who you’re talking to or, you might have to go back and think about it yourselves and then come back to me. I’m in a situation right now where we decided on all this information and then I created the prototype. And they went back and showed it to more people and they realized actually, we are talking to someone different now, I think now that we’ve seen it, so it’s very iterative. I don’t think people should beat themselves up for having to go back to the drawing board or not having the answer immediately. It is hard. It’s a hard thing, but, it’ll be even harder if you skip it.

Yusuf: 

What that then potentially means is that if you are creating a dataviz, you want to go through that iterative process and not try to get it a hundred percent correct the first time, because you might need to change it anyway. So not spending all of that time, going through all of the individual details to get it to that sort of perfect state. Is that fair?

Alli: 

Yeah, I think it is fair. That brings me to my next big mistake people make is that they use the default chart type. You have to think a little bit more. So once you figure that information out, or at least you have a good starting point, then I would say to go back to your data and start exploring your data, because that’s a huge part of your data visualization. Sometimes you have an idea of how you want to visualize something, but your data doesn’t really support that well. and what a lot of people do, is they have their data, in Excel usually, and they just go into the chart options in Excel and they pick something and, you know, and then they call it a day. But, the problem is a really good example of this, that I see a lot is like, you create a pie chart because you want to show the proportions of a category in your data, right? It’s fine, you know, it works, but then you start thinking, well, actually I want to show how the proportions changed from 2019 to 2020. So then you put two pie charts right next to each other. But then if you think about your audience, they think, okay, so category A in 2019 was this wedge. And then now I’m comparing this wedge to this wedge, and then I’m comparing wedge B to wedge B over here and they’re just going back and forth. And then you realize actually the pie chart wasn’t the best choice for this. So what I like to tell people to do. You have your success criteria, then you’re going to think about the data relationship that you need to show. And I have a really, great resource that I always go to. It’s the Financial Times visual vocabulary. It’s free to download. They have this really nice graphic where there’s about 60 different charts and they’re all grouped by the data relationship, which is like distribution, correlation, magnitude, changes over time. So think about what you’re actually trying to show. And then choose a chart type based on that and experiment. So what I like to do is I go to, like, I want to show for that example with the pie charts. Okay. We’re actually showing a change over time. So then I would go to the Financial Times visual vocabulary and look on the column for change over time. And they have a whole bunch of suggested chart types. And then I pick one plug in my data. And try it out. Maybe that doesn’t quite work or maybe it’s not showing exactly what I want to do. And then I’ll try something else in there. So it is very iterative and you’re always thinking, okay, this chart type is this showing what I want it to show to my particular audience in the way that I want it to. And then sometimes it happens where you do that and you realize, maybe we’re actually talking to this other audience and then you go back and forth. So you hop in between it, but you do need to set a success criteria because otherwise you’re just going to create a default chart, and you’re not going to know whether it’s succeeding or not, because you didn’t set a success criteria in the first place. So success criteria, chart type based on your data relationship. And you’re off to a great start.

Yusuf: 

So the first mistake was not understanding your audience. The second mistake was choosing a default chart type. What’s the third common mistake that people make?

Alli: 

Feeling like you only can use one chart. A lot of times, you guys know this, you’ve created a chart, a very complex chart with your analysis, and you’re like neck deep in this analysis. And you think, well, wow, this is way too complex for me to give my client. But I know that it’s too complex for them, but how am I supposed to dumb this down, it’s so complex, to just one chart. It needs to be simple for them, but you know, one chart’s not going to do, but that’s a false choice. You can have multiple charts. a lot of people are scared to, or maybe not scared, but just don’t know that it’s okay to use more than one chart. And, it’s hard to know which charts to use or how to sequence charts, but Stephanie Evergreen, she’s a prominent dataviz practitioner and trainer, she has this method that I’ve used a lot and it works really well. She says, sequence your charts by answering three questions. What, so what, and now what. What; what do we have, what are we seeing? So what; what does this mean to me or to you? And then now what; some sort of chart suggesting something. And this works really well. You don’t have to dumb down all your information into one simple chart. You can sequence charts. You can have charts that work together and it’s very effective.

Yusuf: 

And that might make it easier to understand. if you do split it up between those three, then it’s easier for people to ask and answer those questions, sequentially.

Alli: 

Yeah, and it’s really great. If you have to take your report and put it into a presentation because that sequences really well and you can be presenting a what chart, and a so watch chart and then a now what chart. And it really has a nice cadence to it when you’re presenting it to someone and it doesn’t overwhelm them.

Yusuf: 

Those are the three key mistakes. Now, quite often, the nature of the work that auditors do, particularly in the performance audit field, which is the more, public sector area, the audience could be the general public, right? So just down the road from you is the Government Accountability Office. And they would produce a range of reports that then are produced for Congress. But they’re made publicly available. So some of what they do would be with Congress as the audience, primary audience or maybe even the secondary audience and they have the public, the greater American public, as the primary audience. What do you then do when you have those multiple audiences and also audiences that you don’t necessarily know exactly what they want because they are so diverse?

Alli: 

Yeah, that’s really tough because, it’s very common for you to create something and it solves the particular need of your particular audience. But then someone’s like, you know, this isn’t a good chart. It doesn’t answer this question for me. It’s like, well, every chart can’t answer every question. So you really have to prioritize what question it really is important for you to answer. So I would say to take into consideration your main audience and your main goal first. And a lot of times what we do, is to create one version that’s more for a technical audience that knows all the jargon and then to create, kind of almost like the social media version, where we take details out and we highlight certain sections. I don’t think there’s any magic pill to that problem. It’s very difficult, but just know that even if you have the most perfect chart for the most perfect situation, there’s always going to be the person who says, but it doesn’t answer this question, so make sure you’re satisfying your first audience first. And then you can build alternate versions, if you want. I have done quite a bit at a work where we create three different versions of the same thing. Like the complex person who’s in the weeds version, the kind of mid-level version, and then the social media version. And each time you’re just taking out complexity, taking out extra details, and replacing it with bigger, bolder, some call-outs, bigger annotations, that kind of thing. I find, the more general your audience is the less time they’re going to spend with it. So you have to think a little bit more about how are you going to get their attention, and what’s the one or two things you really want them to pay attention to.

Yusuf: 

This is more on the sort of explanatory visualization side, when you develop a dataviz output, would you title the graph or the chart based on the question that it’s looking to answer? Is that something that would make sense? So this is the question, and then the chart answers the question.

Alli: 

Yeah, I’ve seen that as a very effective tool. For explanatory, putting in there, what you want the person to take away in your title is also very effective. You know, sometimes it can be considered like, Oh, you’re kind of pushing them to a certain conclusion. But I mean, if this chart has a very obvious conclusion, like product Y had the most sales in 2019. I mean, it’s not really biasing the situation. If you say, you know, product Y had the most sales. So I think calling it out in the title, for those explanatory things, especially for a general audience who might not be spending that much time with your graphic, just putting kind of the takeaway in the title. People can get it real quick. And then if they want to dig into the X axis and the Y axis and actually see what’s going on, they can, but otherwise they’re still getting the information.

Conor: 

Performance audits go for an extended period six to 12 months. Sometimes they go even longer. So the auditor’s doing these performance audits, you know, do a lot of detailed and granular analysis and work. And there’s always that risk or trap that they want to show all of their analysis in one chart or on one dashboard, to sort of signify, the fact that we’ve taken a long time. We’ve done a lot of important work here. We’ve been in this position ourselves, we’ve done this work and we think let’s just put everything in together. So what are some of the things we can tell them to think about so they don’t fall into that?

Alli: 

That’s a great question. My best tool for something like this is to test it on other people and to ask them to test it on other people if you can’t do it yourself. Them hearing that information or seeing someone in person struggling, figuring out what they’re trying to say. Then you can say, your manager is going to be doing this exact same thing when they’re looking at this and then they’re just going to give up and move on. I like to tested on other people. And there are websites. If your information isn’t confidential, which I’m sure is probably usually in your line of work. I think it is, but if it’s not there’s websites, usability hub is one that I’ve used before, where you can upload your graphic and then a series of questions and even, the kind of the demographic of people that you want to, focus on and they’ll send it to those people and they’ll answer your questions. Or you can even have like a private email address where you send your questions and your graphic to a person. So if you have someone that you want to test this on, they can look at it and they can answer your questions. If you can’t be in the same room as that person, which is very valuable in this pandemic time. So there are ways, and even if you just email it to the person, email it to someone who is on the team, who’s privy to the information, but not, you know, neck deep in it, like you are. And just ask them to answer one or two questions about this graphic. Like what’s your main takeaway from this graphic. And then they’re probably gonna come back with, it was confusing and there was a lot, but … Conor: Don’t be disheartened you get, because that’s because that’s good feedback, because you’re still experimenting with what you’re trying to show. Yeah, exactly. And I think if you are in the position where you’re trying to convince a client, that that graphic has too much information, you know, act like you two are a team and Hey, let’s send this over to Bob and see what he thinks about it. See what he takes away from it. And that person can fight the battle for you. It’s like, well, if Bob was confused, then maybe we should try to figure out a way to simplify this.

Yusuf: 

Yeah. And look, I mean, that happens to all of us, right? So you spend a couple of days working on a viz and you’ve produced what you think is a fantastic output. But you’ve been in the weeds, and so you can’t really see. You then send it across to, I send it across to Conor and Conor goes I don’t know what you’re talking about here. I’m like what do you mean you don’t know what I’m talking about, it’s all very, it’s all very, come on.

Alli: 

Exactly. I know. And you think, well, it’s so obvious, but, and it’s really hard too when you first start doing that, it’s hard to not be defensive and be like, well, I obviously meant X, Y, and Z, but you start to realize that having a fresh pair of eyes is so valuable. And sometimes I talk to my husband and I show him progress shots on things, but sometimes I won’t show him anything at all. And just so I have his valuable, fresh eyes, like towards the end of a project or in the middle of it. Because it is a huge resource to have someone who can give you fresh eyes. So the sooner you can get over your ego and realize that fresh eyes are really going to help you with, making your visualization as effective as possible, you can really make some great strides.

Narrator: 

The assurance show is produced by Risk Insights. We provide data focused advice, training and coaching to internal audit teams and performance audit teams. You can find out more about our work at datainaudit.com. Now, back to the conversation.

Yusuf: 

All right. So you’ve been involved in something that I don’t know much about, but viz for good, can you tell us a little bit about that?

Alli: 

Viz for social good is a nonprofit and they pair up with charities or other nonprofits. The charity gives them their data and then viz for social good distributes it among their volunteers who are dataviz practitioners. And then they visualize their data, and try to help them achieve their mission or further their mission. So, they had one where it was Bridges to Prosperity, and they helped them visualize something and they have a particular thing that they want to find out of the data. And then everybody tries to help them figure it out. And either creates them a dashboard or something explanatory. And it’s really amazing to have these dataviz practitioners come together and try to help them solve a problem because hiring a consultant can be out of a charities budget, but it’s a really great initiative and I’ve contributed once or twice to their projects.

Yusuf: 

Okay. It sounds like something that you would take a lot of personal satisfaction from. Is that community growing?

Alli: 

Yeah, it definitely is. It’s uh, let’s see. I don’t know exact numbers, you know, you don’t have to participate in every single one. I think about every month they have a project. but they usually get around 20 submissions for each one and I think it is growing over the years.

Yusuf: 

How can people find out about that and participate? Is it still sort of open for people to participate in and get involved?

Alli: 

Oh, yeah, totally open, totally free. You know, you don’t have to commit to anything in particular. Because I’ve been on their emails for a long time and then as a project comes out and I have time, I work on it. So just go to vizforsocialgood.com. You can click a link on there to sign up as a volunteer, signs you up for their mailing list so that when a new project comes out, you get emailed and notified and it gives you the data set and what this charity is looking for, you know, who their audience is, what they’re hoping to get out of the data. And then you can just start visualizing and then submit it on their website.

Yusuf: 

Fantastic. So there’s lots of auditors that are looking to improve their data visualization skills, and if you can do this, and at the same time, so, you know, you get a dataset, you can have a look at it, try to figure a few things out and learn and contribute as you go. It sounds like a win-win really.

Alli: 

Oh, yeah. And it’s so amazing because people come from different fields, you know, dataviz practitioners come from different types of fields. So, auditors have a different perspective than a journalist. So that skill set is definitely needed.

Conor: 

Auditors have been stereotyped over the years as sort of automatons or robots or people who think in a certain way. And I was struck recently in reading some of your material about a key tenet in your work as creating wonder in your visualizations. And I started to think about that a little bit and auditor’s and risk people also like to have a sense of wonder. And I wondered if you could just give us a bit of a brief overview of what that means in terms of how do you go about creating wonder for your audience? What are some of the key things we should think about?

Alli: 

Obviously you can feel wonder in a lot of different ways and the way that it translates to visualization, thet easiest I think is through context. So either feeling like something bigger is happening around you or to make you feel small or to make you feel connected. A really great graphic from Reuters recently was about the amount of plastic bottles that we create. They visualized this pile of plastic bottles on this kind of 3d map Of New York city. So you were able to see, wow, this pile of plastic is looming over the city. That when you’re in it, you feel so tiny in the city, but this plastic mountain is looming over it. So kind of makes you feel small and comparison. And you can use some techniques like that in your work. A visual metaphor is something that’s very effective. You don’t have to have a lot of illustration skills to do a visual metaphor. So, What I like to do for the visual metaphor is think about kind of what we were talking about before, like, think about your main goal of your visualization. And this works the best in like a process type thing. And you can think about, am I trying to explain like the structure of something or maybe the flow of something, and then just slowly start to describe your process a little bit more abstractly and just keep trying to abstract a little bit more. And using adjectives, like main words that you have, and then find adjectives for those words. And then find word associations with those words. And slowly, you just start, you can start jumping different fields. Like for one that I did, we were trying to describe this process for, finding the, vulnerabilities of your, enemy. It was like one of those war game type things. And it was a very complex idea. And I did something like that, where I described, okay, this is kind of the structure of this war game. And then I slowly described the process in a more and more abstract way. And then I ended up describing it in the way of like a house and a power line and like cutting the power line is like cutting the certain vulnerability to this particular, war game. So it was something that was very, technical and hard to understand if you’re not in it, but having the visual metaphor of a house and a power line, it’s something that people can understand. So you’re kind of using a visual metaphor as a bridge from complex information to something that someone already understands. And you know you don’t have to be an artist to draw a house. So, I think these kinds of techniques are more available to people than they really think.

Conor: 

The audit community has got a little way to go in terms of getting better at that storytelling really. And because we’ve treated traditionally, findings or analysis in isolation and said “this is what we found, here it is”. But we haven’t been able to contextualize that and put that against, bigger thing or a smaller thing to say, this is what it actually means though. That’s certainly something that we can improve.

Alli: 

And you can even contextualize things in terms of something that the audience understands a little bit better. Like this particular thing cost the company $800 million. That’s the equivalent of two jet planes or something like that. Contextualizing that information in that way, isn’t very hard. And you can easily find data on those kinds of things. So putting it in the context of something, your audience maybe deals with day to day this is the cost of 10 years of groceries or something like that. It’s pretty easy, for anyone to make those context comparisons.

Yusuf: 

In terms of the work that you do nowadays, where is that focus and how do you help organizations with their dataviz?

Alli: 

My main focus would be with people doing kind of research or scientific research, and helping them communicate their research to a general audience. I’m also working with a real estate agent he’s looking to use data visualization as a way to communicate his industry knowledge with a general audience so they can see, Hey, this is the person with the data. This person knows what’s happening in my local area really well. And he’s explaining it to me really nicely with nice graphics and I trust him. So using a way using data visualization as a way to effectively communicate your business, to a general audience. Is very valuable. And I think that people are just beginning to realize that something you can do because of places like the New York, New York times and Washington post, they are really growing their database and their graphics teams because they realize that’s, what’s getting the clicks and that’s, what’s getting people to buy newspapers. are all these really interesting and communicative charts like that one? visualization. They had about COVID Washington post, early on where it was a simulation showing how, COVID spread or a virus can spread exponentially. And it was kind of showing a model on how, how this thing spreads. It was the most read piece in the Washington post. So I think journalism is kind of leading the way, showing that people are interested in this. Small companies are just starting to see that they can also have something like this. You don’t have to be a big corporation with a big design team. You can be a one person, two person consultant team and hire a dataviz consultant to help you design graphics for a brochure or report that you put out to your audience. So that’s a long way to say that I do mostly for scientific and research people, but also for small businesses, and I think that’s becoming a more popular thing.

Yusuf: 

What’s the most challenging aspect you find whenever you start, or are working through a data visualization project. What’s the thing that creates the most complexity or difficulty for you during those projects?

Alli: 

The thing that is the hardest, which doesn’t sound that interesting, is just the state of the data. Someone has maybe a PDF of data and I have to try to figure out how to get it out. Or they have a whole bunch of tabs in Excel. And they’re labeled very poorly or it’s free text. Those kinds of things, which are very boring but that’s usually the long pole in the tent is trying to figure out what in the world the state of the data is in. And then closely followed by, helping my client figure out their audience and their goal. it’s something that’s often skipped. And I think it’s because even if you know to ask it, you might get it wrong or you might not know. You feel like maybe it’s going to be even harder for you to figure that out then to create the visualization. So you just want to skip it. But I have found that the projects that are the most successful are the people who really invest in figuring out who their audience is and what their main goal is. So, that is a very difficult part of my job, because I don’t think people really realize that that is part of the job. They just think, okay, you know, we have the data, let’s do it. And they like to wash their hands of it after that. But I do need a significant amount of information before I visualize something.

Conor: 

Sounds like maybe a fair bit of patience as well at times.

Alli: 

Yes, yes, yeah. That has definitely been my number one finding is that, people are people, and no matter what job you’re in, your number one skill is dealing with people.

Yusuf: 

What do you have in store for 2021? what would your key activities and goals be for next year?

Alli: 

My personal ones are, I have really started enjoying creating more tessellations and I’m currently exploring how to, integrate tessellations with data visualization. I don’t know if you guys are familiar with M.C. Escher? I’m sure you would recognize it if you saw it, but he really, he, yeah, he really popularized those tessellations the metamorphosis ones where it like turns from a fish and to a bird, um, you would, you would know it if you saw it, but, that was something I was experimenting with, with my latest vis for social good visualization. I created a tessellation and. the data was about, was actually about vis for social good and what the topics of all their projects was on. They had 20 projects on health, so I had a heart tesselation. Or if it was on youth, and then I had like a baby bottle tesselation and then they all were connected into one thing. And I think there’ll be an interesting opportunity to figure out how to integrate tessellations with data visualization, because maybe not so much in your guys’s world. I think maybe understanding and implication are more important to your audiences. But for a lot of audiences having a data visualization that’s eye catching and beautiful is It’s like the big currency. So, I think that, that might be a really interesting thing to try to, create, especially now with a lot of people using visualization for social justice type things. Like, I don’t know if you saw those warming stripes where. They had these almost like barcode colors that represented the global temperature. Blue was cooler and then red was hotter. And then you had all these stripes, right next to each other, explaining like the temperature of the world from, you know, one year on to current and you can see how it’s slowly getting redder and redder. And then people were like, putting them in on billboards and putting it on face masks. And this dataviz on someone’s face mask blew my mind. So I think that having data visualization move into more of a cultural, beautiful pattern element, that’s something that I’m looking to explore in 2021.

Conor: 

Are there particular types of analyses that lend themselves more to tessellation?

Alli: 

So far, my findings have been things that you would normally put in like a unit chart unit chart. Like, if you have a group and 52, are women 48, are men something like that? And then you have 52 men icons and 48 women icons, whatever. something like that, where you have a unit chart and you have an icon representing each group, I think tessellations really lend themselves well to that or something where you showing proportions or I’m thinking like a stacked bar chart or a stream graph type situation, like things with proportions, maybe things with icons so far, that’s what I’m seeing lends itself best to tessellations.

Conor: 

Yeah, we’re certainly keen to see your progress in that space. We’ll be following it very closely Alli.

Alli: 

Yeah. Thank you. It will be exciting.

Yusuf: 

You obviously work on a range of different projects and people in our audience will be interested in those and potentially connecting with you. What’s the best way to reach you?

Alli: 

Yeah, go to my website. I have my email address there and a contact submission form. Allitorban.com.

Yusuf: 

Alli, thank you so much for joining us. Lots of interesting tidbits for auditors to take away, and I’m sure we’ll have more discussions around specific aspects of dataviz as we go and as we get deeper into the subject with internal auditors and performance auditors.

Alli: 

Yeah. Thank you. Thank you guys for having me on it’s so nice to see you guys again.

Yusuf: 

Yeah, you too. Thanks Alli.

Alli: 

Bye.

Narrator: 

If you enjoyed this podcast, please share with a friend and rate us in your podcast app. For immediate notification of new episodes, you can subscribe at assuranceshow.com. The link is in the show notes.