Ann K. Emery Drops her Top Do’s and Don’ts of Dashboard Design

[00:00:44] LP: Hello my dear listener, and welcome to the 73rd episode of the Present Beyond Measure Show, the only podcast at the intersection of presentation, data visualization, storytelling and analytics. This is the place to be if you’re ready to make maximum impact and create credibility through thoughtfully presented insights and ideas.

Today, you are here either because you like data and you design dashboards for a living and you want to learn how to do it better. Or like me, you are a raving fan of today’s super special guest. Now, as usual, I am super stoked for today’s guest. That never changes. But in particular, this guest has been on my radar for a very long time. Where her practical and practitioner-centered approach to data visualization and communication has garnered her a very well deserved and raving fan base.

Let’s jump to it.

[INTERVIEW]

[00:01:55] LP: All right, everyone. Today’s guest is an internationally acclaimed speaker who equips organizations to get their data out of dusty old spreadsheets and into real world conversations. She is considered one of the fields foremost experts. Each year, she delivers over a 100 keynotes, workshops and webinars with the aim of equipping organizations to visualize data more effectively. She has been invited to speak in 10 countries, thousands have enrolled in our online training academy and she’s consulted to 200 organizations, including the United Nations, the CDC and Harvard University. Hmm! I don’t know, heard of those. She’s been a personal hero of mine since the beginning of my journey into data visualization awesomeness. Please, help me fangirl and welcome the latest guest in my superstar women in analytic spotlight, Ann K. Emery.

[00:02:53] AE: Thank you so much, Lea. That’s great. What a great intro. Wow! Geez! I’m blushing.

[00:02:59] LP: Well, it’s a podcast, so half the people won’t see that. But no, honestly, it’s very well deserved. This is a moment in the making for me for 10 years now. Your work has had a profound impact on the development of my journey as both a date of vizier, and a speaker and trainer in this space. I’m such a huge fan and I’ve just been working so hard to get you on here, so this is an exciting moment.

[00:03:23] AE: Great. Great. Great.

[00:03:25] LP: Everyone loves a great superhero origin story. You’ve been at this for quite a long time, but we’d love to really hear how did you fall into this wacky and wonderful world of data viz.

[00:03:40] AE: Well, in college, I knew I liked spreadsheets. My dad is an economists, so I just kind of inherited the spreadsheet gene. I remember mentioning to my – I think you’re just kind of born with it.

[00:03:50] LP: I have that.

[00:03:51] AE: Yeah. Some of us are. We’ll talk about technical people in spreadsheets, I’m sure plenty. I remember, I didn’t know what career paths you went into with data. This was before data science, or data analytics, or big data, or whatever the big AI, whatever the big terms are these days. This is like back in college time. I mentioned to a professor, “Oh! I guess I’ll be an accountant. I think I’ll be an accountant. I think that’s what I’ll do.” He was like, “No, you would be terrible at that.” I thought he meant that my spreadsheet skills were terrible. I must have showed that on my face. Like “What?” And he said, “No, you’re an extrovert. You can work with data, but you have to do something working with teams, and people and like bringing people together with that data. There’s some other angle of it that’s not just looking at spreadsheets all day every day like accountants.” I was like, “I don’t know. That sounds like the best thing in the whole world to look at spreadsheets every day,” which I get to do for big chunks of every day and is great.

After college, I got the typical corporate jobs, and then one thing led to another. I started speaking at conferences, as I know you do a lot too and people just started to ask me to do this at their company. I would give a talk and they’d say, “Wow! This was so helpful. Can we fly you out and you give the full day version of the star company?” I did a couple of those to get my feet wet and turned it into a full-time job. And here I am, year eight of working for myself.

[00:05:12] LP: Wow! It’s only been eight years. That’s amazing, because that’s not too far from how long I started. But I feel like you had such a strong presence even back then when I discovered you. You really must have hit the industry hard.

[00:05:28] AE: Well, I think, I started blogging and YouTubing before that, when I had a salary job, just for fun as a pure, honestly, as a pure hobby. I have three kids now, so this was back pre-kids when I could like wake up on Saturday morning and write a blog post. Not anymore. Not anymore at all. But yeah, I started doing like blogging and YouTubing as a hobby, pre-self-employment too. That might be the timeline that more people are familiar with for me.

[00:05:54] LP: Well, that’s already an amazing tip right there is, if you’re really passionate about the subject, that you’re a specialist in, get out there and start writing a blog or writing videos and you never know what path that could lead you to. Right?

[00:06:07] AE: Absolutely. And do it before you have kids, before your time disappears.

[00:06:11] LP: Yes. Hire a team, automate. Yes, completely. Well, that is a really fascinating story. I’m so thrilled that you were rejected from a career in accounting in order to bring so much knowledge to this space. Your work has really covered so much of the gamut of data visualization. You’ve been on so many podcasts. I used to use your chart choosing tool and things like that. But today, we’re very interested in talking about dashboards. These things that like I get asked about all the time, and everyone loves, because I’m more on the more boring, non-sexy data presentation side, but people love their dashboards. I think this is one of the areas where the little changes can make big, massive improvements. You actually brought today an idea, like a framework that you have for helping people understand what kind of dashboards are out there and when to choose them. I’d love to hear more about that.

[00:07:14] AE: Okay. I’m going to give you a quiz first. This is not a hard quiz. This is a poll.

[00:07:19] LP: Okay. Let’s see.

[00:07:20] AE: Yeah. Turning the tables. You’re in the hot seat now. Okay, Lea. Give me your definition of dashboards, because there’s 50 million correct answers. What’s your definition?

[00:07:31] LP: All right. On the fly, my definition of a dashboard is a single view suite of data chart modules, or data form modules that are designed to alert a lay audience to the most critical business key performance indicators, business metrics in a way that they can either make very basic decisions on their own, or alert them to ask the proper teams to investigate those changes further.

[00:08:04] AE: Wow! That is a very sophisticated definition. Not that I expected anything less of you, but that’s probably the best I’ve ever heard, to be honest. You get a trophy. You get a Friday afternoon trophy. Gosh! I love asking people that because you truly get so many different answers. Then those are your words to describe it, but the picture you have in your head or ones you’ve seen, that’s different for everybody too. Because dashboard software programs have only existed for, I don’t know, 10 years. When did Tableau become a thing? Five years ago maybe. I don’t know. They’re growing and growing in popularity all the time.

[00:08:42] LP: I want to say at least 10 years, because right when I discovered this whole field, they were there. It has to be around the 10-year mark.

[00:08:51] AE: Oh! You know what? You know how I mentioned, I started speaking at conferences, and then people would ask me to go speak. That was my first conference presentation was, I think 2009 about dashboards. That was 12 years ago now, about dashboards when they were the buzzword. If you talked about dashboards at a conference, it was a guaranteed standing room only presentation, which might still be the case now. Maybe, I don’t know. I feel like there’s other topics that are – you said dashboards are sexy. I feel like presentations are sexy. We should switch jobs.

[00:09:23] LP: Right? Exactly.

[00:09:25] AE: Nowadays, people have entered the field maybe more recently, might think of a dashboard as “Oh, it’s Tableau. It’s Power BI. They think of maybe a certain tool with a web-based interactive dashboard, where the user can drill down, and use checkboxes, and drop-down menus and explore the dashboard on its own. I think it was – it’s interesting. You mentioned something about maybe a single screen or a single view. Some people use that in their definition, some people don’t, some people are like, “Oh! I need multiple tabs.”

[00:09:54] LP: Or scrolling is another option, yeah.

[00:09:57] AE: Before these software programs existed, 2009 and prior maybe, what people called a dashboard. This is my end of one personal opinion. I think they meant like a static dashboard, a printout or a PDF. That used to be the thing for dashboards, especially with the groups I work with. I think we have very different groups that we work with. I work with a lot of federal and state government agencies. I work with a lot of universities. I work with nonprofits and foundations. That still a lot of times how they think about dashboards, honestly, is it’s a piece of paper at your annual meeting. It’s the packet at the board meeting with your key facts and figures on it. Just getting people to sit down and recognize that there’s different types of dashboards is so critical, because if you work in a data role, everybody has been asked for a dashboard at some point, everybody has/

[00:10:51] LP: What we need right now is more dashboard. It’s the one I hear a lot.

[00:10:55] AE: I told you, I don’t know if this is before we started recording or after. But I told you I was struggling with my computer this morning working on a dashboard. I’m helping a group revamp their dashboard, like dashboard in air quotes. It’s just an Excel file with columns and rows. There’s not a single visual anywhere to be found. It’s a very well-organized table that could feed into graphs, but it’s like that’s the lingo they’re using in-house to mean dashboard. The file name says dashboard, the title on the document says dashboard. When we’re emailing back and forth, and talking over Zoom, it’s all dashboard, dashboard, dashboard. I open it and I’m like, “No, this is a table. This is a data set.” Yeah. Every single day, I have to kind of clarify. “Wait. You’re asking for a dashboard or you say you have a dashboard. Tell me more? What does that look like? What does it feel like?”

[00:11:43] LP: Right. The word kind of creates certain expectations, I think, and I think this is important for both dashboard designers and consumers to understand is spending years in an agency and on the client side. That was almost the first thing people ask for it, “Oh! We need a dashboard for this. I don’t know what I want, but we need a dashboard in.” Getting very clear on the expectation of what that’s supposed to accomplish on an ongoing basis, I think is the first critical step for making that ask. I don’t know what do you think about that?

[00:12:15] AE: Absolutely agree. I’d say, step zero is recognize the dashboard idea we have in our minds, is going to be probably a little bit different from the dashboard idea somebody else has in their minds. Then step one is get clear, get clear on what the expectations are for that dashboard. This is where the four types come in that I’m going to try my best with my frazzled brain after my Excel was breaking on me all morning. I’m so angry about this. I would go to insert a row and it would just freeze for like 60 seconds. I’ve Googled all the things. I restarted the computer. It was it was fine for a couple of hours, and then I’d go to insert a row and it just wasn’t having it.

[00:12:57] LP: Maybe this is some dark ad where from a dashboard companies saying, “Man! Come to us.”

[00:13:03] AE: Stop using everyday software programs, pay more, use our fancy software. Maybe.

[00:13:09] LP: No, no. They would never do that. They would never do that.

[00:13:11] AE: Okay. Where were we? I’m going to try to describe the four types of dashboards that – I have a visual for this. I think I have it on a blog post somewhere that we can put in the show notes. I’m going to try to describe it [inaudible 00:13:23], so we’ll see how we do. This is the planning convo we’ve all got to have when people ask us about dashboards to make sure what people are expecting is what you make for them, that there’s a nice cohesion there. Step one, figure out if you need a static or interactive dashboard. I’m going to be like, I’m thinking about kind of quadrants here. There’s kind of this quadrant of – you’re either on the static side or you’re on the interactive side.

[00:13:53] LP: That’s like the spectrum going from like left to right.

[00:13:55] AE: That’s your left, right. Yeah, your left right and your – gosh! We can geek out on your podcast, right? It’s your X axis, horizontal line. Static or interactive? Static is, you really just need a meeting handout. Once things go back to in-person more and more, you really just need like, there’s a staff meeting in the conference room. What you really need is a on pager of key facts and figures that might be printed out and passed out as a meeting handout. Or static would be an email attachment. People say, “Hey! Do you have those key numbers and you say, “Sure. Here they are. Please see attached” and you send them a fairly short PDF. It doesn’t have to be a page. I think that’s a made-up rule some people have in their head, but it’s not a full report. It’s a short-ish document. Where else would static come in? It’s your printouts. It’s your PDFs.

[00:14:49] LP: Could it also be informed by the level of acuity or savvy of your audience where maybe an audience, or decisionmaker really wants to have a little bit of narrative summary with a few graphics, and like it all tied up. Read to me like a mini bedtime story about our data, make it a page and I don’t want to be drilling down and I don’t want to –

[00:15:15] AE: Agree, 100%. Absolutely. Yeah. I have in my notes. It’s more for the non-technical audiences or what you might call a lay audience. I think you used lay earlier. It could also be the busy audiences, which is like every supervisor or person in a leadership position ever, ever. I mean the really smart people, the really highly educated people. They just don’t have the privilege of time. Providing them some little details, like you said, it’s going to be so, so helpful. It’s a gift that you’ve said. I’ve already dug through all the messy data. Here’s what I think are the key findings for you in this tiny little package.

[00:15:56] LP: It’s like putting the cooked dinner in a five star restaurant and lifting the metal, whatever that is, whispering the ingredients in the ear and then backing away quietly.

[00:16:07] AE: Lea, you’re so good. That’s so good. Yes.

[00:16:11] LP: I love analogies. They’re like my thing. Thank you.

[00:16:14]AE: I love drawing diagrams, so you put us together, we’ve got an analogy diagram going on. Static, we know when that would be really helpful. We know what that looks like. The other option, that’s also correct sometimes is interactive. That’s going to be, somebody opens, it’s going to have to live on a computer or everybody’s got phone apps that are dashboards that you can drill down, maybe you have your own data. I don’t know. The apps that track your runs, or your bike rides or something and you can say, “Oh! Just show me this time frame right now” or “Just show me my –” I think Strava does this. I started using Strava again for exercise. It has some nice little charts in there and stuff, but it’s digital. It’s digital. It’s not a PDF. It’s something that you use the drop-down menus, you drill down yourself, you interact with it.

[00:17:02] LP: Right. Interact with it.

[00:17:04] AE: The trouble with interactive dashboards is, they’re not always time appropriate for everybody. People like me, probably people like you who like data we’re like, “Great! Let’s make a dashboard” and we like exploring the dashboard, and we like drilling down because data is our favorite thing in the world. We’re like, “Yeah, let me explore it, and find the nuances, and find the insights, and figure out what’s going on and know everything about this data set.” But then the people we’re giving the dashboard to, they often just need those quick insights, the key takeaways. I see a lot of mismatches happen in the real world a lot where the data people like using interactive dashboards, so they make interactive dashboards, but they’re making it for somebody who really just needs a one-page PDF. There’s just this constant kind of mismatch in what’s designed and then what people actually need.

[00:17:58] LP: You may enjoy cooking the meal, at least a little bit, or prepping it, but your audience may want to have nothing to do with that.

[00:18:07] AE: They just want to eat it. They just want you to do your little reveal.

[00:18:11] LP: Right, exactly. Yep.

[00:18:15] AE: Okay. Many more interactive dashboards should be static dashboards. They really should. I hate saying that aloud. Nobody wants to hear it. But it’s just –

[00:18:24] LP: But that’s so not innovative, yeah.

[00:18:26] AE: It’s so boring. Really? Are you sure my boss just wants a one-pager? Yeah, they do. They just want a one-pager. How do I know it’s [inaudible 00:18:36]?

[00:18:36] LP: Especially if you got an effective one.

[00:18:37] AE: Yeah, exactly. A good one. Not a terrible one, a good one. Okay. Static or interactive? That’s a key decision-making point, and it really depends on your audience. Non-technical audiences, give them the static dashboard. Technical audiences, give them the interactive dashboard.

[00:18:57] LP: Got it.

[00:18:58] AE: The other decision-making factor, so imagine this one is your vertical line on the quadrant, like your vertical Y axis? If we’re making a little like a plus sign kind of quadrants. That one’s really easy. Do you need a single dashboard or do you need lots and lots of matching dashboards?

[00:19:15] LP: Multiview.

[00:19:15] AE: The first time we worked on dashboards, I’m thinking of that 2009 million years ago conference presentation. I used to do education data, so I would work with the US Department of Education or state education agencies. A single dashboard would be like all the schools in a state combined, however many hundreds or thousands of schools, the state overview. Then a series of dashboards would be one per school, like school A gets a dashboard, school B gets their own dashboard, school C gets their own dashboard.

[00:19:48] LP: Got it. The single is more like an aggregate level for that really bird’s eye view and then Multiview is allowing you to get more granular for more specific users. Okay.

[00:19:58] AE: I also see mismatches there too sometimes. People will think like, “Well, doesn’t the school also want to see the aggregate dashboard for all the schools combined?” I mean, sometimes, what they really want to see is their own school’s data. It kind of just depends on somebody’s position within a project, The people in the state education agency, they want to see all the schools combined. They might – they might want to see a couple of schools individually, but really, they want to see like, how’s our state doing? The people working in the schools, they want to see how is my school doing. There’s not a right or wrong answer there. It’s just one of those planning considerations that we’ve got to talk about upfront to make sure we’re making the right dashboard and we don’t run into so many issues later on.

[00:20:42] LP: If I’m hearing you correctly, humans, who are the people consuming these dashboards generally want to see data at the level that pertains to us?

[00:20:53] AE: Yeah, of course.

[00:20:55] LP: I hope everyone heard that.

[00:20:58] AE: Say it again for the people in the back, Lea. We want to see useful data, right? We want to see data about us, about our own project, which is so obvious to say aloud. It’s one of those things you do need to say aloud at the beginning of a project, because otherwise, I see groups spend months working on dashboards, and so much money and – or they hire a consultant and the consultant might be very skilled at dashboards. But there hasn’t been that conversation upfront about what type of dashboard. The consultant does their best guess, they of course, they do their best guess and they deliver what they think the people need. But it’s just so much time and money lost, just because these conversations that take 15 minutes haven’t taken place.

[00:21:49] LP: Well, I really think it’s in the spirit and this goes for data presentations, too. But it’s making these outputs audience centered. If it’s designed for more than one audience, in one single view, there’s going to be a red flag that someone or all will not be happy about some aspect of their view. You’re right, an external consultant might not have that insight. They may come in and apply principles and use specific frameworks, but it’s understanding that internal culture that is really essential.

[00:22:22] AE: Yep. I think a lot of times, that’s like step zero zero too, which I think you and I know because we do this for a living. But thinking about who the audience is, and really trying to narrow your audience to one. Everybody has multiple audiences, so you know, I’m sure you’ve done this with groups like you’re making some presentation slides. Who are the slides for? When is your presentation? Who’s going to be in the room? Obvious questions like that, right?

[00:22:50] LP: What needs to be covered and how does everyone get their needs met?

[00:22:54] AE: Absolutely. Okay. We talked about the four types, static or interactive, single or series. Then depending on what type you need, depending on which of the four quadrants you fall into, that determines what software program you’re going to use. Software comes later. Software comes later. Everybody is like, “We just got Power BI and what can we do with it?” I don’t know, “What do you need to do?”

[00:23:20] LP: Right. Did you need it?

[00:23:22] AE: I started switching from Excel to sheets to see what sheets could do. Great! But it does the same things as Excel pretty much. Like what do you need it to do?

[00:23:32] LP: Why? The question is why.

[00:23:35] AE: Yeah. You know, Lea, I’m a big fan of everyday software. I use Excel, and PowerPoint and Word as much as possible. It tends to be the common language among all the groups I work with anyway. I think people are usually pleasantly surprised to learn how much Excel can do. Everything on the static side, Excel can do. If you’re making a one page PDF, make it an Excel, set it up to be PDFable. Just set the page margins, you can adjust the page breaks, you can add page numbers, you can adjust all the font sizes, you can make it look like it was made in some other software program. It can look very beautiful. It’s just Excel saved as a PDF. You can even do a lot of interactivity with Excel too with slicers that are a lot easier to use than people might realize too. Slicers sound confusing, it’s just a filter. There’s a million Google resources and YouTube resources on how to add slicers. They’re not complicated or time consuming at all.

[00:24:31] LP: Right. Though this is great. One of the questions I feel like would bubble up for practitioners since I spent many, many hours creating static dashboards in Excel. Well, I sort of exited this realm when things like Google Data Studio really started ramping up where they were kind of creating an Excel-ish experience, nothing insanely groundbreaking in terms of visualization. But these platforms, we’re aiming to alleviate some of the headache that practitioners experience in actually creating and designing them of cleaner data, import, and hygiene and making the process of getting raw data into actual visuals more streamlined. That could still be PDFable. I’d love to hear about tools like that, where you don’t have to intend for it to be interactive, where a consumer gets lost in it. But they might solve for some of the like very rote manual data issues in Excel.

[00:25:33] AE: Yeah. I’m not too familiar with Data Studio, to be honest. I’m familiar with my Depict Data Studio, but not the Google Data Studio. But in terms of Google’s sheets, I put Excel and sheets and then if you’re on a Mac numbers, those are the same bucket in my mind. They do almost the exact same thing. There’s little nuances that could give us all a huge headaches, but they’re mostly some – it’s like a Venn diagram [inaudible 00:26:00]. I can’t help myself. [Inaudible 00:26:03].

[00:26:05] LP: Yep, yep. Well, I guess that’s something to think about, again, because it’s about the why. Right? Someone might choose to use a platform Data Studio, definitely is more of a visual interface to like a repository of data, which I have found useful for pulling stuff in where you can pull in from sheets. But you can also pull indirectly from other data sources so that you’re not extracting data and putting it into Excel, which is the way that I used to do it. Do you have any tips for that in terms of making that process easier and more streamlined? Or it just is what it is?

[00:26:45] AE: Google Data Studio specific tips would be, check out my friend, Ben Collins. That’s what he does for a living. It’s benlcollins.com. He is like the Google and data trainer. He’s got all sorts of courses. Honestly, that’s why I don’t have to learn about sheets or Data Studio because I’m like, “Go ask my friend.” He already knows how to do all of that, why would I learn? Just refer people to him. But then in terms of getting data into Excel, that I don’t know, that’s so tricky. Every organization is so different.

I’ll give people a really practical tip, I was working with a woman on Wednesday, just a couple days ago. Anyway, I don’t know when this is airing, but it’s just a couple days ago. She works at a foundation. so she has a foundation database, where they all enter in information about their grants that they give out to organizations in the community. I don’t know. It’s like a grant level type of database. She wanted to look at some patterns over time. Who doesn’t? She was getting ready to pull one year’s worth of numbers. We’re trying to make it easy for her to pull the next year of numbers, so she could do some time comparisons. What she had been doing – what has she been doing? She had been like downloading the data one year at a time. 2020 data was its own sheet, and then 202 was its own sheet. She couldn’t kind of compare across sheets and I told her, “Just do one download. Do one download and include that column in your export that says the year” and then we put that into a pivot table and filtered by year.

If people who are listening work with data, they’re like “Duh! I understand exactly what you’re talking about.” If you don’t work with data, I’ll just say, “Try to keep all your data in one spreadsheet when you download it. One single spreadsheet to use is going to save a lot of time.

[00:28:38] LP: One hundred percent. I think the hardest lesson I learned was all of these discrete imports and then placing them on sheets. I always thought to myself, can I create a column that allows me to enter this as just another dimension of the data. Saved so much headache. Very cool. That’s it. That’s a great tip, because I know there’s a range of expertise of people coming into this, of who listens. Everyone loves the do's and don’ts. First, I want to start with, what are some of the more effective, awesome, either examples of dashboards you’ve seen or techniques that are being used that you’re like, “Yes, please more of this”?

[00:29:23] AE: Well, anybody who has a dashboard, five gold stars, well done, because getting your first dashboard set up is the hardest part. So anybody who’s listening, if you have one, you know how hard this has been. It might have required hiring a new staff member, it might have required going to training, it might have required lots of lots of long hours, and thought process and just mentally thinking about, what’s the important data to include? Where do we get that data? Getting the first dashboard is super, super hard. It’s hard to give examples of specific really good dashboards because so many groups dashboards are private. I get to see them because I sign all the papers with them to give me permission, but a lot of them aren’t public facing. That’s changing though for a lot of the foundations and government agencies they work with. More and more data is becoming public all the time, which is really cool. But a lot of the good ones are behind the scenes.

[00:30:15] LP: That’s something to look out for. It is true that a lot more data is being socialized, especially in the last couple of years, where data has been sort of top of mind for a lot of people. But maybe this question then might be a little easier to answer. What do you think are the biggest challenges or mistakes practitioners are making when it comes to building their dashboards, whether it’s charts, colors, formatting, layout? What do you think are the biggest hang ups?

[00:30:48] AE: I think the biggest challenge a lot of us run into is volume, just how much were trying to include in a dashboard. That seems to be a roadblock for everybody working on a dashboard? How do I fit all the charts on this page? Do I scroll or not? Do I have different sheets or tabs on the bottom? If it’s a printed dashboard, how many different pages can I have to fit everything? If anybody listening is running into that problem, I would say, timeout, go back to the whatever 0.00 question, try to pick one audience that this is for. Because a lot of times people are like, “Well, this is for my boss, but it’s also for our Board of Trustees. But it’s also going to be public facing on our website someday and it’s also for that staff meeting. But I also want to share it at that concert with my peers” and that’s so many different data priorities, and data needs to try to fit into one dashboard.

I just see a lot of dashboards that looks very full, just so many different charts.

[00:31:48] LP: Bulging.

[00:31:49] AE: Bulging, bursting at the seams, absolutely. Just take a step back and say, “Okay. If I had to pick one audience, and prioritize just the data points that that one audience needs. That solves almost every other problem. It really does.”

[00:32:04] LP: Right. That is interesting, because I’ve often wondered about layout in dashboards. Does the layout of the modules matter? Because often, I’ll go and I’ll see like on a dashboard platform, they’ll show you examples of what their dashboards look like, what the reporting looks like of different reporting platforms that you can use. They’re like platinum view dashboard, just really very random. Seems like a very random selection of as many different kinds of chart types as possible to almost create like a candy shop feeling of, “Ooh! Look at all of these different things.” What is your take on actual layout? Does it matter? Should it have all different chart types?

[00:32:50]AE: The chart types discussion is interesting. Because I come from a research background, like my first job was – I worked in a university and I churned out peer reviewed journal articles for an NIH funded study, like very academic, very technical, very researchy. I had all bar charts all the time. It was, I think, I’ve seen, what if I maybe see in my first couple years of working. I had seen bar charts. I had seen a pie chart at some point, I hadn’t used them. Maybe lines, if we collected data over time. Scatter plots, maybe scatter plots are kind of what you use in a stats class. They’re out in the real world. They can absolutely work in the real world. That’s about it. I don’t think I’ve ever even thought about including maps or any of the other options.

Anyway, people should know my personal weakness is, I have way too many bar charts. I always do it in all my drafts so that I know that about my – like we all know our strengths and weaknesses, right? I know my weakness is, my go to his bar charts. Then as I edit my slides, or dashboards or reports, I try to add more variety so that I’m not looking at it saying, “Oh, geez, Ann. Have you not learned anything? You have all bar charts over and over again.” But then sometimes I hit the phase where I’m like, but this is just for an internal audience, and they’re all technical and the bar charts perfectly sufficient. Why am I trying to do anything that looks like a candy shop and just use variety for variety sakes? I don’t know. Honestly, I’ve gone back and forth in that decision about whether like everyday charts like bar charts are okay, or whether you do need variety. Honestly, I’m not sure where I sit as of today.

[00:34:33] LP: Wow! That’s insightful. In reading Nancy Duarte’s Data Story, she talked about how their company conducted this massive research into thousands of their clients executive presentation slides. They were trying to look for the most common chart types. They were expecting elaborate innovative visualizations and things. What was amazing is, it came down to three where the vast majority are bars, pies, or doughnuts and lines. That was it. Well, I personally believe you can include a few more in there, especially with executive presentation, like bullets, stacked bar, 100% stack bars, things like that. Sometimes they’re going to require the right training, because those three people immediately understand what they’re supposed to understand from seeing those charts. There’s no learning curve, there’s no training gap. I think that’s why we end up going back to that.

To your point, asking that question why is still so critical? Why am I changing it up? Is it for variety sakes or is it because this is truly the best chart type for this? Or if it’s kind of a neck and neck, like they can work equally well. Okay. Then let’s throw some variety in there if that’s really called for you.

[00:35:59] AE: Yeah. Or am I just picking it because I have a chip on my own shoulder and I’m just embarrassed looking at my work from 10 years ago, these all bar charts. I’m like, “I got to jazz it up now, because the bar's been raised.” I didn’t realize that. Oh my gosh! I had this conversation a couple weeks ago with a group that had – this is super, super nuanced. But you have a dataset, savvy audience, maybe they’ll appreciate it. But it was a category with subcategories with sub subcategories. That’s how their data was structured. Some people call that nested data or hierarchical data. It was three levels of data, like the broadest and then the subcategories sub sub, right. Okay.

There’s a lot of different ways you could display that. What they previously had was they previously had three separate charts on three different slides, which didn’t work because these were related. You couldn’t tell that something was a subcategory of something else, because it was physically on a different slide of the presentation. It would take a very skilled presenter to be able to click through and say like, exactly, to just say, with their voice, how these charts were connected. It needed to be one cohesive chart, not three separate charts on three separate slides. We’re thinking, “Okay. How do we bring these three disparate charts together?”

What are all the things we brainstormed? We tried a tree map to show how the nesting or the hierarchy worked. We tried a sunburst diagram, which is basically a donut chart, with layers going outward. You could see like, pretend one donut is a fourth of the donut, and then that breaks off into the next rung, and the next rung. Those, just based on their data, because they had some zeros or some 1%, or some little tiny slivers that just didn’t show up. That didn’t quite work. Then what we settled on as being the best approach was a Sankey diagram, which basically looks like stacked column charts kind of in a row. Here’s this broad one, subcategory of sub sub covers. It was going from left to right across the screen.

Well then, we ran into a different hiccup, where you can’t really make that in PowerPoint or Excel. So then how do you make that? The guy was like, “Well, I know how to make this in Tableau,” but then making sure the formatting from Tableau matched everything else they had in PowerPoint, or do you go to like sankeymatic.com and make it there. But then that’s an extra step. I wish it was some simple discussion of like, “Yeah, just use a bar chart if this, this or this or just use a Sankey diagram or this, this.” But it’s always more nuanced than that of like, “Yeah, we could make a Sankey diagram, but at what cost of staff time.? Is it really that much better than –?” I mean, honestly, we haven’t finished this project yet, but we’re probably going to end up with some bar charts. We probably are. We probably are going to end up with bar charts to save time. Because in the context of this broader presentation, these charts are not the star of the show. They matter, but they don’t matter matter. I don’t know.

[00:39:02] LP: Well, I think what you’re getting to is, again, people are always blocking this line between form and function. Even though appeal, visual appeal is important, like people will actually trust something more because they like the way it looks, which is something I picked up from Cole Nussbaumer Knaflic’s book. But the thing is, I think we make a lot of sacrifices for form and we sacrifice function. This is why that why question is literally like, I’m constantly asking this myself. Why would I choose that? By the time I’m done looking at this, can it answer these questions? It’s like something I call the grunt test of like a cave person coming and looking at something, and grunting, and knowing exactly what they were supposed to get from it or they’re grunting in like confusion.

[00:39:56] AE: You need a sound effect of grunt.

[00:39:57] LP: Yeah. I think this is why intentional design and data viz is like the core pillar of my entire approach to things, because I never want to do anything for the sake of doing something either because I’m not thinking about it and I’m just not conscious. Or the reason isn’t in the highest service of comprehension and understanding for the audience.

[00:40:28] AE: Agreed. We’re speaking the same language.

[00:40:33] LP: Yes, indeed. Well, good luck with the sand key that might end up being [inaudible 00:40:39] and that’s perfectly fine.

[00:40:39] AE: Yeah, we’ll see. To be determined. Stay tuned.

[00:40:44] LP: You have to keep us updated. All right. Okay. We’ve arrived at our upgrade segment, which is a tool, a resource, a book, a something that you are loving right now that’s really taking your practice to the next level or something that was really integral to your journey as a practitioner. What do you got?

[00:41:15] AE: All right. I have so many good books to choose from, but I’m going to show the one that I am currently just fangirling over.

[00:41:22] LP: Welcome to Ann’s Bookshop.

[00:41:24] AE: It’s called I Am a Book. I Am a Portal to the Universe. It is by – here’s the author’s, Stefanie Posavec and Miriam Quick. I had to order this. I think it came from the UK, so there’s no like Amazon one day shipping if you’re in the US. It still comes in about a week though. It is amazing. In terms of text placement. Text placement, Nobel Prize for this. I told you, I come from a really boring research background of all academic reports, Times New Roman, size 12. This book gives me permission to use like font size 5,000 and to not have them all go –

[00:42:08] LP: This is going on my Christmas list.

[00:42:09] AE: – to not have them all linear. This whole book is a story. The story is based on the art of how the text is displayed. It’s an educational story. It’s about facts and figures in science. But it’s just the text plate. It’s just absolutely a masterpiece. It makes me think –

[00:42:27] LP: So visual.

[00:42:28] AE: – why have I been so boring with paragraphs. A paragraph doesn’t just have to go from left to right in a certain font size. I don’t know if this one – I don’t know if they have an eBook version. You really have to get the physical one, I think to understand it.

[00:42:42] LP: Yeah. What sounds amazing about that is, it’s so helpful sometimes when something comes along that breaks us out of a paradigm that we are so enmeshed in of reading paragraphs. Like I’m writing my book right now and it’s paragraph format, with visuals, but that’s what I’m used to. That’s what I’m used to reading. That’s what I’m able to produce at this time. But even just the pages you showed me, it’s amazing to see what your brain does when it’s challenged a little bit. It’s like this crazy house that I heard about once, where they built the house with a floor that was completely uneven and nothing was even more predictable whatsoever. Like the stairs were all janky, and the floor had these constant slopes and you had to constantly watch your step and think about how to move across the space. You literally couldn’t check out, like there’s a word I’m looking for. But you couldn’t go on autopilot, right? I think a book like that is going to be amazing for jiggering some creative juices and seeing how that can flow. That’s amazing. I’d love it.

Before I enter the final question, is there anything you’re really excited about with the future of data storytelling right now?

[00:43:58] AE: I cannot wait till I can stop talking about some of the topics I’m talking about.

[00:44:05] LP: I know exactly what you mean.

[00:44:08] AE: Let’s rerecord this in 10 years.

[00:44:10] LP: You got it.

[00:44:11] AE: Our dashboard conversation, I cannot wait for that to be a different conversation. I hope. I hope. That’s what I’m working towards, that we no longer have to say, “Guess what, not every dashboard has to be interactive. It’s okay to have a one-page PDF” or we don’t have to say, “Hey! Really, we need to narrow down our audiences.” That we just don’t have to talk about that because everybody gets it, and they’ve already moved on to bigger and better things. I cannot wait for the types of topics that I’m blogging about, and talking about podcasts about and going to speak at conferences about. I can’t wait for this to shift someday. I don’t know if that was the answer you wanted me to say, but that’s what’s been on my mind a lot lately.

[00:44:48] LP: Actually, I completely resonate. There are certain parts of my training where I’m like, I dream of the day I never have to say this again. But for now, I think as advanced certain things are getting with mobile dashboards, and predictive modeling and AI and things. I think what we’re really seeing come to the surface is this gap in the foundations. We need to put a few cinder blocks in there before we start to build the glass sunroom. That’s why I appreciate your work so much, because I think we’re just going to have to keep doing that for a little bit. That’s okay. It’s called job security.

[00:45:27] AE: Yeah. Somebody the other day was like, “Are you worried the robots will replace you?” I was like, “Absolutely not. No. Not one bit. There’s no robots to replace critical thinking skills for any of us.

[00:45:38] LP: In fact, thanks to the robots. It’s probably going to be around for longer. All right. This is our final wildcard question. Think very hard here and imagine this very plausible scenario. You’re backpacking along the coast of Spain with your family in tow, when suddenly you trip and fall into a vortex that pulls you back to the moment you’re about to deliver your first presentation. Because we do talk about that here, too. I’d love to know if you remember, what are you presenting about and what advice would you give to yourself?

[00:46:14] AE: Oh my gosh! Can I give you something that probably you were weren’t expecting? Why not. Why not. We’re recording this on a Friday, I get to be my source.

[00:46:21] LP: Of course, we love twists on a data storytelling show.

[00:46:24] AE: We’ve talked about this one. Before, at my first conference presentation was 2009. It was about dashboards. The presentation was like, “Hey! We should have dashboards. We shouldn’t just have 100-page reports. We should have dashboards.” People get individualized data and it was my first presentation. Nobody was there to hear a little old Ann. Nobody had heard of me. I was like, very entry level person at my first conference. Everybody was there, because the term dashboard was in the conference session table. I was very unpleasantly surprised to be speaking to a standing room only audience along with my colleague too. I don’t know if I could say this. I’ve started the story, I guess I have to finish it.

I wore a dress. I wore a very conservative business casual dress that was about knee length, but I had people sitting two feet in front of me on the floor. And the whole 90 minutes, I was just worried. Are they seeing more than they should see? Since then, I’ve almost always spoken at pants. That’s what I would do differently. I would just, I would have worn pants. Also, I would have brought handouts too. That’s probably the answer that I should have given.

[00:47:34] LP: By the way, handouts.

[00:47:35] AE: By the way, handouts, because we were talking about VBA coding, and like nobody in the audience knew about VBA coding. We should have just set – we talked about these four steps. Like, “Here’s what you do with VBA coding to get started.” We should have just had a one pager that had those four steps typed out, and maybe we did, but maybe we brought 20 copies, not 150 copies. I don’t remember, but I think I just thought, a presentation is slides. Well, kind of but a presentation is also the people presenting and it’s also supplemental materials like slides. Now, I always have handouts. Okay. I’ve given you two answers.

[00:48:10] LP: I love where that – I love the journey that took, but building on the handout aspect, especially if you’re considering conferences. I think absolutely one of the best things to do is to have a handout that’s more readable, prepared, rather than when they inevitably ask you to send your slide deck, which, if you’re designing it correctly, is most likely not suitable to be a printable handout. But bring that and I bet you that people will remember that talk so well, because you’re helping seal your information into people’s brains just by having that more readable version accessible.

Yeah, I love how we just glossed over — no. Thinking about wardrobe, wardrobe is essential too. It’s very important. All right. Well, Ann, unfortunately our time has run out. I have loved this discussion so much. There’s such useful stuff here. Tell the listeners where they can keep up with you if they don’t know already.

[00:49:08] AE: I’m really easy to find online. If you just Google Ann K. Emery data viz, you’ll find my website. I have a bunch of blog posts. I have a bunch of YouTube videos. I have a newsletter. All the usual stuff so we can stay in touch.

[00:49:20] LP: Perfect. You also have what sounds like a pretty robust data viz course library too. Why don’t you share a little about that?

[00:49:28] AE: I do. I started making courses in 2018, which was pretty early for online courses. Nowadays, the technology is advanced a lot of people have courses. But I have courses on data analysis, data viz best practices, how to make graphs and edit them in Excel. Several dashboard classes on the different types of dashboards we talked about. Presentations, reports. I love making courses, so I have all the courses.

[00:49:56] LP: Well, what’s funny is I think we’re in the same course building like entrepreneurial community on Facebook, where we accidentally saw each other in there, where you were like, “Lea?” because I posted something. Then you actually shared some kind of spreadsheet that shows all of the course topics you’re tackling. I was like, “Wow! She is a machine.”

[00:50:17] AE: I think about courses, I have all the diagrams and spreadsheets for the courses as well.

[00:50:22] LP: And I’ll be borrowing some of those.

[00:50:23] AE: Go for it.

[00:50:25] LP: Awesome. All of these links are going to be available on the show notes page for this episode. Ann, thank you so much for being on the show today. Such a long time coming. I’m so glad we make it work. Your work has been such a major influence for so many, including myself, and I feel privileged to know you and get to share your wisdom here. I look forward to some future collaborations with you.

[00:50:48] AE: All right. Thanks, Lea. Thanks for having me.

[END OF INTERVIEW]

[00:51:00] LP: All right! Wow! What an incredible moment to finally have Ann on the show and have her share her brilliant wisdom, very practical wisdom around creating dashboards. Which, rather than where we tend to get caught up with the fancy sophistication of our chart types and interactivity, sometimes it really is about boiling it down to the basics.

To catch all of the links to register for her course, check out her website. For all the resources mentioned in this episode, visit the show notes page at leahpica.com/073. I would love if you could leave a comment, or suggestions or even a question for Ann, because we want to hear about the challenges you face when you’re presenting your vital information. Please, if you’ve liked what you’ve heard, if you’re a fan of the show, please, if you’re on Apple podcast, just scroll to the top and hit that little plus sign to subscribe. Or if you’re on Spotify, you can hit the follow button. Even better, leave a rating and review. Rating and reviews are extremely appreciated because they affect the rankings of the show and I read out my favorite ratings here. Last, if you’re a fan, I welcome you to reach out to me on LinkedIn. I’m very easy to find there. If you’re a fan of the show, just drop a note, make sure you drop a note with your connection invite and I will be sure to accept it. I’d love to connect.

I’ll leave you with today’s little bit of presentation inspiration by my Greek homeboy, Aristotle. That is, “Excellence is never an accident. It is always the result of high intention, sincere effort and intelligent execution. It represents the wise choice of many alternatives. Choice, not chance determines your destiny.” Whew! That might be one of my favorite quotes of all time. As someone who takes a lot of pride and working towards excellence where I can without killing myself, it is so important to understand that when you move towards a path, towards excellence, it really is about identifying all the little choices that you’re able to make along the way. It’s doing things like dashboard and data visualization design not by default, but by design. Do you get me?

Think about all the ways that you can make intentional choices when you are building dashboards, or data presentations, or reports, whatever you have to use to communicate your insights. I promise that the more you identify those choices and make them intentionally, the closer you will get to excellence. That’s it for today. Stay well, stay safe and namaste.

[END]