Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Dashboards That Deliver: A Conversation with Amanda Makulec
128 Plays36 minutes ago
In this episode, I talk with Amanda Makulec about what it really takes to design dashboards and data products that people can understand and use. We dig into why so many dashboards fail, how designers and analysts often misjudge their audiences, and what it means to take a truly human-centered approach to data visualization. Amanda shares insights from her work leading the Data Visualization Society and from her book, including practical ways to think about context, cognition, and decision-making. We also discuss common misconceptions about dashboards, stakeholder expectations, and the gap between technical correctness and real-world usefulness. This conversation is packed with ideas for anyone building data tools meant to inform decisions, not just look impressive.
Subscribe to the PolicyViz Podcast wherever you get your podcasts.
Become a patron of the PolicyViz Podcast for as little as a buck a month
Pick up the new book, Dashboards That Deliver.
Follow me on Instagram, LinkedIn, Substack, Twitter, Website, YouTube
Email: jon@policyviz.com
Recommended
Transcript
Introduction to Amanda McCulloch and 'Dashboards That Deliver'
00:00:12
Speaker
Welcome back to the PolicyViz Podcast. I'm your host, John Schwabisch. On this week's episode of the show, I am joined by Amanda McCulloch, former executive director of the Data Visualization Society and a co-author of the new book, Dashboards That Deliver, How to Design, Develop, and Deploy Dashboards That Work.
00:00:30
Speaker
with her three co-authors, Andy Cotgrave, Jeff Schaefer, and Steve Wexler, who, as you might know, were co-authors of the previous book in this two-book series, The Big Book of Dashboards. I'm very excited to have Amanda join the show. We talk all things dashboards. We talk about how she got hooked up with these three jokers. Yeah, I'm going call them jokers. these three jokers to write this book. We talked about how they solicited examples, talk about the process, how four people are going to go through dozens of examples of dashboards to come to an agreement on what should be included in the book and then how to write about them.
00:01:09
Speaker
And then we talk about what I think is a real value add of this book. And that makes it such a nice follow up to the big book of dashboards, which is the process of creating dashboards, both good dashboards and bad dashboards, but really more following on how a team or an individual can produce, develop and publish a more accessible, a more useful, a more beautiful dashboard. So we talk all things dashboards, we talk all things data visualization. We go pretty deep in this conversation and if you are in the world of creating dashboards, I think this is the conversation for you. So
Amanda's Background and Data Visualization Society Involvement
00:01:49
Speaker
let's head over to my interview with Amanda McCulloch here on the PolicyViz Podcast.
00:01:57
Speaker
Hello, Matt. Oh no wait, not Madison. Hello, Milwaukee friend. ah Hi, John. How are you? Oh, good. i always um I mean, I had Wisconsin in my head, but then, you know, I immediately go to Madison. It doesn't help when there's two major cities that are both starting with an M. And as I appear, I have heard that Camp Randall Stadium is the fifth largest city in Wisconsin. when it is filled at the university of wisconsin madison because of how small the other i could see that yeah because it's probably like 80 000 at this point yeah something like that and they're also like both cities are like at the same latitude like an hour across an hour separate yeah it's pretty good yeah it's great um how are things in wisconsin wisconsin is fabulous i love living in wisconsin it's uh it's so you know, calm, quiet, peaceful here. And we're getting absolutely dumped with snow this winter, which yeah my kids are loving. So we can complain about that. the the snowball The snowball and snowman making potential is much higher with Wisconsin snowfall than it is in DC. You got to run out but there, even got to run out there and like catch it in the first like inch of snow and try to make a tiny little teeny tiny snowman. But it's been great. We really like it. You like a day and then it's gone.
00:03:08
Speaker
yeah i date Yeah. A day and it's gone is basically it. Then it's just
Creating the Book: Process and Practical Steps
00:03:11
Speaker
gray slush and then it just feels sad. So yeah it just kind of I like blanket the blanket of snow we've got here. It looks very winter wonderlandy. Yeah, yeah, yeah. Um, okay. so you're out in Wisconsin now.
00:03:23
Speaker
Great new book, dashboards that deliver. Um, it's a follow-up to the big book of dashboards. Um, let me ask you two questions simultaneously. First, for people who don't know you, maybe give a little, a little spiel on your bio, but then the more important question is, um, how did you get hooked up with these jokers, Andy, Steve, and Jeff to, to, to, to get in on the, on the second book?
00:03:46
Speaker
Great question. So first, I am Amanda. For those who don't know me, I am a health data visualization designer based in Milwaukee, Wisconsin now. But I was in Washington, D.C. for more than 15 years where I helped to build out the local data viz D.C. community. And it was where I was when I helped to be get the data visualization society off the ground.
00:04:09
Speaker
um I was a founding board member for DVS and served as their executive director for four years. And so I'm a big fan of any way that we can build community in the data visualization space. I also probably am known for writing and talking a lot about visualizing data responsibly, especially in the context of health and healthcare. care and spent a lot of time speaking around that topic during the COVID pandemic. So it was exciting actually coming full circle with dashboards that deliver because we actually have the John Hopkins COVID dashboard as one of our scenarios in the book, which was so exciting to get to get to dive into as a public health data viz person.
00:04:44
Speaker
Yeah. i know but But you want to know about the jokers and how I got- I to know about the jokers, yeah. I mean, I'm sure i'm sure they're cringing when they hear that, but I'm going to- Your word, not mine, John, so that's okay. Yeah.
00:04:56
Speaker
So I actually met Andy and Steve together at the tapestry conference in 2018. You were there too, right? Weren't you speaking at that conference? you Yeah, probably. Well, there was one in Nashville and one in Miami. Miami. Yeah, that was the Miami one. So Miami one where back in the day, Tapestry, for those who don't know Tapestry, yeah was a great, really kind of niche, smaller data visualization conference. Tableau co-sponsored it, I think at the time. And the 2018 one was hosted at the University of Miami by Alberto Cairo and his crew there. They're great speakers. i mean, I think Cole was speaking. Elijah spoke and did his third wave talk. And they had an open call for lightning talks or Pecha Kucha talks where they called Pecha Kucha talks. yeah And they said they said, you know, we need some more people to do these five minute Pecha Kucha talks. And
00:05:50
Speaker
Being the, I think I was five months pregnant at the time with Caspian and just like lumbering around and came to this conference and had never been before. I was like, I like to give talks. I'll give a talk. And so I did this short little five minute talk about the mediocrity of ah medical test result data visualizations that get sent out to folks. and talked about like how awful some of those reports are that you're getting bad news. And so told it from a very personal perspective in terms of my own experiences. And it's all just like seas of red with no good kind of indication of risk ratios for you compared to other people and the general population. And after I gave that talk, Steve Wexler, he will tell this story too, decided he's like, I would like to know her because she talks about how to make data practical and easier to understand. And so he introduced himself to me and we had a nice chat at Tapestry. i got to meet Andy, who we may have crossed paths in the past, but met him at some of the happy hour stuff that was happening at that conference. And a little while after that,
00:06:55
Speaker
Steve reached out and invited me to go join him and Jeff as a guest on Chart Chat. And they they brought me on as a guest. We talked about some really interesting data visualizations that I think Mona Chalabi had done recently. i think This was back in 2019. And I hadn't realized that it was an informal audition for being a Chart Chat co-host. And if you don't know Chart Chat, Chart Chat came out of the first big book of dashboards as a space for Andy, Jeff, and Steve to kind of continue on the conversations they had in the book writing. And so after that first Chart Chat, I got invited to be a recurrent co-host on Chart Chat and we're hosting Chart Chat number 66 this month. wow So we've kept it going. And so when a new the idea for doing a new book, whether it would be a version two or a follow-on to Big Book of Dashboards came up, they were kind enough to invite me to be part of the writing team and continue our collaboration. They liked me enough to write with me as well as inviting me to chat with them every single time. Yeah, because the writing part is a lot more ah lot more of ah of a dedication to being with them. That a two-year runway, I think, from when we actually did the book proposal and got it all approved to when we actually had everything come out in print. Because
Diversity in Dashboard Submissions and Selection Criteria
00:08:08
Speaker
collaborating with four co-authors where every one of us actually did read and get input on every single chapter, even though there's lead authors noted on the different and chapters and people who have strong opinions about things, took some time, took some time. But not only that, for folks who haven't written a book specifically or particularly a data or data of his book that is image heavy, like the part of getting the images is kind of its own...
00:08:37
Speaker
universe of tasks. Yes. We have 293 images in that book. And I know, cause I know where the folder is and I know the count cause we had a ah spreadsheet tracking, you know, yeah where, yeah where are all of the images? Like what's the status on the image? Do we have the right file? And it's funny, it's things you don't think about too, that it's getting the images all in there in your draft. and You've gone through this with your layout that's maybe in like a Word document. yeah And then they do the layout of the book. And we gave very specific instructions. And shout out to all three of my co-authors who were able to give clear guidance to Wiley saying, here were the problems that happened last time where, you know, you flip open the book and you're reading and it says figure like 2.7. And then five pages later, you finally find the sign, figure 2.7.
00:09:24
Speaker
yeah And so it's important, I think, as you're thinking about the book process that the first draft and having all the images there is one thing. especially all these screenshots from different dashboards, but then actually getting them laid out in a way that makes for a good reading experience becomes an entirely separate kind of part of that process that took a lot of optimizing and sometimes rejigging kind of aspect ratios and organization and graphics and images where we could. So things would fit together better on these two page spreads in this square book that we have.
00:09:57
Speaker
Yeah, and my, I think it's both of my books with Columbia, Better Data Viz and Better Presentations, the charts are not numbered. Because I just personally like that experience of like, okay, in the figure below or a figure on the facing page, which means the layout thing is much more difficult, right? Because you have to make sure when you say the image below, it actually ends up below.
00:10:21
Speaker
But what I didn't realize is that that strategy doesn't work for e-readers. No, we specifically have it. So our e-reading copy, if you buy a digital version of dashboards that deliver, you will be reading through basically a PDF of the exact layout yeah that we have in the book for that reason. yeah Because even with kind of orienting and making sure the figures are on associated pages, A lot of effort went into that. And that would be very difficult if you were scrolling on an e-reader and realizing that everything moved around on you. Yeah, yeah, exactly. Yeah, you could, if you can change the font size, if you could, you know, you can zoom in and zoom out in an e-reader saying the you know, figure three is over, is next on the next page. Well, next page doesn't mean anything in the electronic version. Anyway. Yeah.
00:11:09
Speaker
Okay, so let's talk about where the examples came from. Because my recollection is, and I think they did this with the first book, right? Was a public solicitation for examples. And when people submitted, you all went through them, selected out. You can tell us what that was like. Did you then reach back out to people and say, you made it and this is what we're going to do? Or we said, you made it, thanks a lot. And you'll see the final version when it comes in print. gosh, that would be a scary thing to say to somebody about presenting their own work. I would imagine we would have had lots of people opt out at that point. Yeah, I would think my JotForm submission is the basis for a whole chapter in the book. I think it's probably worth saying to those who haven't peeked at the book that the book is broken into three sections, right? So part one is an overarching kind process around data visualization and dashboard design, very agile user-centered.
00:12:02
Speaker
Then the second section is these scenarios. And there's 15 chapters that are each a single dashboard and the detailed story about the process through which that dashboard got developed. And so we can dive in a bit more look at like what's in those scenarios. But the third section is some kind of shorter essays on key themes and data viz right now, Gen AI and data viz, those kind of things.
00:12:24
Speaker
When we did the call for submissions, it was, you are correct, similar to what happened with the first book where a call was made for submissions for a big book of dashboards.
00:12:34
Speaker
We had a much larger response this time around. I mean, Steve, Jeff, and Andy are now well-known and have a best-selling book on dashboard design, right? So we got a much larger response. We got upwards of 130, 150 submissions in terms of total submissions in for people pitching dashboards that they'd like to have included.
00:12:54
Speaker
We also at the same time went and said, what are the dashboards that we think should be in this book? That's how the Hopkins one ended up in. We reached out to them directly and said, we would be really interested in having the John Hopkins COVID dashboard in this book. Just as that moment in time, something widely used, 2.5 billion page views over the life of the dashboard, like, why not? So we did reach out to some folks directly as well.
00:13:18
Speaker
But it was interesting because as when we went through the submissions, we actually spent an extra day after the Tableau conference sitting in Jeff Schaefer had an upgraded suite with like the executive suite with a big fancy room in it for us to sit in at the Hilton at the Tableau conference. And we sat with a computer hooked up to this big screen, and into the big TV screen, going through and actually reviewing all of the different submissions and identifying kind of a first call, which ones kind of make the cut or don't make the cut for consideration. And then talking more granularly about different ones. And before that, we had we tried to classify different submissions by topic, right? Like the first book is a lot of business-y kind of dashboards. This book has a lot more diversity in topic and type. So you have dashboards on things like healthcare care analytics, you've got things on banking, you've got things on KPIs. We've got things on professional racing teams. It's very cool. But we wanted some diversity on types. And one of the ways in which we were going through and culling dashboards initially was looking for dashboards that actually functioned, and this is going to sound so obvious, but functioned as dashboards rather than being data visualizations
00:14:32
Speaker
built in a tool like Tableau on a panel called a dashboard. Like, you know, if you if you're working in Tableau, you open up a a dashboard, but people will build scrolly telling pieces or really great infographics in Tableau on a dashboard, but they don't necessarily make a great fit for a visual display of information meant to monitor conditions or facilitate understanding.
00:14:55
Speaker
They're more story-driven. They're more information presentation. And so that was one of the things I think that surprised us a bit going through the submissions was there was a not insignificant number that were amazing data visualizations. Like I think some of them got submitted for IIB awards and things like that. yeah They weren't really dashboards. And because we wanted to focus not just on what is good or bad design in this book and focusing on that piece. We also went ahead and we said, do we think this dashboard or does this dashboard based on what someone filled out on the form, have a clear story of how it got built and have a clear kind of use case for how it's being used in practice. Because people build lots of really interesting looking dashboards for open challenges or makeover Monday or whatever else that could be interesting design conversations.
00:15:45
Speaker
but they were not necessarily interesting process conversations around how a team or a person built a dashboard and then how it actually got used. and So we really looked for examples that had a real story to them and really were being actively used in the real world. And think one or two scenarios were more prototype or template type dashboards, but most have really interesting stories that go into adoption and use which I think goes beyond the scope of a lot of other dashboard design books.
00:16:16
Speaker
Yeah.
Understanding User Needs and Measuring Success
00:16:17
Speaker
I'm curious about the diversity of topics or sectors. Sure. Did that sort of naturally fall out of this decision process or so so you have this 150, let's say there's a whole bunch of them that are just infographics or not just, but you know, they're not. Very cool, but not great for the. not dashboards. Right. So we throw those out and then you have this, you know, set less of left or whatever it is.
00:16:41
Speaker
did you Did you all consciously say, okay, so we have, here's a dashboard that's you know public health versus you know business. let's Let's just use the public health one because that gives us more variety.
00:16:53
Speaker
Or did it just naturally was so long ago. I don't think we said no and dropped anything because we had too much of something. I think it kind of naturally fell out because I think that's what you're asking. Was there like a point at which we had 30 people? And then we're like, oh, we're going to drop these two because we've got too many business examples. And I don't think that ended up happening. We had a a little bit of attrition from our shortlist based on just people like we'd one that we dove down the whole writing process and then they couldn't get internal approval on sharing it, those kind of things. wow So we had some instances where we had some attrition there. But I think the overall makeup and kind of diversity of the different dashboards, both in design,
00:17:32
Speaker
And in topic, I think worked out pretty well. The one miss I will say is we, no matter how hard we've tried, did not get a lot of Power BI submissions. And so I was hoping we'd have a little bit more representation from Power BI dashboards in the book, just because I think we're associated as being people who are more connected to Tableau, but so many people use Power BI. Right. But we didn't get a lot of submissions despite a lot of nudging and asking. And I think that's in part because we've seen that community, especially in the last couple of years with some of the work Shannon Lindsay and others have done to really build that community structure. I think that I've seen a lot more of that kind of community engagement. And maybe if we made that same call now, two years later, we may have gotten more submissions. Who knows?
00:18:16
Speaker
I mean, maybe we could have a whole conversation about the different development of communities between these different tools. sure But but to this to this point, so what share of the dashboards that are in the book are Tableau? And what are some of the other tools that you ended up? I mean, the Hopkins thing was custom.
00:18:34
Speaker
Hopkins thing was custom, but it's in Esri, I think. I think their front end is Esri. Oh, yeah, that could be. Yeah, that might be right. Yeah. think so. um I think about a third are Tableau. About a third.
00:18:45
Speaker
yeah maybe more, but about a third are Tableau. There's some custom D3 dashboards. There's some that there's an S3 one in there. Maybe more like half is Tableau. I'm like trying to mentally map my way through all of them. I mean, the the NASA Earth Information Center is a whole kodgepodge of various different data visualization libraries and codes and types and things all aggregated together with a a wrapping kind of tool that lets them put things that were built in different spaces together.
00:19:14
Speaker
i am not technical enough to be able to explain that in great depth, but you ask Alex Gervich or Marcy Barrow or Alaini Costas and one of them could probably tell you more. But it was a good mix. I think it definitely veered more towards either Tableau or Customs.
00:19:29
Speaker
for the most part, and versus having other kind of bespoke data as tools like Click or Power BI or something else. yeah For me, that kind of, I think, tracks a lot of what I feel like I see, which is like and and the public versus sort of the internal. So it doesn't surprise me necessarily because I feel like Power BI especially is like a lot of internal stuff because it hooks into Microsoft so quickly. And then Click as well. I mean, there's not like a huge external Click community the way there is with- Well, and Click is a lot of healthcare that I think gets more complex. yeah Like we had to do a lot of work with Lindsay Betzendahl to make sure that any representation of the Children's Hospital Association dashboard, even where they had used like made up hospital names, everything he says actually Betzendahl Hospital now on all the examples. Because we wanted to make it very clear that this is not anyone's actual or real data. And there's a lot of kind of hashes or just placeholders instead of numbers, just to make sure that it was as masked as possible.
00:20:26
Speaker
Yeah. um Okay. So the other thing that's that's interesting, I think, about this book is it goes beyond sort of like what's good and what's bad. i mean, you talk about that a little bit, but it really...
00:20:39
Speaker
I think where it kind of pushes the boundaries of what the field is doing is it talks about teams. You've mentioned agile processes, talked about teams and processes. So like what in your mind from that part of kind of the main takeaways that, that readers should expect.
00:20:56
Speaker
think the biggest one is that you should be spending time understanding who's going to use a dashboard or data viz tool more broadly. You should understand who's going to use it, how and why, before you even start building something.
00:21:08
Speaker
And that sounds so obvious to me because I think almost every good data viz class, you mean, you teach data viz classes, John, starts with like, know your audience. You should know your audience. Think about who you're communicating with. No duh. But dashboards, when you're building a tool, you want someone to engage with. beyond just like reading the kind of key takeaway of a chart, I think has a higher bar and requires us to have a little bit more of a product mindset where we think about how we're building something we want someone to engage with and use.
00:21:36
Speaker
So it's not just about having those early interviews and conversations and mapping out a list of requirements, but engaging with those people where you can as partners, or at least engaging with their proxies as partners so that they're giving input throughout the design process.
00:21:51
Speaker
You want to have not just that early stage interview, what do you want on this dashboard? You want to ask better questions than that. But you also want to make sure you're showing people examples. Do you prefer things presented this way or that way? Which is more analytically intuitive for you? Can I watch you interact with this dashboard the way I've designed it and see where you get stuck trying to answer certain questions? And by engaging with people as partners, you help to cultivate champions who are excited to use the dashboard. You help to engage users that they feel some ownership. And when the dashboard actually goes live, you're much better suited to have wider adoption and use.
00:22:27
Speaker
So I think one piece is that piece of engaging with users. And we give specific tools and activities and points that you should think about in terms of when and how you do that. And the scenarios talk a lot about that as well in terms of how people do that in practice.
00:22:39
Speaker
I think the second piece that ah comes out really strongly in the book, and I would hope people take away from it, is the value and importance of early rapid prototyping. And of spin things up, draw sketches, create initial wireframes and mockups to get the overall flow and structure right. Before you spend time kind of pixel perfecting every single thing and kind of working off of a maybe even sample data or a real data model. Oftentimes you're modeling your data as you work through your requirements to figure out the granularity of the data and all those different things that impact performance.
00:23:11
Speaker
And I think that the third big thing I would say people should take away is that when we think about how we measure success of dashboards, measuring success is not just about speed to adoption. And I think we make the and and speed to release, right? yeah So we have all these tools that tell us how quickly we can build amazing new dashboards. And at the end of the day, it's not about speed to getting a dashboard up and out there, but you don't want to drag your feet for six months if you can avoid it. but instead thinking about kind of how else are you measuring success? And we talk about little s success, like the monthly active users and other kind of typical digital metrics you might get off of a Tableau server. And those are the outputs in my world of monitoring and evaluation that we might look at. yeah
00:23:51
Speaker
But you also want to find ways to actually ask your users about how the dashboard is being used in their work and getting those kind of big S success impact measurements that allow you to say when someone says, well, we spent all this money and time on this dashboard. What did it do for us? Which is a very fair question, especially in a world that's very enthusiastic about where can we automate away things with ai And we want to be able to say, well, we built this dashboard and not only did it save like four hours a week in manual reporting processes that someone was having to do. But it also enabled people in X, Y, and Z meetings or other spaces to immediately have access to data points they need to make such and such decision. And for me, this all comes kind of full circle back to my roots working in public health and global health, where for a very long time, until I really transitioned into more of like the tech consulting world in 2017,
00:24:46
Speaker
My title was not data visualization designer. I was a visual analytics advisor. I was a monitoring and evaluation associate. I was someone focused on these data demand and use projects where data visualization and dashboard was one tool we used to help get people more interested in using data and more actively using that data for decision making. And so for me, that kind of question of how we measure success comes back to those pieces. where success isn't just about counting views. I'd rather have three of the right people looking at a dashboard than 300 people casually glancing at it because they think I made a cool sankey diagram.
00:25:24
Speaker
So how do I get the right people looking at it that can actually use it and are excited to use it? And maybe that even means we're automatically emailing them the PDF on a weekly basis. yeah I'm on board with that. I'm like, I'm the terrible person in the world of dashboards where I'm like, whatever helps you start to use this tool more, screenshots adapted into a slide deck, I'll help you make those screenshots better and annotate them effectively. You want to have an email to you each week? Fine. You want to a self-service table as your last page in the dashboard that you can export to Excel?
00:25:53
Speaker
I'm on board as long as there isn't significant security issues. yeah I want to make sure we're building tools that actively get used rather than striving only for dashboarding perfection. I agree with everything you said.
Feedback Mechanisms and Relationship Building
00:26:05
Speaker
So this, Lionel, do you want 5,000 people or the right five people? I think is a key question.
00:26:11
Speaker
when I talk to people about this, I think a lot of people say, well, how do I get that feedback from people? I mean, I think that's a big question because, you know, as we all know, getting someone to just go to the dashboard or the website is hard enough, but then getting them to even like,
00:26:27
Speaker
you know, send, you know, fill out a form or something. So either from your own experience or for the folks um that you worked with in the book, what are some strategies that people use to sort of do that feedback solicitation that, you know, that to try to measure impact?
00:26:45
Speaker
So I think one is kind of the structured form approach. You can have that and make it as close to the dashboard as it can be. Don't bury it in the documentation on page number 52. Go ahead and actually have a link on the dashboard homepage or on the landing page or even in the corner of each page that says, you have feedback for us that you want to share? And have some ways of making that into a good data collection mechanism where first someone says, I'm here to submit feedback or stories or something positive. I'm here to submit data request that there's a bug or issue, data quality problem or design change. or new feature and find ways to create structure in the form that makes it more usable, but doesn't make it cumbersome to fill out.
00:27:28
Speaker
The second is honestly, if you can, going and looking at who your users are over the course of a period, that period might vary. Like when I've worked on USAID dashboards before where data was coming in every quarter, it wouldn't make sense for us to go check and look at things on a weekly basis in terms of usage because the data was only being updated quarterly. So you look at it on a less recurrent basis.
00:27:50
Speaker
Look at who's using it. Reach out to people who are those kind of power users and ask them if you can have 20 minutes on their calendar to ask a few questions about it. So most people don't like filling out forms, but they'll talk about yeah what that looks like. So have a conversation with those power users.
00:28:05
Speaker
I also think when you're doing that analysis, it's good to, if you had specific maybe senior leaders in mind who'd use it, look and see who's not on the list. And if there are people who are not using that dashboard who you expected to, especially if they were involved in the process, reach out and ask them why too, to get better feedback from an anti-pattern perspective.
00:28:24
Speaker
But I think that having kind of a clear mechanism for getting feedback with a form or other structured approach, And actively pushing out and asking, don't just wait for someone to come to you, either with those meetings where you have meetings with targeted people, or you push out a quarterly email that's like, we're just collecting short stories.
00:28:41
Speaker
They just need to be a few sentences. And if you can just get someone to give you a little bit of information, you then know who you can follow up with for more details. yeah But I always think that it's those conversations where those kind of magical moments come out, not usually through a form.
00:28:55
Speaker
yeah Yeah, for sure, for sure. I would add to that. in sort of a longer term thinking, like building relationships. If you have, if you know that you're, you know, like, like for my day-to-day job, right? Like if ah if I'm leading a project where we're doing work in the city of Baltimore or something, like building the relationships with people in the city, advocates, nonprofits, government, and and then being able to sort of like say, you know, in a year from now, oh, we've had this relationship. We put out a new dashboard last month. Like, have you seen it? Right. And sort of like the relationship building, I think works here. Well, and I think that when you when you talk about those relationship pieces too, I think that that speaks to how big is the user base for a dashboard. For some dashboards, you're really kind of aiming at those five people and you know who you can reach out to. In other cases, you're really kind of casting a much wider net, which might require a different strategy in terms of how you approach that. I think also if you have a plan for when you're going to make any enhancements or changes, and I'm a big fan for advocating for the The dashboard designer should not just be constantly making little bitty changes to dashboards. A, you can break things. That's not good. You could break things. yeah B, takes a lot more time to QA. C, context switching is a real challenge in terms of productivity. So batch all your kind of update ideas.
00:30:12
Speaker
prioritize them in terms of what's going to get done and what are the features people want most, and then make those changes. And when you're about to make changes to a dashboard in that structured way is a great time to reach out to users because you can both give them something they might want, remind them that you're looking for input for changes to the dashboard. And by the way, if you've been using it, We'd also love your stories of what's worked well to make sure we're not changing anything that you really like. So there's kind of a carrot piece to that. And the closer, and on internal dashboards, the closer that the data team is to being part of the day-to-day operations of whatever the business unit is or the team is, the more they're a partner who are hearing the stories. They're going to be in the meeting where someone pulls the dashboard up on a screen, right? They're going to be there and actually see it in practice in theory, or they're to see the chart from the dashboard show up in a slide deck.
00:31:04
Speaker
which is where I think it's great to think of data teams as partners, not just as a service desk that someone's going to, that then their charts and dashboards go somewhere else. Yeah. um
Planning with Sketching and Wireframing
00:31:14
Speaker
The other thing I wanted to ask you about was um you mentioned wireframing and drawing as part of the process. And I think that sort of sentiment comes up a lot in in these conversations of good data viz development, which is to draw and sketch and try. And I think a lot of people, in my experience, kind of struggle with that a little bit because they talk about, i need to understand the data.
00:31:38
Speaker
um How do I draw before I understand that fully understand the data? So in in your experience, again, either your own or with the folks that you work with in the book, like how does the drawing, the sketching, the wireframing fit into this process of really understanding the data before you can figure out what the graphs are, you know even before you can sort of lay them all out in ah and a view?
00:32:01
Speaker
Yeah, I mean, Andy has a great example story and vignette in the book that is functionally like, I drew the most perfect dashboard ever, then we connected it to the data. And it's like, there's one big bar and everything else is teeny tiny. And he's like, and it's like, there's no insights, the scatterplot that had this beautiful kind of correlation on it that you could see. He's like, it's just chaos. And so he very much said, like, you can't wait for forever. to translate your wireframe into something with real data, because you may find that your data isn't coming through in those stories the way you want it to, and you have to. Right. So but I think that if you spent time mapping out what kind of those initial goals and requirements are, what does the dashboard need to answer? I like user story structures where I'm like as a someone. I need some kind of information in order to do something. That often gives me some clues as to either how I might structure the pages in a dashboard where each page responds to a different user story. or looking at kind of the you an individual views in a dashboard. But I think that the power of wireframing especially is getting the overall information flow and architecture right first and giving you the shell that you can start to fill in. Instead of working with a blank screen or canvas when you pull up Tableau or Power BI or something else, you're saying, okay, so here's my kind of series of boxes.
00:33:17
Speaker
In those series of boxes, I know I need some KPIs across the top because it's just kind of good design practice and dashboards to have some KPIs with some context on them. And then I know that one of the things they need to know is something that's around geographical kind of information and where things are geographically. So there's going to probably be a map here. And the more you start to kind of just map out ideas and maybe just write them, like what's the chart type that responds to things or the boxes actually having questions, this question is going to be asked here. And thinking about starting with the most important stuff in the upper left, then moving down in that kind of Z or F reading pattern, you start to create a shell that you can start to fill in with different charts or graphs that actually will make
00:34:00
Speaker
The process of building out that dashboard faster later. Now, there are some people who are super confident and comfortable in, say, Tableau, where they want to prototype in Tableau and and more power to you if that's where you want to start. But if you're going to get stuck thinking, i can't make that like beast worm plot or dumbbell plot because I don't know how to make it in Tableau.
00:34:20
Speaker
That's where the power of actually drawing things out first comes in, because then you can go find the guidelines on how you build that chart type in your tool of choice. And so I think that starting with like boxes, don't try to start with trying to sketch the charts, because I think that that sends you down a path that can actually be really tough to do because you feel like you don't have all the information. But start with kind of the boxes and the organization of information and the wireframe before you even get to that space of sketching charts and prototyping.
00:34:48
Speaker
Yeah. um
Book Structure and Reader Approaches
00:34:50
Speaker
so So to the book, you mentioned earlier, there's these three parts, process, scenarios, and then and then the real world piece. um Obviously, different readers are going to come to this with different experience, different levels of skill. how like um i Maybe this is a two-part question, but like...
00:35:11
Speaker
Is this a kind of book people read cover to cover? Is this a kind of book people can drop in and drop out? Where would a begin where would you recommend a beginner start versus someone who's you know more experienced? We debated this actually on if we, as authors, did we write this? We kind of talked about who our prototypical readers would be, right? And there was some debate about, is this a book for people who are new to dashboard development? Or is this like the next year intermediate book for people who do dashboards and want to build them and make them even better? yeah i my perspective is that this is the book I wish I had 15 years ago, as I kind of spent a decade learning and cobbling together a lot of things from user-centered design and how to use agile approaches to manage different data projects and different pieces into the process piece. So I would say if someone's new to dashboard design, the first section of the book, which I know it's a lot of pages, but it's lot of pictures too. Like I said, 293 images in the book. There are a lot of pictures. So I think that I would start with reading through that process piece in part because it sets you up with the language that we then use in the scenarios to describe different parts of the design process.
00:36:25
Speaker
Because in each scenario, we don't go through and say, like, here's the discovery phase and here's the prototyping phase, here's this. We picked and chose what was relevant and interesting.
00:36:36
Speaker
and was a good learning moment for each of those. It isn't every single part of the process for every single scenario. So if I was a newer person and I was using this as a learning book, I would start there. If I was using this as an experienced dashboard designer, i would think about if my goal is to gain inspiration from real world examples or think about different ways to approach my design process. In which case for the first, I would start in section two and look at the scenarios and you could read those chapter by chapter, pick one that's relevant. There's a section at the start of each scenario that says specifically, here are some other similar examples from entirely different fields. So if you look at it and say, when am I ever going to make like a racing team dashboard for an IndyCar team?
00:37:20
Speaker
When am I ever going to do that? I'm not. That's not going to be in my job description anytime soon. But the idea that you have some kind of map of where things are happening geographically and then associated distribution type plots that show more granular detail is actually really relevant if I'm looking at things like vaccine stockouts and supply chain issues across a country or a region.
00:37:41
Speaker
And I want to see high level geographic information on ah some kind of map and then the more distribution based information separately. And so I would suggest looking at the scenarios and looking at those correlation pieces. I also think established designers might really appreciate and enjoy some of just the foundational layout stuff that's in, i think, chapter maybe chapter eight. I should know my chapter numbers better.
00:38:03
Speaker
But where we talk about kind of core principles of dashboard design and layout, that can be really helpful, especially if you're working on a team and making sure you're all kind of designing with the same framework in mind of how you lay out and approach dashboard design. And i'm not trying shortchange section three. Section three is great, but you can read those as individual essays, basically. So right read it cover to cover if you want. But I mean, it's a it's a kind of chewy book. Yeah, yeah, yeah. So hop around, you can hop around, and I i hope people like stick Post-its in it and use it as a reference book that serves them when they need it.
00:38:34
Speaker
I mean, I do kind of feel
Conclusion: Resources and Future Learning Opportunities
00:38:35
Speaker
like that's the newer, um this kind of new wave of data, data of his books is less sort of cover to cover and more reference, because I think it is more, the field has evolved, people are more experienced, so you have this opportunity now to be like, I need to work on this part of my skill set, and so yeah that's chapter Well, and I would say, if you do skip to the second section, you could just read chapter like two that lays out the design process, and you'll get the language that we use. So when we talk about discovery, that's ah that's our way of describing that whole kind of initial section where we're digging into user requirements and mapping mapping out the vision for the dashboard and doing all those things. And so when you see discovery later, you'll know what we're referring to, which I think is always helpful if you're reading a book.
00:39:19
Speaker
Yeah, I think that's a good point because there is stuff. I mean, I'm just flipping through it right now, right? Like there's a description of the agile software development. There's the planner versus the optimizer. Like, yeah, just knowing the languages. I mean, and also like, I think probably most people read the first couple chapters of every book. and They can get the, anyone who wants it can get the first three chapters for free on our website.
00:39:38
Speaker
Okay, so this is what I wanted to ask. So um people can get the book where you know wherever they buy their books. And obviously, they should also get the first book, Big Book of Dashboards. um But what else? can They can also, you mentioned Chart Chat, so they should sign up for Chart Chat so they can watch that.
00:39:54
Speaker
um Okay, and so there are three chapters they can get for free? Yeah, so the first three chapters, including the framework chapter, are all a free download on dashboards that deliver.com. There's also a Dashboards fan club that was our easy catch-all way to let you sign up once and be able to download extensively any resource that we publish.
00:40:12
Speaker
Right now, there are various different templates and tools to support and facilitate discovery processes, and we'll have more of those resources being released over time. So it's a great place to go if you want downloadable resources that go along with the book.
00:40:24
Speaker
Awesome. Where else should people go? What um what else um should they be prepared to learn? I think you have an online course coming out, I do. So I'm going to be launching an online course on dashboard the dashboards that deliver framework, designing dashboards that deliver on Maven coming up in 2026. So if you're interested in getting updates on when that course and other workshops or live talks that our author team are doing. You can always sign up for our newsletter on dashboards that deliver.com and you'll be the first to see when those signups go live.
00:40:59
Speaker
But we'd love to have folks join to go ahead and walk through and do some hands-on learning around the dashboards that deliver framework. if you're someone who's thinking about how do I put this into practice and wants to have a little bit more of that engagement and conversation. around how we approach building great dashboards and data tools and data visualizations more widely. Would love to have folks joining that conversation. It'll be a cohort based course so that people can actually engage with each other for some peer to peer learning and conversation, as well as getting to learn from some of my experiences and some of the content in the book.
00:41:28
Speaker
Awesome. Well, I'll make sure to put the link to dashboards that deliver.com in the show notes. so listeners, you don't have to memorize it. you can just click and you can go over there. And congrats to my co-authors. I know I'm sitting here alone, but I have three amazing co-authors who all deserve their fonders too.
00:41:44
Speaker
No, that's all right. we don't We don't need to congratulate them. They couldn't take time out of their day to come here. It's fine. The Joker is the other Jokers. All right. It's great to see you.
00:41:54
Speaker
Thanks for coming on the show. Always. Thanks, John.
00:41:58
Speaker
Thanks for tuning in everybody. Hope you enjoyed that. Hope you'll check out the new book, Dashboards That Deliver by Amanda McCulloch, Andy Cockrave, Jeff Schafer, and Steve Wexler are now available wherever you get your books. And if you have a moment, sign up for my sub stack where you can get more information on books that I think you need to be a good data visualization developer and creator. And of course, if you have a moment, click those five stars wherever you get your podcasts, rate and review the show so that I know you're enjoying it and helps me get better and more guests. So that's all I've got for you. Until next time, this has been the PolicyViz Podcast. Thanks so much for listening.