Tableau Public Tableau User Group 17 Sep 2026

0:10 20 minutes on the clock.

0:14 The pressure's running high.

0:17 Goodreads data on the screen.

0:20 Let the creativity fly.

0:25 Analyze design and tell the data story. One shot on the stage, gotta claim that glory. Three contenders, one big stage. No time to overthink. From blank sheet to brilliant story. Faster than you can blink. It's Iron Viz. Let the data come alive. 20 minutes, one big chance to visualize. Make it smart, make it shine. Make a viz that sticks. Analysis, design.

0:44 Tell a story, do it quick. Three authors, three chapters, three stories to spin. 20 minutes to build it. Let Iron Viz begin.

0:55 Uh. First up, Kevin Wee busting myths on the page. Are graphic novels just for kids? Nah, they're all the rage. Fiction, fantasy, sci-fi, watch those numbers ignite. Then a custom book chart brings a story to life. A clever little masterpiece built right from the start. He made the numbers jump right off of the chart. Showing graphic novels have much more in store.

1:13 They're not just superheroes anymore. It's Iron Viz. Let the data come alive. 20 minutes, one big chance to visualize. Make it smart, make it shine. Make a viz that sticks. Analysis, design. Tell a story, do it quick. Three authors, three chapters, three stories to spin. 20 minutes to build it. Let Iron Viz begin.

1:34 Brian Moore came in with the battle of the books. Hemingway versus Hogwarts. Let's take a look. Started with dots, seemingly simple and plain. An animation kicked in and changed the game. Turns out Hemingway classics like The Old Man and the Sea got nothing on genres like modern fantasy. Wait, what's this bar chart? Did he lose the plot? Nah. Finished early, had more time to alight.

1:55 It's Iron Viz. Let the data come alive. 20 minutes, one big chance to visualize. Make it smart, make it shine. Make a viz that sticks. Analysis, design. Tell a story, do it quick. Three authors, three chapters, three stories to spin. 20 minutes to build it. Let Iron Viz begin.

2:13 Once upon a time, there lived a woman named Ann who entered Iron Viz with a Venn diagram, but everything changed when the genres combined. She turned to a journal and drew outside the lines. Romance, fantasy, maybe something dark. She maps the mashups and finds what makes them spark. Her hand-drawn design made the whole story sing. She cracked the code and became the Iron Viz queen.

2:35 Queen, queen, queen. Queen, queen, queen, queen.

2:39 Q- q- q- q- queen. 20 minutes, clock is ticking. No more time for overthinking. Pages turning, data finding. Build the story, keep it climbing. Drag it, drop it, make it hit. Find the insight and commit. No rewrites, no second plot. 20 minutes, give it all you got.

2:52 It's Iron Viz. Let the data come alive. 20 minutes, one big chance to visualize. Make it smart, make it shine. Make a viz that sticks. Analysis, design. Tell a story, do it quick. Three authors, three chapters, three stories to spin. 20 minutes to build it. Let Iron Viz begin.

3:12 Let Iron Viz begin.

3:26 Rob Taylor: Hello, everyone, and welcome to this September Tableau Public TUG. We've got a huge Iron Viz special for you today. I'm wearing my presenter's jacket, uh, replacing Dua Lisa, who is the au- author of that song. Album coming to a streaming service near you soon, hopefully. Um, we've got a jam-packed agenda. We have got an amazing cast, amazing crew, and we can't wait to share it with you, so thank you everyone for joining us.

3:50 If you've got your Iron Viz banners, get banging them now.

3:55 Show, show you're all cheering and celebrating at home. So welcome to the TUG.

4:00 Today's lineup, well, like I said, we've not just got one Iron Viz contestant, we've got all three from this year's final. We've got Kevin, Brian, and Ann, and they're gonna show us all sorts of secrets about the vizes they created on stage in 20 minutes. We're being generous, we've given them 30 minutes today to talk, so they've got a little bit of time to spare.

4:16 Although, Brian apparently has got even more time to spare. He, uh, he nailed his Iron Viz in record time. So can't wait to hear from all three of them speak.

4:26 Your hosts today, you've got myself, Rob. I'm a, um, I'm gonna be just hosting the main thing. Then we've got Chris, Preethi, and Agata, who are gonna be hosting the, uh, the talks. They'll be there to field your questions from the chat. So if you've got any questions for, for the guys as they're going along, feel free to put them in and they'll, they'll reach out to the, the presenters and get them to answer your questions.

4:47 So a little bit about myself. I'm Rob, I'm a three-times Tableau Ambassador. Um, I love building crazy things on Tableau Public. It's my playground. It's like, I like to experiment with different charts, learn lots of fun things, and just generally have a good time sharing and learning from other people in the community.

5:04 Next up, we've got Chris.

5:07 Chris Westlake: Hello, I'm Chris. I am a data viz designer at JLL. I've been using Tableau for about eight years. Also love doing funky stuff, mostly with circles and radials in Tableau Public. Um, I won Iron Viz back in 2024, so super excited to see how these guys brought, what they brought to the stage this year.

5:33 Preethi Lodha: Hello, everyone. Preethi Lodha here. I am a director of data analytics in higher education in Boston.

5:40 Uh, my connection with Iron Viz is I participated a couple of times, but for la- past two years, I have been the judge for the qualifiers that has been happening, and it is an amazing experience to get to see all the creativity flow through and the experiences. Uh, I am currently a five-time Tableau Ambassador. Really excited to come back this year as a Tableau Ambassador.

6:05 And I, while I've not won Iron Viz, I did win a couple of data visualization competitions, so I'm really excited about that.

6:13 The visualization on to the left-hand side corner, how many, How Americans Spend Their Money, is my app- Award of having a million views on the Tableau Public. So that is a visualization that has over a million views. Uh, apparently if you visualize what people want to look at, it does get a boost up. I've been also published in a lot of featured articles and stuff for the data visualization work, so that brings me joy.

6:42 And I'm really excited, uh, to hear from our Iron Viz finalists this year.

6:52 Agata Ketterick: Hi, everyone. I'm Agata. Um, I started out building dashboards in, in, uh, Tableau, and most more recently I've turned that into a career in data strategy at USA Today.

7:04 Um, so these days I spend more time connecting, um, business needs, people, and data. Uh, but Tableau is still my favorite way to bring data to life.

7:14 Um, I've been part of the Tableau community since 2019. Uh, and I am a six-time Tableau Ambassador.

7:26 Eric Howard: Hey, everybody. I'm Eric Howard. I'm a five-time Tableau Ambassador, so I'm the founder of this TUG and one of the leaders of the, uh, Nashville Tableau Users Group.

7:36 Uh, I do a lot of, uh, presentations at, uh, TUGs, been, uh, three times at the Tableau Conference.

7:44 Um, if you want to present at Tableau Conference, get involved with your local TUG and present there. Uh, I tried for years and never got accepted, and it wasn't until I started doing things at my local TUG that they like to see things that have been tried before that work. So get involved with your local TUG, and we'll see you at Tableau Conference '27.

8:07 Rob Taylor: Thanks, Eric.

8:08 And this slide's quite nice. We also have Lisa Trescott as one of our hosts who has been at Dreamforce all week, so she's, she's not, not currently on camera. But, um, yeah, Dreamforce has been happening this week. Um, hopefully you've all managed to either attend if you're over there or check out some of the, um, recordings and watch them on streaming.

8:26 Um, the Tableau keynote was apparently really interesting. There's so much cool stuff they're doing. It's gonna change the way we're using Tableau completely, so really excited by that. So if, if anyone's watched it, feel free to put in the chat anything that excites you or anything you found interesting. We'd love to, uh, chat about it.

8:43 But now, without further ado, we are gonna move on to our first speaker, who is Kevin Wee. And, uh, Chris We- Chris Westlake's gonna be hosting him, so I'll pass over to you and Kevin, Chris.

8:53 Chris Westlake: Thanks, Rob. It is an absolute honor to introduce Kevin, um, who is a lead BI analyst at Capital One. He's a three-time Tableau Visionary and a Tableau Public Ambassador. With over 10 years of experience in data analytics and a PhD evaluating scientific visualization, Kevin specializes in visual storytelling, information design, and user experience.

9:18 If you've seen Kevin's work, that probably makes absolutely no surprise to you at all, as he was a 2026 Iron Viz finalist, and he's widely recognized across the DataFam for his UI design talks and free educational resources. Today he's gonna present his finalist dashboard, busting the myths about reading graphic novels. Kevin, great to see you. Take it away.

9:42 Kevin Wee: Great.

9:43 Well, um, hello, everyone. My name's Kevin, and today I'll be covering my, um, Iron Viz, um, the visualizations for the final round, uh, where I'm on stage competing with Ann and Brian. Let's get started. Um, let's see here. Yep. As Chris said, my name's Kevin Wee. I am currently a Visionary and an Ambassador. Um, for those that speak Mandarin, um, my Chinese name is Huang Koen.

10:09 And, um, I've been learning Tableau since 2019.

10:14 And, um, in the last couple of years, I have had opportunities to present on stage or in various TUGs, where all my videos, most of my videos were uploaded, um, in my YouTube channel.

10:27 I love giving back to the DataFam community, and in the future, if you have any opportunities for knowledge sharing, feel free to reach out to me via LinkedIn.

10:37 Um, as Chris said, um, this May, uh, the three of us had opportunities to compete on stage and, uh, donate monies to the charities. Um, and, uh, I created a visualization which I will share today.

10:51 So, um, for my Iron Viz in last November, I submitted one visualizations for the qualifier round, where I disclosed, uh, the nutrition values of various dim sum dishes.

11:07 And this May on stage, I created a visualization where I share my passion on reading graphic novels. And today we're gonna focus on the later visualizations. We're gonna talk about how it was built. We're gonna talk about some, uh, spark behind them. So let's get started.

11:32 Note that I'm not gonna give you the three-minute presentation that I did on stage, but I'm gonna walk you through the visualization. So what you're seeing here, it's the whole visualization actually started with, um, this version, where just a lot of redacted information's on a visualization.

11:50 I was trying to keep a s- a secret what topic I will be talking about, um, what kind of data on Goodreads I'll be extracting.

11:58 Then I disclose actually I, uh, will be ta- I would be talking about graphic novels.

12:06 Um, I started with presenting a myth about graphic novels where people thought it's mostly for kids or it's mostly superhero comics. But turn out, if you look at the data, only, um, 18% of them were, um, sci-fi.

12:23 And you can see that, um, I'm sorry, you can see that only 13% of the graphic novels on Goodreads were about superheroes. Only 7% were about children.

12:34 And the top three genres were actually fiction, fantasy, and sci-fi.

12:39 And if you, um, further, uh, look at the number of reviews, um, superheroes only, uh, had only 14% of the reviews were about superheroes. Only 16% of the reviews were about children. So maybe the s- slightly disproportionate high, um, comment count for students, uh, I'm sorry, for children's, um, graphic novels were why people kept thinking that graphic novels were for childrens, but it still didn't explain why people misperceive, um, superhero as a predominant medium.

13:16 So what I did is I created, I presented a, um, rank chart where I showed the popularity of different genre over time in the last, uh, I believe, um, six decades.

13:30 And you can see that, um, superhero had its peak and, uh, trough, and we can see that, uh, the medium actually peak in the, I believe in, where is this? This would be, uh, in late '80s, and then it dropped back, and then it bounced back again in, I believe, 2000. And I think, uh, this re- what do you call that? Renaissance of superhero genre, it's caused by, uh, the MCU, the DCEU, the Marvel, uh, Cinematic Universe, the, uh, DC, uh, cinematic, um, extended universe as well.

14:06 Maybe these, um, uh, cultural saturations coming from the TVs and movies were what, um, misdirect, uh, were what indirectly led to the misperceptions of the related graphic novels were the predominant medium.

14:25 Um, while I was creating this chart, I also presented that, um, turnout fiction, young adult, and fantasies had been, uh, the predominant, uh, genre in the last, um, two decades. But we did see that there's some disruptions happening in the '80s. Um, turn out in the '80s, there's one graphic novel named Maus. Um, it had, it fell under the classic, um, nonfiction war history genres.

14:53 Um, it's about, uh, the survival stories of the author's father as a, um, a Jewish in, uh, World War, uh, in the World War II era. It's actually the first graphic novel that won the Pulitzer Prize. Um, confirm, not confirming, uh, supporting the fact that graphic novels can actually carry serious topic, not just a children's medium.

15:17 I also share some of the, uh, re- more recent rising genres such as romance, LGBTQ, queer, and adults art. So, um, turns out graphic novels, it has so many different, you know,

15:31 topic that it could cover.

15:34 I ended my, uh, talk with recommending some of the books based on the genres that people tend to like, including, um, romance. Um, I recommended, uh, Heartstoppers. Uh, you can see that, uh, this franchise has been very well-liked and well-perceived by the readers.

15:51 And for the one that like history, um, graphic novels, uh, I recommended Maus, um, as discussed earlier. And for those that love, uh, fiction and fantasy, I recommended Sandman, which is a story about what it means if dreams could manifest as a physical entity.

16:09 Um, I ended my presentations with a quote directly from Sandman, talking about, um, the importance of reading something fiction because it represents humanities and human dreams.

16:23 And that's it. So that was my presentation. Uh, you should go check out the one-hour video from, uh, Tableau Iron Viz, um, just to see the details.

16:33 So after the demonstrations today, I will be covering a couple of things. Number one, um, one inspire my present, uh, my presentation, how I prepped the data set, and how I created a visualization.

16:46 And then we will be discussing the design and storytelling behind them, and I'll show you what people like the most, uh, in terms of, uh, the perceptions of the audience.

16:56 Um, and by the end of my presentation, you should be able to, number one, describe the makeup of my visualization. Number two, um, list at least one to two graphic design concepts that I have deployed in the visualization. And number three, you should be able to describe one to two data storytelling techniques that I leveraged in my presentation. So with that said, let's, we did the demonstrations already.

17:23 Now let's talk about inspiration.

17:26 Um, as, um, some of you know that I love things that are sci-fi and fictions. As a matter of fact, I have so many graphic novels in my house. I just love plowing through them.

17:38 And I'm a book nerd. I love, uh, novels and graphic novels. And what I like about them, it's they combine, um, storytellings and visual design, which as a data visualization practitioner, you can tell that I just love this.

17:54 Graphic novels, um, because you can create something that's so visually striking on the s- on the paper, um, it's often being used to depict something like superhero, something fictional, right? On the right side, you can see that this is a graphic novel called Klaus. Uh, it's basically telling what if, um, Santa Claus is a superhero. So you can see that how it used a lot of visual language to depict, um, this very classic entity.

18:21 Um, it also covers a wide range of topics, right? On the left side here, it's Sandman. Um, it tells a story, what if dream itself It's a God. What would happen in that universe? And on the right side here is a very realistic fictions depicting the love story about two teenagers in high school.

18:42 Um, because of its nature, um, graphic novel, graphic novels have been increasingly used to, uh, disclose a lot of, uh, profound subjects, including Maus, as we discussed earlier, talking about author's, uh, father's experience with, uh, uh, Holocaust and a concentration camp. It has also been used, uh, in other medium to dis- depict what it's like to live or escape an oppressive, uh, government.

19:08 So, um, graphic novels, it can be very a relaxing reading medium. It could also be used to depict something very, um, historical and realistic.

19:19 That's why I love, um, uh… That's why I decided to dis- uh, to talk about graphic novels in my Iron Viz.

19:28 Um, now let's talk about the data preparation part. Um, for those who are not, uh, fam- a- aware of this, uh, the data set that we used this year, they are, they were about Goodreads book review data. Uh, the data set's ginormous. I think, uh, we have 6.5 million rows of, uh, book records, and each row represents, uh, each row we have so many informations here.

19:52 We have the, um, the book title, the different kind of IDs, the publication date, the first publication date, summary, and then, uh, the top 10 genres reported on Goodreads. We also have informations about the rating and then how many people rated, uh, one star, two star, three, four, five stars. We also have how many percentage of the raters rated one to five stars, and how many reviews that we have.

20:18 Uh, we also have several informations about the authors. So when you are given this large chunk of data set, you really need to decide what kind of story you need to tell and start filter your data set based on that. Otherwise, you will be, uh, you would have been boiling the ocean.

20:36 And, um, in terms of personal data set, I actually created two more supplemental data sets I used to blend in with the Goodreads data to create a final visualization, including one, uh, table just about, um, all the supplemental text occur on the fi- in the final viz. I also created one book chart, um, data set, uh, for data densification. I'll explain more about it later if I have the time.

21:07 I use Tableau Prep mainly to filter, clean, correct, pivot, aggregate, joined, um, the data set or created, or created any, um, calculations re- regarding the visualization themself instead of the data analysis. Um, I can show you my Tableau Prep here. You can see how crazy it is. I think, uh, I speak for the three of us that we really max out the functions of Tableau Prep.

21:33 I can just zoom, like this, I can't even fit everything here.

21:36 Let me just zoom in a little bit, uh, for readability. You can see how, uh, the starting point here, I started with a 6.5 million rows of data set.

21:46 I,

21:48 I cleaned the data set. I update some of the value that were wrong. I deduped it, and then I export, uh, versions of the records, and I further branched out and then did more cleaning, more pivot aggregations.

22:01 Or I even blend it with my book chart data set, you know, create so many calculations and a lot of pivoting. Yeah. So I think Tableau Prep is definitely the unsung hero, uh, in, for Iron Viz, and that if you were to, um, participate in Iron Viz this year, uh, I urge you to master Tableau Prep.

22:23 Now let's talk about, um,

22:27 um, how I created a visualization, and then talking about the design and storytelling aspect of it. Um, here it's the final visualization, right? Let's go back to page six here.

22:42 Um, before I show you how I did it, I want y'all to give it a guess.

22:46 How many worksheet I used to create this visualization?

22:52 Give me a number. At least six.

22:55 So, "One plus map layers." That sounds like a Brian Moore thing. I, I don't think I can do everything in two.

23:02 Seven.

23:04 One?

23:06 12. Okay.

23:09 Eight. Okay. So, yeah, y'all got pretty close to what I had.

23:14 So I use four, um…

23:18 There's only four, uh, workbooks here. So one is for all the supplementary text at the background, including the boxes and the speech bubble.

23:30 Uh, one, it's the book chart. One, it's the rank chart. Another one is the scatter plot. So only four, uh, charts. Know that because I, we had to create everything under 20 minutes. I, I don't want to create like 10,000, you know, worksheets and then struggle to put them together on the screen. That, like, I feel like my hand's gonna jitter if I had to, if I had that workload.

23:50 Um, let me show you where these, um, four charts are. So basically, you can see that the giant worksheet's at the back, right? That is the only tile item. Everything else floating.

24:03 And, um, let me go back to one again. Yeah. So we started with this. With the book chart floating on top of the text worksheet, and then as I progressed, the book chart would disappear, and then we sh- now we show the rank chart.

24:22 And then as it progressed, the rank chart would also disappear. Now I show the, uh, scatter plot and the related items about scatter plot. Know that all of these swappings were done using dynamic zone visibility.

24:35 Which has been effective for our presentations when you have to show and hide different elements, different containers on a screen.

24:42 And you will be like, "Kevin, you only have 20 minutes, uh, and it seems like you have to align everything very precisely on a screen. How did you do it?" So, right, you can see that things has to align very clearly, otherwise the whole visualization will look very chaotic. So I

25:03 actually

25:07 created a cheat sheet here.

25:09 Can you see that?

25:10 Like, when a parameter is zero, I had these lines and the reminders on the screen telling me what to do so that I remember to put all the worksheets, like aligning the line.

25:23 This is actually how I, um, make sure every worksheets were precisely landed. Let me show you this.

25:30 This is how chaotic it is. If I turn off all the… If I turn off all of the, what do you call that, dynamic zone visibility, this is where everything was. This is where everything was. You can see that the, um, what is- what's this? Uh, let me turn this off.

25:46 Um, the book chart align with the lines for the book chart, and then the rank chart align with the line of rank chart. Um, yeah. That's why the last five minutes, uh, when I was on a sc- when I was on the stage, I was, my hand was shaking and I was like, "I need to align this, like, to the pixel that I des- I designated them." And then I have to remember to go through this list, so that I remember to study all actions and everything.

26:13 And, yes. And my final stage, uh, step on the stage was to, you know, move everything to one, so now everything look more, uh, closely clean. So, yes, this is how I made them.

26:24 And, um, now let's talk about, uh, design elements. So we've been going through this for multiple times.

26:33 Um, when I designed this visualization, there are several things I deployed to, uh, make it immersive to the audience when I talk about graphic novels. Um, number one it's I use a lot of, uh, design items that related to graphic novels or just about books. Um, people love the book charts, right? So this is clearly, I'm trying to play with the metaphor that, uh, as I talk, we flip through books.

27:01 We flip through books. Um, sadly, these animations were not activated on Tableau Public, uh, because some loading issue. But, uh, on the stage you can actually see the animations where the book were flipping.

27:15 Um, I also use things like, I believe, um, speech bubbles. Right? This is a speech bubble. This is also speech bubbles. Just to, um, play with the idea that this is more like a comic book, less about, uh, traditional literature.

27:32 Um, other than that, I believe when I was presenting you can also feel that I use a lot of, um, design elements to tell people where to, um, the s- importance of the informations, right? The more important texts are bigger or in bold.

27:49 And then the more, uh, important part of the visualization, they are often thickened, right? Like now if I want you to focus on the book chart, the book chart itself is thicker than every other lines on the screen.

28:02 Um, here. When I want you to focus on, um, the rank chart, the rank chart is thicker than, uh, the tick marks. Also, when I want you to s- focus on specific genres, I use color to guide your attention.

28:17 So yeah, these are some of the design elements I use. And also you can see that I align everything to the T. So just turn this on. Uh, by the way, if you press G on your keyboard when you're doing dashboard design, you can summon the grid line. You can see how… Oh, you can't see it that way, but here you can see that how I align everything. Like, all the text are align.

28:42 Right, I believe if I, let's say go to the rank chart. You can see that, um, titles here, they're all align. Yeah. So as a control freak, as a control freak, yeah, this is how I make sure all the informations are well lay out.

28:57 Now let's talk about the storytelling part.

29:00 What kind of, um, techniques I've been used, uh, I've been using in, uh, my presentation. Number one, it's I control people's mental load. I understand that that was the second day of Tableau Conference at 4:00 PM. Everybody was tired. No one had the mental load to learn something that's too complicated. So when I was presenting, I started with something very little, almost empty, like almost noth- no- nothing of importance on the screen.

29:29 And then I gradually re- uh, reveal more information, more information, and more information. Kind of walk people through, especially when I'm talking about a novel chart type.

29:40 Um, I also leverage animations to draw people attentions, as I showed, and I also, I leverage a lot of color highlight in my, uh, presentation.

29:51 Color is an effective way to direct people's attention because they tell people that, oh, the color part is the information. The gray out part is the secondary information. You don't have to pay attention for it now unless you have to.

30:05 So these are the different techniques I use to, uh, tell an effective story. I hope that it's effective.

30:12 So now let's go back to my presentation.

30:17 Yeah. The wow factor.

30:20 Um, pretty, I'm glad that you asked, uh, where did the idea for the book chart came from. So, yes, um,

30:28 by the end of the Iron Viz, uh, someone reached out to the three of us and then show, uh, us, uh, this photo, uh, this illustration that they drew while we were competing on the stage. Um, you can see that people really love the book charts. Um, they find it very, um- Uh, fitting to the topic.

30:46 And, um, and as a matter of fact, of course, uh, it is one of the more complicated charts, uh, I have, um, created on the stage. Um, it is parameter triggered. Uh, it has animations. It's, uh, curvy, right? Non-conventional. So people love it. Um, so what inspire it? So here's the fun fact. When I was pr- when I was prepping for my Iron Viz, I did not come up with a, a book chart at first.

31:12 It was only the rank chart, KPIs, and then the scatter plot. But then, uh, my sous viz, Gus Frazer, told me that, "Kevin, uh, people love your dim sum visualization. People love when you create something round, like for individual topics. Do you think you can recreate some kind of magic like that?" So what I did was I spend the next weekend, I just keep sketching.

31:33 I was like, "Okay, it's about book. So now let's do book." Right? So I sketch different versions of books, um, three-dimensional one, two-dimensional ones, just to create something book like. And then after that, I moved to Figma just to keep on creating. Okay, if we are, if, if we're gonna do a book chart, how do I draw that book out? You can see that a lot of iterations, they, they look like a speedometer.

31:55 They don't look bookish at all. Turn out that part was really hard.

31:59 I eventually went online, find some, um, cross sections of a book, and then I just keep on tracing it until I, you know, come up with something that look like a book.

32:08 After that, let me show you.

32:11 On, um, Excel, basically I come up with this model, um, where I basically, I pinpoint, like create a, what do you call that, the coordinates for each point of the book spine, of the, you know, the curve part, and then some dots just to create this book chart.

32:30 I created this data model. I think this is just one book. For one book I have, what?

32:36 Uh, 440 data points.

32:39 And then on Tableau Public, I'm sorry, Tableau Prep.

32:44 Let me just, oh, move things around here. Go, go, go, go.

32:51 I blend the book model data set with the existing records about

32:58 the top genres.

33:00 Oh, you can't see now.

33:02 But the point is, uh, I blend them together and then I created, um, a new data set where, um, it represents how different elements moved on Tableau Desktop when I click through the parameters.

33:18 So yes. So that's how I created a book chart.

33:22 Now, let's wrap things up together. If we have more time, so we can open for questions. Um, earlier today, uh, I talk about these are, um, some of the objectives of learning I hope you get out of my presentation. Um, number one, let's talk about the makeup of my visualization. So I only leverage four worksheets. One background worksheet that contains the text, the boxes, and the cheat sheet, and then we have three floating charts.

33:52 Um, the whole visualizations was powered by the dynamic zone visibilities where I show and hide different containers as I click through the parameter.

34:02 I also use parameter scrolling to tell the story.

34:05 The data processing was mainly done using Tableau Prep.

34:10 And in terms of graphic design concepts, I deployed visual metaphors such as flipping books and speech bubble to, uh, make it immersis, um, immersive for my audience.

34:22 I also, uh, leverage a lot of, um, graphic design concepts such as typography, layout, and color, uh, to create, uh, hierarchies of information to direct the audience attention.

34:33 In terms of storytelling techniques, I leverage a gradual revelations to control the audience mental load.

34:40 I also leverage animation and color highlights to guide the audience attentions to a specific part of the visualization.

34:48 So with that said, um, these are my, uh, con- uh, contact info. So for those who like to reach out to me, uh, for more graphic design or UX/UI tips or, uh, just to connect, uh, feel free to reach out to me via these platforms.

35:06 And, uh, on my Tableau, on my blog, uh, there are a lot of resources about data visualization, graphic design, book recommendations for, um, data visualization topics, and the slide deck of every presentation that I've done, and a lot of video recordings as well.

35:21 With that said, I'd like to thank all of the Tableau Public, um, user group leaders, uh, for the talk invitation. I'd like to thank, uh, Ashley Dirks and, uh, Gus Frazer specifically for the design of the book charts. I'd like to thank Ann Brian and many Tableau staffers for the awesome Iron Viz experience. And there are just so many more data fan members that I'd like to thank, uh, that support me through the whole pro- six months, uh, while I participated in the Iron Viz from the qualifier to the final rounds.

35:52 Thank you.

35:55 27 minutes. Cool.

35:58 Chris Westlake: Awesome stuff. Thanks, Kevin. Um, I always love hearing about people's kind of Iron Viz journey and sort of-

36:05 Kevin Wee: Mm-hmm

36:05 Chris Westlake: … a lot of people don't really realize the amount of work that goes on in the background, and it's, it's really cool how passionate you still come across about it, I think.

36:17 Kevin Wee: Thank you. Thank you.

36:18 Chris Westlake: I was slightly sick of talking about it. It's like that PTSD of going back over it all.

36:26 Um, did I-

36:27 Kevin Wee: Earlier, uh, when I was writing some, uh, tips down on my, some cheat sheet on my laptop so that I remember some of the talking points, my hand was also shaking. I was like, "This gonna, we're gonna have to do this one more time." Did I miss any questions, um, in the chat?

36:43 Chris Westlake: There were a couple, yeah. Um, one from Rob asking if you'd used Tableau Prep at all before the final

36:51 Kevin Wee: Oh, yes. Uh, so, uh, I learned about Tableau Prep in early 2020 when I started doing Makeover Monday.

37:01 And yeah, so I have a couple of years of using Tableau Prep, but Tableau Prep in 2020 was not the versions that I had last year. I think, uh, things like aggregations, things like parameter setting, these are new functions on Tableau Prep that I had to leverage when I, uh, participated in the Iron Viz.

37:20 Chris Westlake: It's definitely come a long way in the last, last few years.

37:24 Kevin Wee: Mm-hmm.

37:24 Chris Westlake: Great stuff. Um, Ann is also asking, um, she says, "You were so incredibly focused during the build, as if you didn't react to the crowd.

37:33 Was that a result of just being hyper-focused, or did you have earplugs in, or what's the secret there?"

37:40 Kevin Wee: I didn't have any earplugs in. I think the three of us didn't have any earplugs in this year. We just practiced so many times, uh, before, uh, on the stage, I think we just at that moment were like, "We're just gonna keep on doing this and ignore everybody." But yeah, it was very loud, uh, on the stage, uh, to receive the energy and the, you know, the noise, uh, from the audience.

38:02 Chris Westlake: It's just the best feeling, isn't it?

38:05 Um, and then finally, another question from Rob. Um, how many rebuilds did it take you to get it done in 20 minutes?

38:14 Kevin Wee: What do you mean by how many rebuilds?

38:15 Chris Westlake: A lot. So to get your final build down to the 20-minute-

38:20 Kevin Wee: Uh-huh

38:20 Chris Westlake: … time li- time limit, how many times did you have to go through the build?

38:26 Kevin Wee: That's a very interesting question. So I, I think my first time I did it, it was 23 minute. No, no. I think it's 23, 25 minutes, but I have to say that, um, I think part of the perks of being a Visionary while before you step on the stage is that you get to talk to so many other Visionaries and final- finalists beforehand. So I've learned so many tricks before I even participated in the final round.

38:52 So let me just show you some of the things that I did to streamline the process. Number one, you see how I enumerate everything, right? So all the pills, I enumerated it, and I tell myself, I told myself exactly where to place them. So while I was doing this, it's like I didn't have to do any of the cognitive processing. I was on my zombie mode. I was just drag, drag, drag, drag, drag.

39:20 Click, click, click, click, click. Um, it's, um, freeing in that way. So that's why when I was practicing, uh, my time was actually not that long. It was slightly longer, but and then the rest is just like, um, I spent more time on Tableau Prep fixing the numbers and fixing, um, the, the title of the each pills more than, uh, practicing the final round, I'll say that.

39:44 Um, uh, but I think, uh, the, the work paid off.

39:48 Chris Westlake: That is impressive. It definitely sounds like if anyone on the call gets into the final next year, you definitely want to chat to Kevin about tips how to do, do stuff fast, 'cause 25 minutes for a first run-through is-

40:02 Kevin Wee: Mm-hmm

40:02 Chris Westlake: … very impressive.

40:04 Kevin Wee: I, I think I got the… Just to give the credit out, I think I, I received, uh, this tip from Ryan Soares.

40:12 Chris Westlake: Nice. Excellent.

40:15 Cool. Thank you, Kevin. I think that's all the questions we have, so I will hand over to Preethi, who's going to introduce Brian.

40:25 Kevin Wee: Sure. Thank you.

40:25 Preethi Lodha: Thank you, Chris.

40:27 That was really exciting, Kevin.

40:31 Okay. Uh, I'm really excited for our next, uh, speaker, uh, Brian Moore. Uh, Brian and I have a connection, not only that we are from Boston and we co-lead the Boston Tableau User Group, uh, I consider Brian a friend too. So Brian is a data visualization and analytics practice manager at ClearTelligence, as well as a five-time Tableau Visionary, six times Ambassador, has 15, one five, viz of the Days under his belt.

41:08 Brian, do you just keep churning viz of the Days? Is that what the aim is? Like you publish and it's a Viz of the Day.

41:17 And co-leads, uh, the Boston Tableau User Group.

41:20 Uh, Brian is known for his, uh, unique style of building single-sheet Tableau visualizations and sharing his techniques on the blog DoMoreWithData.com.

41:34 Brian is also an active mentor and community leader.

41:38 And as a 2026 Iron Viz finalist, Brian will be sharing the story behind his entry, Exploring Literary Prestige Versus Public Opinions, and it's titled Hogwarts Over Hemingway.

41:53 Brian, take it away.

41:56 Brian Moore: Thank you very much, Preethi. Thank you everybody for having me.

42:00 Uh, I apologize in advance if there's a lot of background noise. Um, I'm having some of the flooring replaced in my house, and they're currently working right above me, so my whole house is shaking. I don't know if you can hear any of it, but I will do my best to, to work through that. But, um, yeah, so as Preethi mentioned, um, my Iron Viz final was titled Hogwarts Over Hemingway.

42:24 Uh, I'm gonna focus a lot more on the kind of technical aspects of the build and a lot less on the story, um, just 'cause I think the, the technical piece is a lot more interesting.

42:34 Um, but the, the viz focuses on some of the really, you know, popular books, um, that we commonly re- refer to as the classics that have, um, a high number of reviews on Goodreads, but a surprisingly low average rating.

42:50 Uh, and it looks at, you know, whether the popular- uh, popularity of modern fiction might be, uh, like partly to bl-

42:59 Um, so when I started thinking about what I was gonna build on stage, uh, I decided that whatever it was, I wanted users to be able to view it on Tableau Public and have an experience kind of similar to if it was being presented live on stage.

43:15 So I wanted to make it really easy to navigate through the story, um, so that there was no question on where to go to next.

43:23 Um, and also wanted to include kind of like all of the commentary, so it would seem like there was still a narrator, you know, even if I wasn't there presenting it.

43:33 Uh, so set it up so users could just kind of keep clicking next, uh, to navigate through the story, and then they could read some insights about, you know, each one of the steps or each one of the slides there on the left-hand side.

43:46 Another thing that I decided very early on was that I wanted to build the entire thing in a single worksheet.

43:54 Uh, so after watching some of the builds from previous years, I noticed that, like, the majority of time was spent on formatting. Um, so, you know, things like setting up multiple dashboards, adding text boxes, formatting charts, uh, things like that. So figured if I could build the entire thing in just a single sheet, it would save a ton of time during the actual build.

44:16 Uh, what I didn't realize is just how much time it would actually save.

44:20 So the actual build, um, I'm not sure how many of you watched it live, but the actual build for what I presented on stage only took about eight or nine minutes.

44:30 Um, and then I spent the first 10 minutes or so building this, 'cause I had a little bit of time to spare. This is just like a little… Call it the ratings explorer that people on Tableau Public could use to explore the data.

44:42 Um, but I used about half of my time to build it and did not present it on stage. So just this main piece here was built in one sheet, um, in about eight, eight and a half minutes, I think.

44:54 All right. So some technical details about the viz. Uh, as I just mentioned, you know, it was built, um, basically in a single worksheet. There are a couple of exceptions.

45:04 Uh, one is that ratings explorer that I just showed you, um, that was added kind of ac- after the fact when I realized I had some extra time.

45:13 Uh, and the other one was the navigation button, so that previous and next button.

45:17 Uh, originally I had that built into the same sheet, but once I published it out to Tableau Public, the interactivity was a little bit glitchy. Uh, so I moved it to its own sheet just for, like, a better, better user experience.

45:32 Uh, next, everything you see is a mark. Uh, I mean everything. So if you look at it closer, so chart titles, uh, axis lines, axis labels, a main title, reference bands, reference lines. Every single thing is a mark here, even like these little decorative lines over here. Everything you see on the screen is a mark. There aren't any images or anything like that.

45:58 Um,

46:00 but to, uh, to be able to build this in one sheet, like, I really kind of had to design it this way.

46:05 Um, I couldn't rely on any of, like, the native functionality for things like chart titles and, and reference lines.

46:12 Um, so really just everything had to be kind of plotted manually.

46:20 Uh, as you may have noticed on that last sheet, I did use map layers to design that sheet. Um, kind of allowed me to, you know, plot everything on the same grid and kind of easily control, um, like the stacking order, so what appears on top of what else.

46:36 Um, and it uses just three different mark types. So there's bars, lines, and text.

46:45 Uh, so going back to that sheet, you can see that there are, um, a total of six layers.

46:53 I'll just go through those kind of quickly. So starting at layer zero.

46:57 Uh, so the only place where this is used in the entire sheet is just for this little reference band here.

47:05 Um, this is actually a bar chart.

47:09 So you can see layer zero is a bar. Um, it's actually a, a stacked bar, um, because I wanted this reference line to start at the axis, or so I wanted the reference band here to start at, like, the axis line, which you can see is about 0.1. But for a regular bar chart, you can't really control where it starts. It's gonna tr- uh, start at zero. So there's actually a hidden section of the bar down here, um, to make that reference band.

47:38 So this is a stacked bar with just the value of, like, 0.1 here.

47:44 So you can see if I was to hide that layer, the only thing that disappears is that little colored band there.

47:54 All right, uh, the next layer, um, this is… uses the line mark type, and it's used for every single one of the charts. Uh, so this scatter plot, um, even though it looks like a circle unit, um, these are actually lines. It's just if you have a line that doesn't go anywhere, if it has the same starting and end point, it'll show up as a circle.

48:13 Um, so this one view here controls all of the charts. So it controls that, it controls…

48:21 Let's see.

48:24 Um, these unit charts.

48:31 The horizontal and vertical bar charts are all controlled just with this one layer here.

48:43 Uh, layer two is another bar chart. Uh, this serves two functions. So the first is this colored section on the left here that kind of changes color as you go through.

48:51 Uh, so this is a bar chart.

48:53 And then there's also another hidden bar here, um, that just hides the beginning of these, like, rounded bar charts so they have a uniform, like, accurate starting point. You can see if I hide that layer, you can see the- Tips of them drop down below zero, this disappears as well.

49:15 Uh, the next layer is another line chart, and this one controls all of the other lines. So, um, axis lines, reference lines, and these kind of decorative lines over here in this section here are all controlled on this layer.

49:34 Uh, the next one controls almost all of the text.

49:37 So any text that you see here that's, um, oriented horizontally is on this layer.

49:45 So that's the, the chart titles, the main title, all this text over here. The only text that is not, um, part of this layer is the vertical text, which is part of the fifth and the final layer. This one is just for those few kind of, uh, instances where the text needed to be aligned the other way.

50:07 Uh, so those six layers account for every single design element that you see as you navigate through this.

50:13 Uh, but you're only ever seeing a subset of those at any given time.

50:17 Um, so if I was to remove this step filter, see what happens.

50:22 Uh, this might look familiar if you watched Iron Viz Live. Here, you're seeing every single design element for every single step all at once, but I'm using a, a filter and a parameter to kind of control what you're seeing, which I'll talk a little bit more in just a minute.

50:42 Uh, so I'm gonna talk more about the data source in a few minutes, but just wanted to talk a little bit about how it's structured, just so that the, the filtering and navigation makes a little bit more sense.

50:54 So in my data source, I have a field called step.

50:58 Uh, each step represents a click as you're navigating through the story. So the story starts at step zero. When you click next, then you're viewing step one. If you click next again, then you're viewing step two, uh, and so on, for a total of 36 steps.

51:13 Uh, so each of these steps, uh, the data source has every record, um, needed to plot all of the different design elements for that step, and then nothing else. So it's basically like 36 little mini data sources kind of stacked on top of each other.

51:28 Uh, and then there's a parameter to track, uh, which step the user's currently on. So when you open the workbook, it'll start at zero. When you click next, it'll increase by one. When you click previous, it'll decrease by one.

51:40 Uh, and then it's as simple as just adding a filter to, to filter the data to only show the records where the step field equals that parameter.

51:48 So here, if I scroll down, you can actually see all of these views by step.

51:58 And then if I add that step filter,

52:05 now you can just see one when it's on one. When I go to two, you just see two.

52:10 When I go to three, you just see three, and so on through the entire presentation. So that's how that kind of navigation and controlling what the user sees, um, was set up.

52:22 Um, so the actual workbook is really pretty simple. It's one worksheet with six map layers, a few pretty basic calculations, and then one parameter action.

52:34 I tried to make it as simple as possible so I could build it really quickly, and so I didn't mess up anything while I was on stage.

52:41 Uh, but to get to that point, it took a kind of ridiculous, uh, amount of planning.

52:51 So before I could start really building in Tableau, I had to figure out what my story was gonna be.

52:56 Uh, as I'm sure you know, there's a 20-minute time limit for building your Iron Viz, but there's also a time limit for the presentation, which I actually found a lot more restrictive. Um, so trying to find a, a good story that you could tell in about three and a half minutes is, is pretty difficult.

53:13 And I knew because of the way that I was gonna build this, that once I started building it, it was gonna be really difficult to, to make changes. So it was really important to get the entire story nailed down right at the start.

53:26 Uh, so I made some really rough versions of the charts that I was gonna use. Um, I wrote out my entire script, and then spent the first couple of weeks just editing that until I had something that I was happy with, and that I could present within the time limit.

53:41 Uh, and then once I knew what I was gonna build, I had to figure out how I was actually gonna do it. But luckily, we had a, a couple of months with the data set, and we were allowed to use Tableau Prep, which I did very happily.

53:55 Uh, so here's a look at the final version of my prep flow, and it's, you know, probably where 95% of the work happened for my Iron Viz.

54:03 Uh, in order to build everything in one sheet and to support that navigation, um, I had to set up the data source in a very particular way.

54:11 And one of the most important s- steps in doing that was, um, to do some data densification, uh, just so I had the records that I needed to plot every one of those design elements for every story, or, or every step throughout the, uh, the story.

54:28 Uh, so here's a kind of simplified version of my densification table, but basically it has, um, you know, a record for every step, and then for every design element within that step, which, um, in here is called component.

54:42 Uh, but this includes things like the chart titles, footnotes, the text, the type of chart. If you scroll down some, some of the more complicated ones, it has axis labels and axes, the, the tick marks, and all kinds of things.

54:59 Um, and then there is also a, uh, a field called order. So for some of these design elements, I might need more than one point. So for example, for a line, you need two points, a start and an end.

55:12 Uh, for some of the more complicated ones, might need more points.

55:16 Uh, so there's a field called order, and then finally a field called join. Uh, and this just has a value of one for all of the, the chart elements, and a zero for all the design elements.

55:27 Uh, and this is just so I can take this and then join it to, like, the standard dataset wherever join equals one.

55:34 Um, so I can have the actual data to create those charts, and then kind of union the records for all of the, um, the, the design elements on top of it after that.

55:44 Uh, so that, uh, kind of like the base level gave me all of the records that I needed to be able to plot every single design element chart, every single mark in every single step of that story.

55:57 Uh, so at this point there was a lot of work to do, but, um, this at least provided the, the data that was needed to, to create all of those calculations and things.

56:10 Uh, so the overall goal of the prep flow was to create the minimal number of fields that I could, um, that would let me plot all of those different design elements and as much as possible to use common fields that could be used on any of the map layers regardless of the mark type.

56:28 Uh, so every single mark needed an X and Y coordinate. So whether it was a chart, a label, a line, everything needed the X and Y coordinate. That's what actually allows you to plot it, plot the position of it on the worksheet.

56:41 Um, so those are used in the Make Point calculations in map layers to, to get the position.

56:49 Uh, every mark also needed a value for color, and by using the same field for every record, uh, I only had to apply color once and it would update the bars, lines, text, everything. So I had one field called color that I could use across the board and just, um, apply my color palette once.

57:08 Uh, the same is true for size. So there was just one universal value or field for size, and so I only had to adjust the size once and it would apply to every mark.

57:18 Uh, each mark also needed a value for layer. So that's what I used in the Make Point calculations to determine which marks belonged on which of the six map layers.

57:28 And then finally, there's the step field, which i- was used for filtering and comes from the intensification table.

57:37 Uh, then there were a couple of additional fields needed for specific mark types.

57:41 Uh, so for all the bars and lines, I needed a unique identifier, um, that was used for the animations and a detail to separate the marks in some cases.

57:52 And, um, and then a field called text for all the text marks. And these were, um, calculations used to actually create all of the text that you saw. Even the, like, the commentary, all of that was done, um, with pretty, pretty manual calculations.

58:11 Um, so I don't have time to dive too deep into the prep flow, but this is where all of those calculations are created.

58:17 Uh, so the blue section up here is for all of the charts in the workbook, so there's a different step for each chart.

58:23 And in each one of these, I'm creating an X field, a Y field, a color field, a size field, um, and then they all get unioned together.

58:31 Uh, same with all of these sections. So in this next section over here, in this pink one, this is just for the mark labels.

58:37 Um, so those aren't actual labels, they… Those are text marks as well. Um, uh, but the same thing here, calculating the position, the color, the size, and all of those common fields.

58:49 Uh, this pink section in the middle is for all of the other texts, so the, the commentary, the titles, labels, um, like access labels and map labels and things like that.

59:00 Uh, this little yellow section here, just that single step, is for all of the other l- like, uh, design lines, so the access lines, reference lines, and those decorative lines.

59:12 This pur- purple step over here is just for that reference band, that layer zero that I showed that's only used on that scatter plot.

59:20 And then this section down here is for that other bar that changes color throughout the, uh, the entire viz on the left-hand side here.

59:28 Uh, so all of those steps are used. Um, the output of each of those has those common fields, and then they're all unioned together to, to create that final data source.

59:42 So if you go through and look at any of these layers, you can see it's just color, size, Make Point, ID, color, size, Make Point, ID, and order.

59:52 Color, size, Make Point, ID. It's all the same exact fields on all of these, and I don't have to adjust them individually because it's the same field. I put color on all of them, and then apply the color. Put size on all of them, then adjust the size.

1:00:06 Um, so that really kinda cut down on the build time.

1:00:11 Um, also it made it really easy to remember how to build it when I was on stage.

1:00:19 All right, um, and if I haven't lost you, um, on the event yet, uh, there's one last thing that I wanted to touch on, which is the animations. Um, they were, I would say, a really big part of my presentation, so wanted to kind of briefly describe how I set those up.

1:00:35 Uh, so for any design element that I wanted to animate, so, uh, if you remember, let's just go back here for a second.

1:00:42 Uh, as you go through this, there's stuff, not in these specific steps, so we can get past that. There's stuff kind of moving on and off the screen, moving around the screen, and things like that.

1:00:53 Um, so for anything that I wanted to animate, anything that I wanted to move, I created two additional calculations in my prep flow.

1:01:02 Um, one was called move from, and that's the direction that the element will move from on any given step. So if it's moving from the right side, then we just have the value of right. If it was moving from the top, it would have the value of top.

1:01:17 Uh, and then move to is the direction that that element will move on the next click. So when you click again, is it gonna go to the right? Gonna go up? And so on.

1:01:26 Uh, then for any record that had those values, I would create a duplicate of the record in my prep flow.

1:01:33 Um, and if it was, uh, for the move from record, I would decrease the step by one, so that way that element actually existed on the step before it needs to appear on the dashboard. And then I would adjust the X and Y coordinates depending on the direction I want it to move from. So if I wanted it to come from the right, I would increase the X coordinate.

1:01:51 If I wanted it to come from the top, I'd increase the Y coordinate.

1:01:55 Uh, and then because I fixed the axes on my worksheet, it would essentially just hide these marks out of the view, uh, until it got to the right step, and then it would move them kind of into the viewable section of the dashboard.

1:02:06 Uh, which I'll… I'll demo- demonstrate that in just a minute.

1:02:11 Um, but just a, a quick note on the animations. I had to pay really careful attention to what fields were on the marks card and how those values change from step to step.

1:02:21 Basically, if you want to animate marks, you can really only change one thing at a time from step to step.

1:02:27 So I could change the position or I could change the color, but you can't really do both. Um, size is another one. So you have all these different attributes. You can change one and it'll animate to change that, but if you try to change more than one, um, then the animations kind of break. So that was a bit tricky, just making sure the data source was set up correctly to, to support these animations.

1:02:50 Uh, but this is kind of a demonstration. So, um, this red section here is the, like what you actually see when I'm presenting. This is a view of all sections of the dashboard. This is on step zero, but you can see the marks for step one already exist. They're just hidden off to the side. When the user clicks Next, this goes to step one, and then that moves into the view.

1:03:15 And the next one, it just changes color, but it stays there.

1:03:19 A few other changes. There you can see something else getting ready for the next step.

1:03:24 So that's gonna move out. That moves in.

1:03:29 And then as you can click, you can just see…

1:03:33 It's not… It's actually was pretty simple to, to set up, um, once I figured out how to do it, but it's really just moving stuff slightly to the right or slightly f- above or below, um, and then controlling what the user can see.

1:03:54 All right. Um, so that's kind of a high level overview of how I designed my build. But, um, the, the whole experience was just really, really incredible. The, uh, the team at Tableau was really supportive and, you know, really did everything they could to help us succeed.

1:04:09 And probably the, the best part of the whole thing was getting to hang out in a, a small windowless room with, uh, Ann and Kevin all week and getting to know them a lot better.

1:04:17 Um, technically it's a competition, but, uh, you know, it felt a lot more collaborative than anything. Um, so it was really, really a cool experience.

1:04:27 Um, that's about all I have. I tried to keep the presentation on the shorter side so I could leave time for questions. Uh, so if you have a question, happy to answer those now, or if you'd rather ask it in private, you know, feel free to reach out to me on the Tableau site.

1:04:41 Preethi Lodha: We do have a couple of questions for you, Brian. Uh, so, um, I think the first question comes from Neha, and she put it out to all the finalists. Uh, she says, "I know it's hard to think about it, as you all had your soul poured in these, but what is that one thing, an afterthought, that you felt you could have changed?"

1:05:07 Brian Moore: Um, you know, honestly, I don't think I have any regrets. I was really, really happy with what I built, and I was really happy with how the presentation went. I don't think I would change anything about it.

1:05:18 Preethi Lodha: Yeah.

1:05:19 Yeah.

1:05:20 Uh, the next question is from Oycina. Uh, "How many weeks did it take you to put together the guides, directions, and references for the viz?"

1:05:35 Brian Moore: Um,

1:05:37 sorry. The

1:05:41 guides, directions, and references? I'm not sure what they're referring to.

1:05:47 Preethi Lodha: I think they're talking about doing the entire viz together and figuring your data strategy out, your, uh, prep flow out, and designing the actual dashboard.

1:06:00 Brian Moore: Um, so I th- I think… I don't remember exactly. I think we had about two months to do everything, and I spent probably the f- first month on the story and figuring out exactly what I wanted to build, and then the second month doing the actual build. Um, I, I spent a lot of time in that second month.

1:06:21 Preethi Lodha: Yeah. Yeah.

1:06:22 Uh, Kevin has a question for you. Kevin wants to know: Do you plan everything on paper whiteboard first before creating the spreadsheet?

1:06:33 Brian Moore: Um, so I had just done it, like…

1:06:37 I just did it in, like Word just so I could incorporate, like, my weird, uh, kind of junky chart examples in with the text. So I just had like a, like a Google Docs document going, um, where I could plan out the, the text and what I was gonna display.

1:06:57 Preethi Lodha: Yeah.

1:06:58 That's interesting.

1:07:00 Uh, Lisa has a question for you, and I think it is more about how did you decide on this particular data story?

1:07:08 Brian Moore: Um, so I think I kinda lucked out. My story came together pretty quickly. Um, when I started looking at the data, I filtered it to just, like the, um, just the titles with the, the most reviews. So I set the threshold at a million, and saw that there were only about 200 books that had more than that many reviews. And just going through the titles, I could see that, you know- A big chunk of them, probably a third of them, were, you know, the classics, and then probably another, you know, 40 to 50% of them were, like, modern fantasy titles.

1:07:44 Preethi Lodha: Yeah.

1:07:44 Brian Moore: So just those two groups made up, like, the vast majority.

1:07:48 Um, and then, so I wanted to do something with those two kind of groupings, and then as I started going through, I realized that one group had really low ratings, the other group had really high ratings.

1:08:00 Preethi Lodha: Yeah.

1:08:01 So before I let you go, Brian, one final question, and this is from me.

1:08:06 So I've seen you optimize to the core, right? Like, uh, I did believe in your fake-out. I was scared when I was in the audience of why this dashboard has not come through, so that was really interesting. But what is one Tableau habit you have developed that saves you most time, not just for Iron Viz, but in your, all the builds of Tableau dashboards?

1:08:31 Brian Moore: Um,

1:08:35 let me think about that one for a minute. I'll drop it in the chat.

1:08:38 Preethi Lodha: Okay.

1:08:40 Brian Moore: When I…

1:08:40 Preethi Lodha: Thank you, Brian.

1:08:43 Brian Moore: All right. Well, thanks everybody.

1:08:46 Eric Howard: Woo-hoo. Everybody enjoying the thing so far?

1:08:51 Spin the wheel to see who wins the swag.

1:08:54 There's the email address, Tableau.public.TUG@Gmail.com.

1:08:59 So the winner can shoot me a quick email, and we will set you up with a $50 gift certificate.

1:09:08 Sam, Cherie. Looks like Cherie is gonna be the winner.

1:09:14 Congratulations, Cherie.

1:09:16 Again, tableaublicktug@Gmail.com, and we will set you up with some swag.

1:09:23 All right. Back to you guys.

1:09:27 Agata Ketterick: All right.

1:09:28 Eric Howard: On to Sharon.

1:09:28 Agata Ketterick: Um, so I am super excited to introduce Ann Pregler, uh, the 2026 Iron Viz Champion. Uh, so Ann is a senior data analyst at BECU, and a Tableau ambassador.

1:09:44 Uh, she is a former math teacher with a masters in data analytics, and she uses her background in education to teach data storytelling through her blog at annepregler.com.

1:09:54 Uh, she is also a Tableau certified consultant, a Vizzy Award winner, and as we saw in her winning viz, an amazing artist.

1:10:05 All right, Ann, I'm so excited for you to show us, show us the magic behind this viz.

1:10:10 Ann Pregler: Thank you.

1:10:11 All right. Uh, let's jump in here.

1:10:16 One sec.

1:10:17 Uh, okay.

1:10:20 So, uh, as Agata said, hi, I'm Ann. I used to be a math teacher, now I do data viz. Um, and this year I did Iron Viz, as you know. Uh, this is a little bit of my work, the kind of things that I create. I typically like to do very story-driven, um, visual pieces. Uh, that's where kinda my, my passion outside of work is in data viz.

1:10:42 How did I get to Iron Viz this year? Uh, my qualifier entry was about packing for Mars, um, how to grow vegetables on a spaceship. A little bit niche, but a very, very fun topic, uh, where I explored the different fruits and vegetables that astronauts could grow in space to help get them all the way to Mars.

1:11:02 But, uh, we're really here today to talk about the Iron Viz finals, of course. Um, and so let's hop over and take a quick look at what I built, just in case anybody hasn't seen it.

1:11:13 Uh, I'm a little bit zoomed in here on my Mac screen, sorry, but we'll, we'll scroll.

1:11:18 Uh, so for Iron Viz, I focused in on genre mashups, uh, so stories that fit into more than one genre.

1:11:26 My viz starts off with this little sort of Mad Libs style Venn diagram, where you can fill in the blanks in a story, um, and as you do, it creates this, this Venn diagram of overlapping genre categories.

1:11:40 And then you can see, uh, how popular that little mashup in the center is.

1:11:45 From there, we start looking at, uh, mashups in general, what kind of mashups are most popular, um, and look at some mashups that have gotten weirdly popular, uh, like paranormal romance with the whole Twilight craze that saw a big, big surge in popularity.

1:12:02 Then the really biggest chunk of my whole presentation and viz is this final page, the next big thing, uh, where we start to look at a couple factors to figure out what genre combinations might be popular in the future.

1:12:15 Um, got this nice little accordion here that expands through, and you can see as I go through the different criteria, our web of possible combinations over on the right side is getting smaller and smaller, um, until we come to just four final potential options.

1:12:33 So that's the viz. Uh, let's talk about the build, uh, the design, the analysis, all of that that led to putting this together.

1:12:44 First step for me in building any viz is always design. Um, I am a very very visual person. Uh, design gives me inspiration for all the rest of the viz, and I actually have a really hard time starting something if I don't already have kind of a feel in my head for what I want it to look like.

1:13:02 For Iron Viz, for the book theme, uh, I chose a design that was very inspired by journals, and specifically, like, bullet journals, these very artistic, um, journals that some people do.

1:13:13 I've never been able to keep one going for more than two days myself 'cause it's so much work, but they look really, really pretty, and so I chose that as my design inspiration.

1:13:23 Um, I made a mood board. I always do this in Figma, pulling together i- images that kind of help to inform, again, just what, what the vibe of the build was gonna be, uh, what it should feel like when you looked at it.

1:13:36 So this is my mood board for the Iron Viz final, and you can see here a lot of elements that I used to make the final design. We've got orange highlighting. That became a pretty big component. Got some of, like, these hand-drawn banners, got these outlines, the dot background, the sort of handwritten fonts.

1:13:55 A lot of these pieces turned into parts of the final design.

1:13:59 And see I've also got here just another example. This was my mood board for Packing For Mars. You can see very different, but in the same way, a lot of these elements turned into the elements of that viz as well.

1:14:14 Once I knew what I wanted my concept to be and how I wanted the viz to feel, I had to figure out how to actually make the dashboard feel like that, because I couldn't just go through and start hand drawing all of my charts.

1:14:26 Guess there's nothing in the rules against that, but I don't think it would go very well.

1:14:30 Um, I needed to figure out a way to get that feel of the hand-drawn journal, while also still actually leveraging and building in Tableau.

1:14:38 And I used three main strategies to make that happen. The first one was background images, second one, custom shapes, third one, chart choices.

1:14:48 Background images, as you can probably see from the viz, were a pretty important part of making this whole thing feel right.

1:14:56 The image we're looking at here is one of kind of my first simplest background concepts. I think this is actually the first thing that I made for Iron Viz this year.

1:15:05 Like, before I'd really even figured out what was gonna be on the next couple pages, I knew I wanted this Mad Libs intro, and I made this background, and this was one of the earliest pieces that I actually put together.

1:15:17 And this is really simple. Um, it's just a bunch of rectangles with drop shadows in Figma, kind of just tilted at different directions, but it instantly sets the tone of what the viz is going to feel like.

1:15:30 From there, I came back to this towards the end of my process to add a lot of hand-drawn background elements, um, to really bring the whole thing together.

1:15:39 So you can see by the time we get to the final version, the drop-downs are, of course, Tableau parameters.

1:15:45 The Venn diagram is an actual chart. The text here is an actual chart. The arrow is an actual image.

1:15:51 But then the drawing, um, that brings in lots of different genre elements, the text, the banner, um, these are all part of the background.

1:16:01 I also leveraged the backgrounds to help me get non-essential text, um, in faster. So for example, this little description here, hover to see top books in this category, this is part of the background, um, and that was just for the sake of time with a twenty-minute build. Everything that's like an essential part of the viz is actual Tableau text, so that it's accessible, but for some of these kind of extras, um, that are more contributing to, again, the feel, I used the background to make it a little bit faster.

1:16:32 Next thing was custom shapes. Um, these were a huge part of how this viz came together and how I created that feel.

1:16:39 If we hop over to the viz itself, here we go, um, let's go look at this Venn diagram. See if I can click.

1:16:48 Oh, I'm in presentation mode. That's why it doesn't like me. There we go.

1:16:51 Here we go. We'll go to the sheet. Perfect.

1:16:54 So this Venn diagram, um, is a little…

1:17:00 I don't know, it's simpler than it looks, um, on the actual Tableau side.

1:17:04 All of the logic that figures out where these circles need to be and how big they need to be is being done with a bunch of trigonometry, um, in Tableau Prep.

1:17:14 The actual graph itself is just a pretty straightforward dual axis. It's just taking those X coordinates, taking the Y coordinates, and it's a dual axis so that I can have two layers.

1:17:25 Um, these are both shape layers. One layer, look at my shapes, is a custom shape, and you can see it right here. Um, I'm gonna click out of this so you can see it bigger, but this shape was drawn in a drawing program using like a highlighter style brush.

1:17:41 And you can see I kept some of the transparency in the shape, um, because Tableau custom shapes actually do maintain transparency. Uh, if your shape is partially transparent and you have like a PNG format, that will stay when you bring it over.

1:17:55 And so that makes this look very sketchy. It looks like it was actually colored in with a highlighter.

1:18:00 Um, the edges are all, like, crinkly and rough. Um, but each of these circles is just one copy of the same shape, um, layered over each other and a little bit transparent.

1:18:12 I also used a very similar strategy, um, on this network chart. Once again, these circles are custom shapes that were drawn using a highlighter style pen.

1:18:23 The only difference on this one is that the outline is actually part of the shape. Um, that's because these are smaller and so it's, it works okay, um, and it was faster than adding a separate circle map layer. But same concept here. If I go and just look at the shape, these are just their own custom shapes.

1:18:44 Um, so that made a pretty big impact in terms of making this feel like someone had come in and drawn these shapes on the page.

1:18:54 Final part, which is chart choices. Um, you can see on this page, none of these charts have custom shapes. Um, none of them actually have anything too crazy going on. I guess the lollipop chart is maybe a little wacky, but for the most part, these are pretty straightforward, almost Tableau native chart types.

1:19:12 Um, I didn't have time to make these fancier. My build time was too short. I originally wanted, like, custom shapes for these circles, and I just did not have time.

1:19:21 Um, but instead of choosing just a straight bar chart or a line chart with consistent width lines, um, I chose styling on these to try to make them still feel a little bit like someone could have come in and drawn these, um, even though they were a little bit more default charts.

1:19:42 Uh, key questions for the design, keeping it clean, make sure it wasn't distracting. That was a big factor with a lot of the hand-drawn elements. I wanted to make sure there wasn't so much there that that became the focus, um, they should be adding to the story.

1:19:55 And then trying to make sure it was as accessible as possible, and very importantly, that I could build it quickly. More on that later.

1:20:03 Step two here was story. Uh, finding a good story was by far the hardest part of this build for me, um, 'cause I wanted to find something that I was interested in, uh, that had enough, you know, components that I could really support what I was doing. And then I was also panicked I would choose the same thing as Brian or Kevin, and then it would make the whole thing feel a little weird, and I was very glad in the end that we all went different directions.

1:20:26 I ended up going with genre mashups, so books with more than one genre category, 'cause I felt there was a lot to dig into there.

1:20:34 Um, as far as considering, you know, what makes that story good, I wanted to make sure that there actually was a story, that it had a beginning, a middle, and an end, that it had a conclusion that wasn't obvious, uh, from the beginning, and that it wasn't just a bunch of facts about something, that it was actually going to tell a story and take the audience on a journey.

1:20:55 Uh, my story flow, you saw this a little bit already, so I won't go too deeply into it. But we started with the hook, which was that Venn diagram and Mad Libs.

1:21:04 The intro, which was actually the second page, just giving some basics and introducing the question.

1:21:09 Then the third page did most of the heavy lifting, uh, with the deep analysis and the actual conclusion.

1:21:18 So the question here was: how do I get this into Tableau? How do I convey the story so that it's clear, so that it will make sense in only three minutes of talking about it, and that it stays interesting?

1:21:28 Um, the ways that I chose to focus on this were, first of all, through the layout that I chose, those three dashboards that you saw.

1:21:36 When I was planning this build, uh, my first version of this story had

1:21:43 eight dashboards, and I took that to Grant, my supervisor, who said, "You cannot possibly build this in twenty minutes." And I said, "Well, fine, you're right. I'll just squish it all onto this third page and use dynamic zone visibility," which was probably more work. Um, but it worked out.

1:22:01 Uh, we did cut it down to three pages, but then I leveraged dynamic zone visibility to make sure that this third page was still gonna have most of the stuff that I wanted it to have. Um, I cut it down quite a lot, but a lot of the key points I wanted to hit on those future pages just became part of this little accordion here.

1:22:21 So this accordion here, uh, uses my favorite thing, dynamic zone visibility. Uh, as you click on one of the headers, um, it expands to a new area.

1:22:31 I didn't have time to add the automatic de-highlighting, which makes me sad, but that's okay. It still works.

1:22:36 Um, but as we click through here, different charts are revealed, which helps to make sure that the audience can focus just on the part of the story I want to talk about, uh, without getting distracted by the next step.

1:22:50 From a technical perspective, there is a lot on this page being controlled by the dynamic zones and by one single parameter.

1:22:58 'Cause you might notice, for example, when I switch between, uh, number two and number three here, not only is the chart changing, the text is all changing, and the background of the page is actually changing.

1:23:14 Um, so look, for example, at this frame right here. Look over here at this circle around four.

1:23:22 That all changes, um, as I flip between these zones.

1:23:27 So on this dashboard, not only do we have these individual containers being controlled, um, we actually, if we go look at the layout, we actually also have these overlay images that are also being controlled using the same dynamic zone controllers.

1:23:45 Another thing, um, that is kind of interesting here from a layout perspective and with those dynamic zones, uh, this section or any of these sections where you've got some text, a chart, and then some more text, feels like it should be a vertical container. Um, and really, it should be, but vertical containers are slow to build.

1:24:05 So this is actually a text box with a big giant space in the middle that I calculated ahead of time to be the correct size to fit the chart, and then the chart is just floating on top of it. Um, and both are being controlled by dynamic zone visibility.

1:24:22 Uh, if we go hop into one of these charts real quick, I will show you, um, this control, uh, because every single thing on this third page is actually being controlled by one single parameter. Um, it is just this accordion state parameter.

1:24:38 And if we look at this, you'll see that it is just a float that can be anywhere from one to four.

1:24:46 Um, so this page has four different states that it can be in, um, and those states are being used to both create the fields that are controlling what dynamic zone is shown, but they're also being used to filter and color-code this network chart.

1:25:04 Um, so if we go look at this one, you can see this network state filter. I'll open that up.

1:25:10 There we go. This is being controlled by the exact same accordion state.

1:25:15 Um, this is being used to show what am I filtering on as I go through each state of the page. And you can see kind of the further we go through, the more filters, um, are being applied to rule out those combinations that the, the different states are ruling in.

1:25:34 So dynamic zones, big part of how I translated this story onto the page.

1:25:41 Uh, likewise, parameter actions. We just looked at that parameter a little bit. The thing that is changing the parameter, um, is actions on each of these headers. So when I click this, the dashboard has A parameter action right there. It's very simple. Um, it's the same one on every sheet, it's just being applied to all of them, that is changing that accordion state to the next number.

1:26:05 Step three, analysis. Um, this wasn't really step three. It kind of happened at the same time as storytelling, but we've got to explain things in some kind of order.

1:26:13 Uh, but as I was coming up with that story, I was also making sure that my data actually supported that story.

1:26:19 Um, this involved, just like Kevin and Brian, a lot of Tableau Prep.

1:26:24 Um, I had to do a bunch of different things in Tableau Prep.

1:26:28 First thing was genre cleanup. Uh, our original source had over 1,000 genres, um, some of which were just really specific, some of which were not actually genres. Um, shout out to the 30 books that listed Google as a genre.

1:26:42 It's not a genre.

1:26:44 Uh, I mapped them down to 17, um, mostly actually using Tableau's grouping tool, um, to help me limit them.

1:26:51 I limited the data to just what was relevant. So, for example, I got rid of all non-fiction.

1:26:56 I calculated all of my metrics, um, and then I separated my sources. Um, this was a big part of optimization.

1:27:03 Every single chart in my build, um, has a data source that is optimized for it. Um, and most of these charts actually… Sorry. Most of the data sources only feed one or two charts. You can see here I actually have six separate data sources that all came from the same original one, um, and each one has been optimized to make sure that it is, uh, correctly, um, set up for the chart that it is going to be building.

1:27:31 All right. This is part of my Tableau Prep flow. It is not as neatly organized as Kevin's. Um, it's a bit of a mess. I put, put comments on, like, three places. So for me, that's pretty good.

1:27:42 Um, this is about half of it. There's about this same amount before that you just can't see. Um, most of this section is calculating metrics, uh, and separating and optimizing data sources. The previous section you can't see was mostly about mapping genres.

1:27:59 My sort of final challenge, uh, for this build was making it fast, uh, making it able to be built in only 20 minutes. Um, the first time that I practiced my build through with, like, I thought I had everything prepared, I was ready, I was optimized, it took me 50 minutes, uh, which was not great. Um, I had a lot of work to do to get this build down to just be 20 minutes.

1:28:25 Um, and of the three of us, I think I was definitely the tightest on time in the end, um, and the most panicked about time.

1:28:32 Um, this was a big challenge.

1:28:35 To get faster, I did a couple different things. Um, I optimized heavily. So I talked about before, like, leveraging the background images to get some of that unimportant text.

1:28:47 I made sure the data sources were exactly ready for their charts. Like, no changes needed to be made, no re-aliasing, no extra fields.

1:28:56 All of my calculations were in a tech stock ready to come in. Um, and then I reused elements whenever possible, like using the same bar chart or lollipop chart from the second page on the third page.

1:29:08 I had to cut a lot of stuff. Um, I couldn't fit in everything that I wanted. There was a lot that I wanted to include that I couldn't, um, and things that I just had to let go.

1:29:17 Um, and then I had to speed up. I had to learn how to move faster in Tableau. Um, I practiced many, many times. I drank a lot of coffee. Um, I learned a lot of keyboard shortcuts. Uh, I'll show you one of those quickly. I found there was some interest in those earlier.

1:29:32 So, for example, uh, this data source i- is used to make the line chart that shows paranormal romance's spike.

1:29:41 Um, and this can be built in one click, uh, using keyboard shortcuts.

1:29:46 So it is Windows + Shift + O.

1:29:50 Nope, no. Alt + Shift + O to add to color.

1:29:54 Uh, then I to add to size. You can see I'm just using the arrow keys to move down.

1:29:59 Uh, C to add to columns, S to add to detail for some reason, R to add to rows, and then all I have to do is click this and make it continuous, and this chart is done.

1:30:14 Uh, so because the data source is optimized and because I practiced exactly what moves I was gonna have to do to make this chart, um, a lot of the individual charts could be built very, very quickly.

1:30:26 Overall, like Kevin and Brian have said, uh, this was an absolutely incredible experience. Getting to do it with them and with the rest of the team was just amazing, um, and it was all very, very worth it.

1:30:39 That is all I have to say.

1:30:41 Any questions?

1:30:43 Agata Ketterick: Do you use any of the shortcuts you learned, like, in your, in your day-to-day life now?

1:30:49 Ann Pregler: Never. It would take so much longer to look up what they are and remember what they are.

1:30:52 Agata Ketterick: That's totally what I feel about them. Like, dragging

1:30:54 Ann Pregler: I have the ones that are generally written down, because I don't remember them anymore.

1:30:58 Um, I do use the thing about optimizing the data source though. Like, that's actually really, really helpful to have data sources that are targeted for charts.

1:31:04 Agata Ketterick: Yeah. Bet.

1:31:07 Um, I did see a question here. I'm just trying to, like, go back up here to find it. Oh, it was Rob's.

1:31:12 Um, he wanted to know if he get a set of guidelines before the final of what needs to be done within Tableau and what else you can or cannot use.

1:31:22 Ann Pregler: Yeah. Um, we had some limitations. Like, we were allowed to do any data prep in Tableau Prep that we wanted with no limitations, um, but we couldn't use other tools to prepare the data necessarily. There were some…

1:31:38 That was kind of a little bit more like a case by case thing. Um, but we couldn't, like, uh, use code to prepare the data or something like that. It had to be done in Tableau Prep.

1:31:47 Um, we were allowed to bring in any images that we wanted to, uh, but all of the charts had to actually be in Tableau. Um, we couldn't use any AI elements. It had to all be us.

1:31:59 Agata Ketterick: Awesome.

1:32:02 And how many times did you practice the whole run before D-day?

1:32:06 Ann Pregler: Uh, look, I found, um s- it really, coffee did make a difference. I found, like, from the evening to the morning after coffee, I was like four minutes faster.

1:32:18 Agata Ketterick: I feel like it would make me shaky after a while.

1:32:20 Ann Pregler: Yes. I was very shaky on s- on stage. We had to make sure I had coffee beforehand. I was very worried about it.

1:32:28 Agata Ketterick: All right, Rob. I think that was it. You wanna wrap us up?

1:32:36 Rob Taylor: Brilliant. Thank you. They were, uh, all fantastic. I love all the tips and tricks and all the clever stuff you showed us. There's so much there for anyone that wants to enter our viz going forward.

1:32:46 Um, uh, hopefully you'll let them reach out to you if they have any questions as well.

1:32:50 Um, thank you all for joining. So, uh, next meeting is October 22nd, uh, and it's a bit of a, a games night theme for the viz. So we've got some great presenters. We've got Tina, Lewis, and Katie, who'll all be, um, coming along to share their dashboards in, in the community spotlight. So we hope you can join us then.

1:33:07 Um, but mostly we just wanna say thank you very much to our three presenters, uh, all our hosts, and everyone that's joined us for this. We hope you've enjoyed it as much as I have, um, and we look forward to seeing you next time. Thank you very much, everyone.

1:33:18 Eric Howard: Absolutely.

1:33:20 Thanks everybody.

1:33:22 Agata Ketterick: Thanks everyone.

1:33:31 Ann Pregler: Bye all.