Skip to content

Summary

The Higher Education Tableau User Group's April Fools' meeting, billed as "the first, last, the only March 32nd HETUG meeting" and held on April 1, 2026, recorded by the group on its own YouTube channel and republished here with credit. HETUG leads Ginny Moench, James Emery, Sam Homier and Roshni Gohil open, amid a steady stream of egg and chicken puns, with Tableau Conference 2026 news: an education registration discount, the 2026 Iron Viz finalists (Ann Pregler, Kevin Wee and Brian Moore), thirteen education-related sessions, How to Do Cool Stuff in Tableau with Ken and Kevin Flerlage, and, for the first time, an education-specific hands-on training session, "The Dean's List of Measures: Mastering Higher Ed Data", built on a new school data source the education TUGs helped create. Ginny also promotes the free Analytics Day on June 11. In "Meet a Community Chicken or Egg", Sam shows a dashboard tracking the eggs from Sam's own hens, with universal title formatting, parameter-driven pages using dynamic zone visibility, and tooltips that adapt to the count or the value, and Ginny walks through Chicken Tinder, a mobile-first "dating app for poultry" viz, with tips on designing the phone layout first, using floating objects for pixel-level control, and checking the result on Tableau Public. Michael McCusker, Tableau tech lead at DaVita and San Diego TUG co-leader, shows how map layers can build non-map charts: plotting points with MAKEPOINT (Y first), turning the background map off, rebuilding a BAN with icons and with selection disabled on individual layers, and combining a twelve-worksheet KPI grid into one fast-rendering view, plus a plug for Data + Tacos on May 4. Finally, HETUG co-leader emeritus Andrew Drinkwater of Plaid Analytics demonstrates open-source data lineage with DataHub (and mentions OpenMetadata), tracing a Tableau dashboard back through published data sources, a data warehouse and an ETL workflow to the Banner registration table, and argues that living documentation, glossary terms and automated metadata ingestion matter all the more as teams adopt AI tools.

Key points

  • For the first time, Tableau Conference 2026 includes an education-specific hands-on training session, built on a school data source the education TUGs helped design.
  • Sam Homier: set title formatting once for every sheet, and build tooltips that adapt to the count (one egg, three eggs) or show fully custom text for each value.
  • Parameter-driven pages with dynamic zone visibility let one dashboard hold several views.
  • Ginny Moench: design the mobile layout first, since it's easier to scale up than to squeeze down, and always check floating layouts after publishing to Tableau Public.
  • Michael McCusker: map layers aren't just for maps. Plot your own points with MAKEPOINT, feeding the Y coordinate first, then turn the background map off.
  • Disabling selection on individual layers removes the blue highlight box while keeping other elements clickable.
  • Combining a twelve-worksheet KPI grid into one map-layers worksheet makes the dashboard render almost instantly.
  • Andrew Drinkwater: open-source tools like DataHub can trace lineage from a Tableau dashboard back to the student information system, beyond what Tableau Catalog shows.
  • Clear definitions, lineage and documentation keep AI tools from "just getting to a wrong answer faster".

Jump to a chapter (41)

  1. 0:00 Welcome to the "March 32nd" HETUG
  2. 1:26 Tableau Conference 2026 announcements
  3. 2:21 The 2026 Iron Viz finalists
  4. 3:30 Education sessions at TC26
  5. 3:49 How to Do Cool Stuff with Ken and Kevin Flerlage
  6. 5:20 The first education hands-on training session at TC
  7. 8:26 Watching TC virtually on Salesforce+
  8. 9:03 Analytics Day 2026
  9. 10:34 Which came first, the chicken or the egg?
  10. 11:23 Sam's Fresh from the Chicken Coop dashboard
  11. 12:38 Universal formatting and parameter-driven pages
  12. 14:12 Tooltips that fit the count
  13. 16:10 Fully custom tooltips by value
  14. 17:44 Meet the hens
  15. 18:48 Ginny's Chicken Tinder viz
  16. 24:46 Design the mobile layout first
  17. 26:29 Checking floating layouts and a Flerlage scrolling timeline
  18. 28:31 Introducing Michael McCusker
  19. 30:15 Data + Tacos and a first TC talk
  20. 32:56 What map layers can do
  21. 35:43 MAKEPOINT: thinking in X and Y
  22. 36:32 Demo: doodling with points
  23. 41:04 Rebuilding a BAN with map layers
  24. 43:08 Disable selection, layer by layer
  25. 46:19 KPI cards: from 12 worksheets toward one
  26. 47:25 Current-year line and prior-year area
  27. 51:36 Custom axis and small multiples in one view
  28. 51:47 One worksheet renders almost instantly
  29. 53:11 Tables with map layers
  30. 55:12 Introducing Andrew Drinkwater
  31. 57:01 The dashboard chicken-and-egg problem
  32. 58:10 Why data governance is hard in higher ed
  33. 59:36 Open source options: DataHub and OpenMetadata
  34. 1:01:11 Demo: tracing a dashboard back to its sources
  35. 1:03:12 Lineage from dashboard to data warehouse
  36. 1:07:20 ETL and the Banner registration table
  37. 1:10:41 Glossary terms and linking back to Tableau
  38. 1:11:46 Automatic metadata ingestion
  39. 1:13:01 Governance as a foundation for AI
  40. 1:14:23 Questions and wrap-up
  41. 1:17:10 Upcoming conferences and close

Transcript

0:03 Virginia Moench: Hello, everyone. Happy New Year.

0:06 Wait, wait, that's the wrong holiday.

0:09 Let me try that one again.

0:13 Happy Friday. Thanks for joining us for the first, last, the only March thirty-second HETUG meeting on April first, twenty twenty-six. We're so glad that you could join us today.

0:27 Boy, have we got a fowl, I mean, educational agenda for you today. Sam and I will be eggucating you about egg, I mean, data collection and designing with your chickens, I mean, your users in mind.

0:44 Um, Michael McClucker, I mean, McCusker, will show us how to use map layers for non-map charts, which I'm personally super eggcited about.

0:56 And to wrap it all up, Andrew Drinkwater will share some open source options for tracking your chicken, I mean, data lineage. Uh, whew, that was a basket full of goodness. So before we get started, does anyone have any announcements that they would like to share? Feel free to come off of mute, put it in the chat, anything that you think would be, um, of interest to our group.

1:20 No announcements? Well, that's okay, 'cause we've got plenty of announcements for you too.

1:26 Starting with the Tableau Conference from May fifth through seventh in San Diego. Lucky people who get to go should be sure to take advantage of this eggucation discount, a whopping six hundred and ninety-nine dollars off the registration price for Tableau Conference.

1:43 Put that in your Easter basket.

1:46 And people who come to TC will learn new skills, see what's coming next for Tableau, connect with community. It's the fastest way to level up if you work with data. And there's going to be eggucation meetup for those of you, um, to meet other crazy cluckers like yourselves, and all sorts of other community meetups like Data + Women, Data + Tacos, which Michael's gonna talk about some more shortly.

2:12 Um, and then events on Monday, May the fourth be with you.

2:16 So I'm gonna turn it over to James now to talk a little bit about Iron Viz next.

2:21 James Emery: Well, just in case you missed the announcement or as of last month, the Iron Viz qualifier results are in, and you can use the QR code to learn more about each of the top ten vizs.

2:31 Now, in those top ten results, you will find our very own HETUG co-lead, Ginny's viz, um, that was inspired by Jane Goodall.

2:40 Um, it's named, uh, Can Eating Less Meat Save the Planet? Now, Ginny will be presenting, uh, about her viz at our next meeting on April twentieth, so be sure to register for it.

2:53 Now, taking the stage at TC will be the three finalists, Ann Pregler, Kevin Wee, and Brian Moore.

3:00 Now, you can learn more about the finalists on their Data + Love podcast interviews. So Ann and Kevin's interviews are already posted, and Brian's is scheduled for the next episode, posted in a week or two. And if you're going to TC in person, a good opportunity to meet the candidates will be at the Data +, uh, Women event that, uh, Ginny mentioned just a moment ago.

3:21 So while Iron Viz is a highlight of TC, it is not the only session to watch out for. So Sam is going to mention some of the other ones to attend.

3:30 Sam Homier: Thank you, James. Yes. So there are thirteen sessions related to or given by people who work in education that we wanna point out to you. So be sure to favorite them if you are attending, 'cause that-that's the organizers know how many people are interested. And then Roshni just put all the links in the chat.

3:49 Roshni Gohil: So there are a lot of can't miss moments at TC, and there's one more session that we have to call out for education. It's How to Cool… Do Cool Stuff in Tableau with Ken and Kevin Flerlage. And you might be asking yourself, "Well, why is this being called out as an education session?" Well, my favorite Flerlage, that guy right there, uh, Ken, used to work in higher ed at Mc- at Bucknell.

4:11 And though he has defected and gone over to Moxy Analytics, he will always be higher ed in our hearts. And I learned at last month's official March HETUG with the Tableau Public and Education TUG, uh, spring break session, that Kevin is actually a little bit more higher ed adjacent than I previously thought. It's not just that he's brothers of- with Ken, but Kevin used to work with Jeff Shaffer.

4:34 That's the r- the Jeff Shaffer of Big Book of Dashboards and Dashboards That Deliver, who's a professor at University of Cincinnati. So Kevin might be stepping up a little bit in my f- favorite Flerlage meter, but Ken still, still at the top. Um, and regardless of whether How to Do Cool Stuff in Tableau is an education session or not, um, these guys are folks that you do not wanna miss at the conference.

4:55 And even Flex, that's right, Flex, this little dinosaur over here, the one sort of in the middle, um, and our opening DJ Cody B, um, he also doesn't wanna miss the session. And he's partial to Keith, so you know it's gonna be a great session, uh, 'cause, you know, we always learn stuff from Ken and Kevin, and even Keith. So make sure you favorite this session, uh, in the catalog as well.

5:20 Virginia Moench: And lastly, but not leastly, we want to call out that EDU is H-O-T-T-O-G-O, and I'm being told… Wait, what is that? Ugh, sorry. Chaperone is featured on the slide, but sadly, she will not be joining us at TC.

5:36 My tiara fell. Perfect timing.

5:39 However, for the first time ever at Tableau Conference, there is going to be an industry specific hands-on training or HOT session, HOT session.

5:51 And guess what?

5:53 We're the industry. The co-leads and a bunch of other education TUGs, including HETUG, CCCTUG, and K through twelve TUG, have been working with Harold Thomas, one of the HOT session instructors, and other Tableau solutions engineers, including Jeremy Wetherall, the New England Brit with magic coding skills.

6:14 To build out a brand new super school data source that is going to be used in this hot session. Can you believe it?

6:24 Roshni Gohil: Woo-hoo.

6:24 Virginia Moench: Are you as excited as I am?

6:27 So what does this mean?

6:29 This means we get real education related use cases to work through in exercises.

6:37 All right, let me say that again.

6:39 Let it sink in. We get real education related use cases to work through in exercises. This is not a joke. I am serious. During the hot session named the Dean's List of Measures: Mastering Higher Ed Data, attendees will work through hands-on exercises to answer questions like, how many students persist from term to term or year to year? And then we'll deep dive into grade distribution and LODs.

7:06 We're in for a treat, and super excited about this new development in education related TC content. So tell your friends, favor these sessions and when you see them in the catalog, and celebrate this amazing win for education at Tableau Conference within the Tableau ecosystem.

7:25 Roshni Gohil: Can we add another woo-hoo to that?

7:28 Virginia Moench: Yeah.

7:29 Roshni Gohil: Excellent. It's gonna be amazing.

7:32 Virginia Moench: It's only been a year or so in the making, I mean.

7:35 Roshni Gohil: Two… More than that. We started talking about this six months, seven months before last year's TC.

7:42 Virginia Moench: That's true.

7:43 Roshni Gohil: So yeah, we're really excited that this is actually happening this year.

7:47 Sam Homier: And then if you are in, uh, education, or sorry if you are in… Yeah, if you're here from the industry, then there is a new discount code for you, which I believe, and somebody correct me if I'm wrong, but it's half off, right?

8:03 Or is that wrong?

8:05 Roshni Gohil: Something like that.

8:06 Sam Homier: Yeah.

8:06 Roshni Gohil: We're not entirely sure. It-

8:08 Sam Homier: But if you-

8:09 Roshni Gohil: But there's a discount.

8:10 Sam Homier: Yes, there's a discount code. It is, at this point, not as much as the discount for education, um, affiliated like institutions, but if you are from the industry and you would like to use that, that is there for you.

8:26 James Emery: Well, if you cannot make it to TC in person, there's no need to bruise because there is a virtual option.

8:33 There are many sessions, including the major sessions like Iron Viz will be, uh, streamed on Salesforce Plus. Now, if you see a session in the catalog that is not on Salesforce Plus, let us know so we can check if, if we possibly do a HETUG presentation in the future.

8:49 And even if you are at the conference, you can still use Salesforce Plus to watch sessions that you missed, so don't forget to register for it now.

8:58 Now, Ginny's going to, uh, mention another virtual conference for you to attend.

9:03 Virginia Moench: Yeah, that's right. So Analytics Day 2026 is coming at you on June 11th from 8:30 AM to 5:00 PM Eastern. You can register, it's free. Registration's open now. Uh, the deadline to register is May 28th, and if you'd like to sub- uh, submit a proposal, do so by May 8th. Um, it… Once you register, you should get a automatic email with your registration information, and also the link to submit a proposal.

9:30 So make sure you register first. And did I mention it's free?

9:35 In addition to that, um, what you can expect at Analytics Day, uh, aside from the many, many excellent sessions that we have every year, uh, the Ultimate viz Showdown is back for the fourth year in a row. So Tableau versus Power BI rematch.

9:51 Which one will make the cut? You be the judge. I think we all know the right answer though.

9:57 And then we're also going to feature our guest speaker, Andy Kriebel. You might know him. He's a Tableau Visionary Hall of Fame member. He's a Tableau ambassador and a 28 time Tableau Conference speaker, as well as the founder of Makeover Monday, Workout Wednesday, and Next Level Tableau. I mean, how does he find the time? It's amazing.

10:17 So lots of fun stuff in store on Analytics Day, so definitely sign up. Even if you can't make it for the whole day, you can pop in and out. If you can't make it at all, you can watch the recordings later, but you have to sign up to get the links for that.

10:31 Roshni Gohil: Up next is…

10:34 Wait, what? Sorry, guys. Flex, he's, he's here. You can see he's here. He's got a question for you guys.

10:41 Um, so he's feeling a little shy. He doesn't wanna ask for himself.

10:48 Ginny, Sam, what came first, the chicken or the egg?

10:52 Sam Homier: Obviously the egg.

10:56 Virginia Moench: No, I have to disagree with you. It's definitely the chicken.

11:01 Roshni Gohil: Flex, what's the answer?

11:04 Oh, that's a good one. Um, Flex says that you're both wrong.

11:08 It's a dinosaur.

11:12 Virginia Moench: Well, chickens are descended from dinosaurs. That kind of makes sense.

11:16 Roshni Gohil: Kind of does.

11:18 All right, Sam, take it away.

11:21 Sam Homier: Okay. Thank you.

11:23 Um, oh, can you all see this chicken coop dashboard?

11:29 Virginia Moench: Sure can.

11:31 Sam Homier: So since we got our chickens back in September, we have been weighing their eggs to just…

11:38 Because I love collecting data and tracking data, so yes, I wanted to start weighing their eggs immediately. Um, they actually didn't start laying until November though. So I made this dashboard to kind of just to answer some questions that I had and to track their egg production. Um, it actually is important because if they start laying less, then it's a sign that they're sick.

12:02 So this date laid chart is actually very useful for me. Um, but I wanted to have some fun with it as well.

12:09 And so as I was building this, I started going through and realizing that even though this was just for fun, a lot of what I was doing was the same as things I build for- Everett Community College, where I work, right?

12:24 Um, and so I have some tips, mostly about tooltips, um, for how to– and I think we skipped my slide, but that's okay, um, for how to just kind of make it more, um, user-friendly.

12:38 So the first thing that I usually do is universal formatting.

12:44 And so in every title on my chart has the same formatting, which I can do all at once. So I can just go here, set the title to… Oh, why is it… Oh, sorry.

12:56 Do different, uh, size, color, font, and it'll change all of them at the same time.

13:02 If you have subtitles though, it's going to mess with that, so… `Cause technically it's, it's all one title. Yeah.

13:09 Um, but another thing that I usually do for all of my chart or all of my dashboards is I'll have pages, um, which is set with a parameter, right? So down here, I have multiple pages, so you can click through and see different information, and it's a l- squished because of the whole resolution thing, right? Um, so I'm just very quickly going through this.

13:34 And this is, yeah, set with a parameter. And so… Oops.

13:39 Down here, I just have the page numbers, and then I use dynamic zone visibility to change it based on the parameter that is selected.

13:49 And so here I have… Oh, where did they go?

13:55 All of the dynamic zone visibility formulas that are necessary for that.

14:02 And then really, um, let's see. So the– Let me pull this up real quick. Sorry.

14:10 Ah.

14:12 The next thing is the tooltips. So almost every single chart that I have up there has a customized tooltip based on, um, the count, right? So for instance, if we're dealing with only one student at a time, I don't want it to say one students.

14:27 So I'll chain– I'll add the S based on how many students there are. And I did the same thing here based on the number of eggs.

14:33 So if they laid three eggs, it says eggs, but if they laid one, it only says one egg.

14:39 And so to do that,

14:43 I created a formula. So here, count of eggs, right with the egg. And so it's just an if only egg or if else if formula. Yeah. So if the egg count is one, don't put anything, and if it's not one, then put…

15:01 Right.

15:04 Um, so yeah, here the hens laid two hundred and fifty-six large eggs up until yesterday. I think the chart says June thirtieth. I have to go in and fix that. But this is up until yesterday.

15:14 But they've only laid one jumbo egg, so this is more than seventy point nine grams or something like that. That was actually a very large egg. Yeah.

15:22 Um, and then you do the same thing here as I was showing earlier. So they laid three eggs versus one egg.

15:31 Um, I've also– So this looks als- uh, zoomed out here, but it looks better when it's actually on the chart.

15:40 But another tooltip that I had to do here was based on where they laid.

15:45 So my chickens really like to lay in the corner of the chicken coop rather than in an actual egg box.

15:51 So for the tooltip, I didn't want it to say box D. I actually wanted to call out that that's in the corner of the chicken coop. I don't know why. It's their favorite place.

16:00 Um, and so now this one says they laid in the corner versus in box C or A or B.

16:10 Um, and then sometimes I need to do a completely customizable tooltip based on the value.

16:19 So here, this is accurate. I double-checked this. We have ninety-one eggs sitting on our counter.

16:26 Um, and we have gifted most of the other ones, to be honest, because we just don't know what to do with all the eggs we get from our chickens. They're– We mostly have them for a hobby.

16:36 Um, but so here the calculation, it's a little complicated, but it's based on if the usage is null, they're still unused. I don't have a use for them yet, so that's an alias there. If it's a breakfast scramble, I wanted to add the note about how that's different from scrambled eggs, right?

16:56 Um, and then if it ends with E-D, then I can just say, "We discarded them, we dropped them, we hard-boiled them." Right? So they all kind of get the same format there.

17:08 And then everything else also gets the same format of we use them to make cookies, we use them to make brownies, and so on.

17:17 Um, so that's completely customized tooltip based on the value that is actually selected.

17:24 And I did, um, something similar for the histogram, right? So based on the bin that you're in, it changes the values.

17:33 And then I did the same thing for the pages. So click to see the weight, egg weights, click to see the calendar, and click to see the actual hens who did the hard part.

17:44 Um, and I really quickly just wanna showcase the hens because they're the whole reason for this dashboard. Um, and one of them started laying blue eggs two days ago, so I'm gonna start collecting color data, and that's Dorothy. She started laying blue eggs, so that's pretty awesome. Yeah.

18:03 But yeah. So those are just some things that I found myself doing, both for fun and in personal dashboards or in work dashboards.

18:12 Oh, yes, and these are the Golden Girls. If you notice their names at the top, this is definitely the Golden Girls. Yeah.

18:20 But, and then I will pass it over to Ginny to talk about her chickens.

18:27 Virginia Moench: I absolutely love, um, your viz, Sam.

18:31 Sam Homier: Thank you.

18:31 Virginia Moench: I really wish that I kept better track of, you know, what my chickens are laying, but I don't have time for all that. We probably do have close to ninety-one sitting on the counter as well.

18:45 Sam Homier: It takes a while, yeah.

18:46 Virginia Moench: It does.

18:48 All right, so welcome to the Chicken Tinder. Yes, that's Tinder with an I as in the dating app, not Tender with an E like me, the one who cares for the chickens. Um, so this dashboard was all about creating something fun, funny, and mobile friendly because most people who use a dating app do it on their phone, and I don't, you know, get a chance to make many mobile-friendly layouts on Tableau, so this was my chance to do it.

19:19 So I start here with this disclaimer screen because, hey, we can't be too careful these days. "Welcome to Chicken Tinder, the dating app for poultry. Disclaimer: This site is intended for amusement only. If barnyard humor offends you, please don't enter." But we're not easily offended, so we'll click enter.

19:38 I probably have to refresh this.

19:46 All right, so once you do enter, you're gonna see this instructional overlay for how to interact with the viz. Um, so I did this using hidden containers 'cause I built this back in twenty twenty-one, uh, and DCV hadn't been invented yet. But if I was going to rebuild this now, I would use dynamics and visibility to include these instructions.

20:07 So in case you're not familiar with Tinder, and I wasn't, I had to look up images of what the screen should look like.

20:13 You start by viewing the profile of a chicken, and after looking at the picture and reading about them, you can swipe left if you're not interested or swipe right if you are interested. And then after going through the reel, down here at the bottom, you can click a star to view your matches.

20:34 And you can click on a heart to change which chicken you're viewing or click the arrows to move the selection forward and backward. So are you ready to meet your potential coop mate? Let's click here to get started.

20:47 Now, Boy Blue is our first chicken, and he's a Bantam Silk Blue Silkie Roo. So let me break this down for any non-Chicken Tinders in the room.

20:56 Bantam is a chicken term for a small bird as opposed to standard, which is a large bird.

21:03 So we know he's not gonna get bigger than this. He's fully grown.

21:08 Silk Blue is the variety or color of the bird.

21:13 Silkie is the breed, and Roo indicates that this is a full-grown male.

21:20 So his profile says, you know, "Some people call me a hermit, but really I just haven't met the one, you know, the one who loves me as much as my mother, who will help me come out of my shell, so to speak.

21:35 Uh, because inside this rough exterior I'm really just a poet with a gentle soul.

21:41 But come near me with that feed bucket and I'll peck you till your hand bleeds." I should say that all these profiles are based on the true personalities of these birds.

21:51 All right, so I'm not particularly interested in somebody who is a, a hermit.

21:59 I need somebody more extroverted like me. So I'm gonna click left and then go on to my next, uh, chicken. This is Candy, a Bantam, again small, Mottled, the color, Cochin, the breed, Hen, which means she's a full-grown female.

22:17 So, "Candy is a cougar seeking a young Roo for walks in the garden and special occasions." Well, I'm definitely not her type. I'm not a Roo, so I'm going to pass on Candy as well.

22:30 Next, number three is Charlie.

22:33 Now, this is our first standard or large fowl chicken.

22:38 She's Cuckoo, which is her color. That is the dark feathers with the, um, like, white tips.

22:46 And she is a Maran. Maran is a breed that lays a very dark egg. I don't know if you have any of those, Sam.

22:55 No.

22:55 All right, and then she's a Hen, so she's full grown. "Charlie's favorite song is Fat Bottomed Girls by Queen, and her favorite food is scrambled eggs. Her favorite coop is the Cochin pen." All right, sounds like a potential match, so I'm just gonna swipe right on Charlie.

23:14 Next up

23:17 is Cookie. Cookie is another Bantam, so she's small. She's Mottled like Charlie. Um, and she's just Silkie Serama. Silkie does a, in this case, a feather type, which is not a designation we've seen in here before, and Serama is the breed.

23:34 And Cookie says, "I may only have one eye, but I'm the sweetest girl in the chicken yard." And that sounds good, so I'm gonna swipe right on her too.

23:43 We'll do one more, Echo Dot. Echo Dot is our Bantam Silver-Laced Wyandotte Hen. Isn't she gorgeous?

23:52 She says, "I would make the perfect spy because I'm good at flying under the radar.

23:56 I believe I lived in Elizabethan England in a previous lifetime. Sometimes I go broody, but it's rare." And that sounds like a good potential match also, so I will swipe right on Echo Dot.

24:10 And I could go through all of these with you. We could hit the arrow and go forward.

24:17 As you can see, the heart is pink when you're landed on it, and it's purple if you're not. But this goes all the way up through nineteen chickens, so I'm gonna let you all Do this on your own time. But for now, let's go back and view my matches by hitting the star.

24:34 So I've got three matches, Charlie, Cookie, and Echo Dot. And that is the Chicken Tender viz.

24:41 But now let's look under the, under the hood, so to speak.

24:46 Um, one thing I wanted to say about when you're designing for mobile layout is that my biggest tip, my biggest takeaway is design the mobile layout before you design a desktop or any other size layout, because I think it's easier to start small and go up from there as opposed to starting big and trying to shrink it down. So, um, if you're not familiar, then when you're on your dashboard building screen, you have this default, and you have, um, different devices that you can default to.

25:28 And once it's locked, you're not gonna be able to get in here and do any editing. But if you did want to go in and edit, you can unlock it so that you can set different things like the default size, fit all, fit width. In this case, I went with fit width, and I did a custom build because I, I did not want to squish it too much. I want people to be able to, you know, scroll and do whatever they need to do.

25:52 And I also always prefer doing a custom layout because sometimes when it does a default, and it just puts things where it wants and not necessarily where you want them to go. And I think for, um, and I think for when you're, you're building for user experience, you, you definitely want to have that control over everything. And that's, um, another thing.

26:15 I use floating, uh, images, floating text boxes. I didn't, um, use any tiling in this whole viz because that allows me to be, like, pixel perfect.

26:29 Um, but also remember when you're doing floating containers and then publishing to Tableau Public, you have got to check it and make sure that it renders correctly because it doesn't always. Some things are always just a little off, and you might have to go in there and tweak it just a bit more.

26:46 Um, so yeah, that's one thing you should always check.

26:50 And the last thing I wanted to mention is that this, uh, parameter-driven, um, timeline is actually based on a timeline that the Flerlage twins wrote a blog about.

27:07 And I wanted to use the arrows to go back and forth because fitting 19 of these into the screen on a mobile layout, it just… It, it was too much. There's not enough real estate space. So, um, I found this blog and was able to incorporate some, uh, some parameter-driven actions to change this, the numbers that are showing up on your screen. And I am not gonna, uh, go into it right now.

27:38 I do have the link that I'll put in the chat so that you can go and read the blog on your own because it's fabulous.

27:46 Whoops. I guess I have to hit enter.

27:49 Um, and really all their blogs are fabulous. So yeah, that was a very important part of this, was designing for the limited space that I had. So that's the scrollable timeline, and that's the Chicken Tender viz.

28:03 And I hope that, uh, you take the time to go and explore the other remaining, what is it, 14 chickens, because they're all pretty funny, and, uh, if I do say so myself.

28:19 Sam Homier: Thank you. That was amazing, Ginny. I, I love hearing about your chickens.

28:24 Virginia Moench: They're pretty funny.

28:25 Sam Homier: Yes. Yeah. Oh, yeah, your descriptions are hilarious.

28:27 Virginia Moench: I need to do a bird one next.

28:31 Sam Homier: All right. And up next we have Michael McClu… Oh, sorry. Michael McCusker, who is a Tableau tech lead for DaVita, a healthcare company.

28:39 Michael recently transitioned to the role from a data analyst and manager after falling in love with Tableau. In this role, he gets to not only train new Tableau users, create a Tableau community, he leads the internal DaVita TUG, explore and enhance our Tableau capabilities, and importantly, he still gets to build dashboards.

28:58 Michael also co-leads the San Diego Tableau User Group. He has a Viz of the Day, and he's presenting at this year's Tableau Conference for the first time ever, so make sure you favorite his session.

29:09 Michael is passionate about coffee and likes to learn where the coffee has come from, how it was grown and processed, and explore different brew methods.

29:16 I have a feeling he's going to be chugging coffee during TC to keep up with all of his engagements, from his presentation to Iron Viz to…

29:24 Did we hear something about a special May the Fourth celebration?

29:30 Michael McCusker: Yep.

29:32 Absolutely did. I can get onto that in a second.

29:37 Sam Homier: Take it away, Michael. Thank you for being here.

29:41 Michael McCusker: All right. Let me share my screen.

29:48 Can everybody see that and hear me okay?

29:52 Sam Homier: Perfect.

29:52 Michael McCusker: All right. I'm embarrassed because I have zero egg or chicken puns in this entire presentation. Um, I was trying to quickly think of edits on the fly, but, um, apologies.

30:04 But this is what I've got for you today. Um, really fun TUG. I'm, I'm thrilled to, uh, join, and thank you for having me on to come and talk a little bit about map layers.

30:15 As, uh, it was just mentioned, I'm Michael McCusker, Tableau tech lead at DaVita, and I'm one of the organizers of the San Diego Tableau User Group. San Diego, of course, home of Tableau Conference this year and for the past couple of years, and- Our user group is hosting an event called Data + Tacos the day before Tableau Conference, and it is on May the 4th, hence all of the Star Wars themes, and you may see those themes continue into the event itself.

30:44 And this is our third year running, uh, Data + Tacos, and it really is just an excellent opportunity to meet and connect with some of the Tableau community, um, in a more laid back and relaxed environment than Tableau Conference. I actually met Roshni, um, at last year's Data + Tacos, and I don't think I ever would in, in Tableau Conference itself.

31:04 So, um, it's a really great way to kick off, uh, your conference week. That's the QR code, and I put the link in the chat as well to our user group if you need more details. Um, and if you want to get registered, we'd love to see you there.

31:20 Whilst we're on the topic of Tableau Conference, I am speaking at this year's Tableau Conference. This is my first time presenting, and I'm really, really excited about it. And if you have room in your schedule and you'd like to hear me talk a little bit about my first Tableau love, which is dynamic zone visibility, in my session that I've called DZV is Easy, uh, please come and check that out.

31:43 A quick little… quick couple of fun facts about me. I'm originally from Scotland, but I've lived in the States for a good long while now, uh, mostly in landlocked locations. But now that I live in San Diego and I have some water nearby, I have been learning to sail, which has been great. And as already mentioned, I'm a huge fan of coffee. I like to learn about where it's…

32:03 comes from, how it's grown and processed. And to sort of double down on both my data nerd and coffee nerd credentials, I've included a picture of my espresso machine there, which gives me real time data visualization for some of the key measures that I'm most interested in when I'm pulling a shot of espresso.

32:22 And I, like many of you, I'm just a huge fan of Tableau. I love building dashboards, sharing my work on Tableau Public, and talking about the things that I've learned along the way. And if you look at my Tableau Public profile, you'll definitely see a theme developing recently where I build a lot of things with map layers, but very rarely is that thing an actual map, which brings us to today's presentation about making it with map layers, where we borrow some of the features of map layers and try and apply them in different kinds of builds.

32:56 But if you're not familiar with map layers, it's just a feature in Tableau that allows us to combine multiple maps into one single view.

33:05 Uh, for instance, I'm planning my Tableau Conference this year, and so I have this map on the left, which has two key data points on it: the location of Tableau Conference and the location of Data + Tacos.

33:17 And n- with these two points, um, in place, I can start to plan around them. On the right, I have a map of all of the available Airbnb listings in the area, and map layers allows me to combine those into one single view.

33:36 I don't wanna commit to any Airbnb without knowing the coffee shop situation, so I have another map with coffee shops in San Diego, and I can layer that into my TC planning map.

33:48 And then just for fun, um, I've drawn a map, a point on the map out in San Diego Bay, which is a really great spot for learning how to sail, and I'm gonna throw that onto my map as well.

34:00 And along the way, a bunch of cool stuff happened.

34:04 Some of those data points came from different data sources, and I did not join, blend, or relate them in any way. Map layers allows us to combine different data sources all onto the same view, something that you cannot do with other visualizations.

34:20 Next up, whenever I added a new layer onto my map, I got a brand new marks card, which allowed me to control things like mark type, size, color, font, labels, all of those things independently, and give me really granular control over every layer of the map.

34:37 And then finally, and perhaps most importantly for this presentation, is this ability with map layers where you can just pick a point on the map and draw something there manually, uh, that you have determined rather than relying on Tableau to tell you where that point needs to go.

34:54 Now, what's perhaps very obvious from this demonstration, and I did not know Dennis was gonna be on this call when I put it together, so I'm a little embarrassed for having just presented it, but I'm not a maps guy. I don't build maps. I don't have… There's no call for me to do it day to day, but I do love, love, love this idea of building up a visualization one layer at a time and having that really granular control.

35:21 So what I wanna do is borrow some of this stuff and apply it in a different way.

35:27 So let's start with a single point. Uh, first of all, that ability to draw a point on the map, and in order to draw a point, I need a way to draw that point, and I need a coordinate system. And so we borrow that from map layers.

35:43 There is a function called Make Point, and the coordinate system is latitudes and longitudes.

35:49 Now, personally, I don't really like thinking in lats and longs. I'm a much more X and Y kind of person. Um, I'm sure many of you in education and math are much the same. Um, but fortunately, the transition or the translation is really, really easy. So if I want to use X and Y, I just have to remember to feed the Y coordinate to the Make Point function first, and then the X coordinate second.

36:14 So for instance, this point here on the grid, I would call that point one, two, and if I want to draw it in Tableau, I would write a calculated field that is the function Make Point two comma one.

36:32 All right. Let's jump into a live demo today. I'm going to recreate a couple of chart types and… in map layers and see if there's ways that map layers can either, um…

36:43 First of all, can I build it in map layers? And if there's anything more or I can enhance the view in a particular way.

36:52 All right.

36:54 So whenever I've not used map layers for a little while, I, um, always come back to this little exercise that I like, and it might seem a little bit silly, but it's very effective at kind of rewiring your brain to think of how, uh, Tableau thinks. Um, and because it's all about creating a point, I'm just gonna draw some points. I kind of think of this as the equivalent of scribbling on a blank piece of paper or the Tableau version of doodling.

37:23 So I have a couple of points already written out.

37:28 You see this a lot, Make Point zero, zero, and I have a second point, which is just gonna be Make Point one, one.

37:36 And so let's bring them onto the view. So I bring the first one on by double-clicking, and I get a point on a map at latitude zero, longitude zero.

37:44 And then to bring another layer on, I'm gonna click and drag point one onto the co… um, onto the maps until you get this orange box that says, "Add a marks layer." And as I do, you see that I get two points on the map, and it has zoomed right in, but, um, it is still a map if I zoom out.

38:06 And we got those new marks cards, uh, one here. And to sort of demonstrate that they are separate points and separate layers, you can turn on and off map layers using this eyeball icon just to the left. So that one kind of turns on and off, and that one turns on and off, just highlighting that they're separate points.

38:24 And because they're independent, I'm gonna start by perhaps making this a big blue circle, something like that, and then making the other point, I don't know, a big green square.

38:43 And as soon as I have two points on the map, I come up here and turn the maps off.

38:51 And this is always my preference. I only borrow the map for a couple of minutes, and s- I always try to get to this grid view. Uh, the reason for that, I like the gridlines. It allows me to much more easily eyeball where my next point is gonna go. Um, and you don't have to worry about things like curvature of the Earth or stuff like that.

39:12 Now, if you are, um, a person who likes organized workbooks, you have all of your calculated fields in, um, folders and all of your tabs color-coordinated. Well, first of all, you're my people, so, uh, hello. Um, and, um, you'll really like this feature. Um, you can rename y- each of these, um, marks cards just by double-clicking on it. So that was my circle, and that was my square.

39:39 And that just makes your workbook much more easily readable and interpretable by someone else coming along.

39:45 I wanna draw a few more points on here, but I don't wanna add a bunch of points to my data pane. That's just gonna grow my data source, um, a lot. So, uh, this is a cool technique where you just bring up a calculated, uh, field window, and let's put something like maybe Make Point two, two in there. And rather than saving this, I'm just gonna highlight it and drag it onto the view.

40:09 Now I'm able to create a new point just directly from doing it like that. And then let's maybe make another one at three, three.

40:21 So two more points made very, very quickly. And then maybe my number two point is

40:27 a triangle.

40:33 And you can… You sort of get the idea there, um, where each point is coming on independently. Each one is getting a marks card. Everything is a little bit different. So one more, and that'll show you. I'm just gonna make this a circle for now, showing you this point up at three, three. If I want, um, I can change this out to be a bar.

40:54 Uh, and what it does, it fills the space between the point and the zero line, and that… It can be very powerful, um, when you're building, uh, things using map layers.

41:04 But let's jump into actual examples right now. So bands. This is the simplest example of a chart type in Tableau, and this is how I would probably build it just regularly out of the box.

41:16 And to recreate this in map layers, um, it should be pretty straightforward. It's just a single point with some text on it. So let's move over to this. I have a point for my band.

41:30 Now, we're only using one point, so it's fairly arbitrary, but I use this point sort of halfway between one and zero. Um, but you can use any point that you like. I'll double-click and add that to the view. And immediately, I'm gonna contradict myself. I just said we need one point, but I wanna turn the maps off. So I'm gonna bring on a second point, and that point's just gonna be there as a placeholder.

41:51 I'm probably not gonna use it in the end, but it's just gonna allow me to go in and turn my maps off.

42:00 Now, if you turn your maps off before you've got two points on, uh, the map layers things just stops working.

42:07 So focusing on my band, that needs to be a text.

42:12 And let's bring on some information. Uh, so I want the metric name, probably the rate, and perhaps the month over month change.

42:21 Drop that onto text and then a little bit of editing.

42:35 We have a pretty good looking band there.

42:38 And now we can bring on perhaps just a few more of them.

42:47 Coming together quite nicely. Now, one of the features of building bands, uh, regularly, uh, I don't love, is if you click on it, you get the big blue highlight box. And I know that there are tips, te- tips, tricks, and techniques to prevent that from happening, and you can apply those same tips, tricks, and techniques in map layers because, you know, we get that same blue behavior.

43:08 But you don't have to. When you use map layers, it's very easy to disable selection on any layer of your visualization just by using this dropdown menu here and clicking on Disable Selection. And honestly, of all the cool things of map layers, that tiny one selection is perhaps my favorite. And what's really useful is you can perhaps have one layer disabled and another part of your visualization enabled, allowing your users to only select on certain things, which is not something that's possible in regular building out of Tableau because it's kind of an all or nothing.

43:41 So I have this m- band. Looks pretty good. But because it's map layers, I can keep building on top of this. If I was doing this regularly, I'm kind of done. I can't go anywhere else. But I want to add maybe a visual element that can… maybe an icon that re- is related to the metric. So I have this other point called, uh, Point Icon. I'm just gonna bring that onto the view and change it out to a shape just so that we can see it.

44:09 And, you know, it had a point. It's just up into the top right. And Tableau has moved things around a little bit. It's trying to zoom in to optimize the space, and sometimes that's good, sometimes it's not. I think in this instance, I don't want it to do the zoom, so you just come in and lock your axis in the same way that you might do for any other build.

44:31 So I think I want this one to run between minus one and two.

44:40 And I could change the X-axis. I'm just gonna leave it as it is just now. And now that I have that shape there on my point icon, let's get something that's a little bit more relevant, and I could even change the color there as well. So starting to look like a pretty good looking band, and this is probably something else that I would disable selection on.

45:03 But you might not want just this, uh, informational graphic. You might want something a little bit more practical. So with a little bit of cleanup, your report might, uh… your band might look like this. Um, and instead of the icon, I'm gonna swap it out for an trend arrow. Like, is my trend a good trend, a favorable trend, or a less than favorable trend?

45:24 Or you could do something like a help icon, and then, and this would be a good scenario where you make this one selectable and your actual band not selectable, and you could add a hover over with extra information about the metric or a click in and use some sort of dynamic zone visibility to allow your user to interact with this in some way.

45:46 Or, you know, it's map layers. You could parameterize all of that and just give that over to your user or give yourself options to your stakeholders.

45:58 So we've taken a band, uh, built regularly and enhanced it visually with map layers and kind of functionally as well, uh, either, uh, with this help i- uh, the metric icon, informationally with this direction, or with this kind of help icon.

46:19 Moving along to my next example is KPI cards. This is something I… I love building KPI cards, and they're, uh, one of my favorite things to do. But dependent on how you're laying that KPI card out, you can start to rack up worksheets really, really quickly. So for instance, this hospitalization KPI card is two worksheets. It's got one that's a band and one that's kind of like this trend, and you smush them together to look like, um, one individual view.

46:47 And I'm only using six metrics here, so that means 12 worksheets, but that could very easily explode to twenty, thirty, forty worksheets all in one view. So let's see if map layers can help me out here.

47:02 So my starting point…

47:04 Well, first of all, what I'm driving towards is something that looks like this all within one single view. And my starting point, we've already built it. The band was our starting point for this. So not gonna recreate that. And in fact, for the rest of this demo, I'm just gonna hide the band because it's done. We'll turn it back on when we're ready for it.

47:25 So the next layer that I want to add would be perhaps the line representing my current year performance.

47:33 So I have a calculation for my line chart.

47:38 And this time you'll notice it's just a little bit different. When I wrote my Make Point calculation before, I picked the specific point and told Tableau to draw something there, whereas here I'm actually feeding it some of my data. And so the Y coordinate, remembering that the Y goes first, is my metric rate. SY means the selected year, um, and then the month of my period date is the X coordinate, and that is the integer version of the month.

48:04 So one for January, two for February, et cetera.

48:08 Now, I am dividing that by twelve just to kind of make sure everything sits between zero and one on the scale. Um, kind of a shorthand normalization, but not really.

48:20 And if I bring that onto the view, first of all, you just get this squished up mess. I'm gonna change it to a circle so that you can see what's going on here. And it aggregates all of that circle together.

48:33 And because I'm using my actual data, I need to give my marks card just a little bit of context, and that context is that month of the period date. So if I drop that onto detail- Tableau draws not one point, but several points. Um, and you can start to see it taking shape here. I can tell Tableau now to connect those dots and turn it into a line by using a line chart.

48:58 Now, Tableau is smart. It knows that month is on the detail shelf, and that it should join it in a particular order. But if you aren't using month or you're us- you want to be really prescriptive, you can make sure that month is on the path shelf. Let's turn, make that blue, and I like these little dots to be in my location.

49:20 Next up, I have another calculation for my area chart, and this is going to be sort of in the background. This is last year's performance that I'm comparing it to, and it's a really similar calculation. The only difference being I'm using my PY version of the metric rate previous year.

49:39 And I just drag that on as a layer.

49:42 And then once again, I have to give it that context of the month of the period date.

49:47 Let's make it a circle first of all, so you can see what's going on. Now, couple of choices here. You could use it as a bar, and it's gonna fill up the space between the point and the zero line.

50:00 And whilst we're here, this is an easier way to see it. You'll notice that the bars are sitting in front of the line, whereas really I'd want the line to be in front of the bars, and that's because the order of your marks card really matters here. It's kind of like the floating order in PowerPoint. Uh, so if you want something to be to the front, you want it to be higher in the marks cards.

50:20 So a couple of ways you can do that. You can either just click on this dropdown menu and move things up or down, and because it's Tableau, lots of drag and drops. You can also just drag those things around. And as I'm dragging, I can't but help rename this be a line

50:38 and rename that.

50:40 Let me call it area because I'm about to swap it out for an area chart which behaves very similarly where it fills the space between the point and zero.

50:51 So maybe make that blue, something like tw-

50:56 So you get the idea. I'm gonna bring my band back.

50:59 So we have recreated that view that I was driving towards from the dashboard. But because it's map layers, we can just keep going. So for instance, I might add this little indicator that tells me if, if the trend is going up or down and if that's favorable. But because that should be in front of the line, I'm gonna drag it up to the top. And I see my, my hospitalization rate has gone up, and that's a bad thing.

51:25 Um, map layers can just keep going. Maybe a goal line. Maybe this is something that you're interested in. And already I'm into the space where I can have all of these things just building out regularly.

51:36 And then I built out this custom axis. This was something I stole from, uh, Prasann who does a lot on map layers. Um, I really, really like it, and you can do whatever you want with it.

51:47 Now because this is all in one worksheet, there's a few things that are now available to you. So first of all, I've managed to take these two worksheets and turn them into one worksheet. So I've gone from 12 worksheets to six. But, um, I think we can do better.

52:06 So because this is all one worksheet, you could use… You… A couple of options here. Calculated field, small multiples, or I just brought in like a small dataset that told me, uh, which row and column to put all of my metrics in. And then when I bring that onto the view,

52:22 it looks a s- little something like this. So one single view returning six different KPI cards.

52:31 If I go back to my dashboard, this is the view with the 12 worksheets, and this is really interesting. This is a nice by-product. It's not the reason to do it necessarily. But as I change the filters on this, notice Tableau just takes a second to render everything, um, because it's rendering 12 worksheets. It's a lot of work you're asking it to do.

52:53 But if I swipe o- switch over to the map layers and do the exact same thing, Tableau's only having to render one time, and notice it's almost instant.

53:03 So to recap, we started building out a band, and we enhanced that using map layers.

53:11 Um, we then took those bands and turned them into KPIs and then put it all onto one worksheet and along the way increased the, um, performance of the report by putting it into map layers. And then one thing we didn't touch on today for, uh, kinda time purposes, this is something you could do with tables as well. Tables are one of those things we'll never escape, um, as developers, and Tableau doesn't really excel when it comes to tables.

53:40 And this is an example table that I would probably build in Tableau, um, and rebuilding it in map layers, you could do something… You could get to the exact same stage, and then you could take it a little bit further where you can add more than one point into each cell, and you could decide what's gonna go in that point. You might have this symbol, or you could do an arrow.

54:06 Totally independent. You can do whatever you like.

54:09 And you could take it even further and just put everything in there. And honestly, I built this table, um, as a sort of joke build to like… almost like taking a feature too far. But then I realized this is exactly the table that my stakeholders love. So if we can provide it for them, then all the better.

54:28 But what I would probably do, because map layers is so versatile and flexible, you, you're no longer constrained to some of the standard table functionality in Tableau. So you could just redesign this altogether to build something that looks a little bit more like this.

54:45 But not enough time for that today. Um, that is my entire, um- A demo for you, but just once again, um, I'm Michael McCusker.

54:58 I'm talking at Tableau Conference, and, uh, Data + Tacos. Really excited about that. Our, our user group is very excited to have you come and join us and host, and host you there, and kick off your conference week. So thanks for your time.

55:12 Roshni Gohil: Michael, that was amazing, and I know you said you didn't have any egg puns, but ya lied. Uh, I heard eggcel, uh, talking about flexibble, uh, tables, and all sorts of fun stuff. So, um, that was amazing. You gave me some really great ideas for some viz that I need to build soon. Um, so, so yeah, thank you. Um, and I'm sure a whole bunch of us are gonna come check out both Data + Tacos, the link is somewhere in the chat here, and your session at TC.

55:42 Link is also somewhere in the chat here.

55:44 But up next, we have our brand-new HETUG co-lead, Andrew Drinkwater. He's back to talk about data. Oh, oh wait, hold on. He's not coming back to co-lead?

55:54 All right, fine. Co-lead emeritus Andrew Drinkwater. Professor Drinkwater. He's back to talk about data relationships.

56:02 Oh, wait. The relationships guy is not here to talk about re- Tableau relationships?

56:07 Oh, he's here to talk about open source data lineage tools that'll help your governance go from scrambled to absolutely eggcellent.

56:18 Eggshell-ent? Uh, Professor Drinkwater. It's a clutch.

56:23 Take it away, Andrew.

56:25 Andrew Drinkwater: Thank you for the eggcellent introduction, Roshni. I appreciate it. Um, and thanks for bringing me back. It's always fun to be here.

56:33 Um, so today I'm going to not present about relationships, uh, because I became known as the relationship guy for a while. But indirectly, we're actually gonna see something about how relationships work, just, uh, a little differently than what we might in, in Tableau.

56:47 Okay. So, uh, let me share my screen here, and we will get started.

57:01 Okay. So, um, this is a presentation focused a little bit on data governance, primarily data lineage and how that interacts with Tableau. So, uh, we'll talk about a few things, and we'll spend, uh, some more of our time together on a live demo and see how well I do live. Okay. So, um, dashboards sometimes have a bit of a chicken and egg problem. Um, and often we find we trust the dashboard because we trust the data, but then when we ask ourself why we trust the data, it's because it came from a dashboard that looked nice.

57:32 Uh, so this sometimes is a little bit circular, um, but we can still have some fun with it. Okay. So, um, moving onwards. Uh, one of the things that happens a lot in my world is we get a fairly simple question, or at least what seems simple, like how many students are enrolled this semester?

57:50 And everybody has a little bit different answer. Uh, it turns out nobody's really wrong, but, um, they often are using different data or filters or time periods, et cetera. Uh, and so what we wanna do is help make, uh, this a little bit more transparent for everybody involved.

58:10 Okay. So, um, data governance as a discipline sometimes struggles in higher ed because a lot of tools are built for major enterprises, not really, uh, folks starting out. Um, they are often tied to ambitious rollouts, uh, whereas we find pilots are often a little bit better fit. Um, and sometimes generic tools can miss the nuances of the, um, uh, s- sorry, uh, enterprise resource planning systems or student information systems that we, uh, know and love and sometimes, uh, curse at.

58:38 Um, so w- we wanted to take a bit of a different tack to that. So that's sort of the background on this.

58:43 Um, Tableau has some elements of data governance built in. So there's, there's great descriptions visible, uh, through Tableau Server or Cloud. Um, you can figure out certification, things like that. Um, and if you have the Tab- Tableau Catalog, uh, part of it, you can also get a pretty good sense of lineage, at least within the analytics layer. So anything that is within Tableau and anything that it reads directly from, uh, is pretty well done there.

59:06 But sometimes you need to go a little further back. And so that's really the motivation for the work that we were doing with the open source tools here, uh, was to, to help, uh, some of the colleges that we work with, um, move towards a more active, or, sorry, more interactive and living form of documentation so they could understand how information moved between, say, Banner as a student information system, uh, through a data warehouse, and eventually through to Tableau, how, how all those pieces fit together.

59:33 So I'll show you that soon.

59:36 Um, there's a couple of different open source options that I'm aware of and have looked into over time. I suspect there's probably more than this, but these ones are both great. Um, so we, we ended up using, uh, a tool called DataHub, which comes from datahubproject.io. It, uh, it's been rebranded ever so slightly, uh, in terms of URLs and things like that.

59:56 Um, but it's a metadata platform that helps with data discovery, lineage, and governance. Uh, there's another one called OpenMetadata. I haven't used it a lot, but it looks excellent as well. So if you're considering going this route, they're both worth a look, uh, and there might be some others out there as well.

1:00:11 So, um, what we found is that often when it comes to things like governance and definitions and, and the like, um, y- you often start with, like, text documents, and they tend to languid- languish over time. Uh, and nobody tends to read them, and so that becomes a bit of a problem. Uh, a live model makes it a little bit easier. So, uh, metadata, not actual data, is ingested automatically.

1:00:34 Um, lineage is connected, which you'll see very soon, and it's intended to be more collaborative, uh, than, than some traditional ways of doing documentation.

1:00:41 All right. So what we're gonna go through today is we're gonna try and start at a Tableau dashboard, and we're gonna work our way backwards, uh, to the systems that are underneath that and making it possible.

1:00:53 And we're gonna attempt to follow, uh, a definition for the term registered, uh, going back to our example of how many students do we have registered. Uh, so we're gonna work our way from right to left here, uh, Tableau workbook back through a data warehouse and, uh, through to the student information system. Uh, and there's a few layers in between, but I don't wanna overcomplicate this quite yet.

1:01:11 Um, so what we s- often find is that we have a, a simplified notion of what happens in terms of data flow and lineage. Um, but actually it can often be much more complex than that, uh, particularly once we start extracting spreadsheets and making local databases and emailing them around. Uh, so with that, let's take a look at our live demo.

1:01:37 Okay. So, um, here we start with a very basic example Tableau dashboard. Uh, so this is about course enrollment analytics. It's really just, like, how many students are enrolled in different terms over time, um, and, uh, some high enrollment courses shown down the bottom. So nothing, nothing super complicated here. Uh, I know all of you are very capable of making fancier Tableau dashboards than this, and I suppose I am too.

1:02:02 Um, but the core point of this is to try and understand how did this information actually get here in the first place. Uh, so let's try and dig into that. Uh, so I've got a link or a button here to go back to my data definition, so we're gonna start by clicking that. And what I have here is, uh, the first look at the tool that, um, that from an open source perspective is called, um, called DataHub, and from our perspective at Plaid is we, we have made some improvements to this, and so we call it Plaid Govern.

1:02:32 Um, but it's powered by DataHub. So, uh, that's the gist of it. So when we first open it, uh, we get a view here that gives us a summary, and we've got some documentation that we created for this. It tells us it's a, it's a view of Plaid University's fall enrollment, um, and what the purposes of it are. And it gets into kind of what the key dashboard is on here and some key metrics and so on.

1:02:52 This is all plain text, so we just, we just made this for the purposes of that. But it gets much more interesting once we start to look at more components that are around it. So, uh, first one is, you know, documentation is pretty much as you need it. It's a, it's a text editor, uh, or a rich text editor, sorry, and you can modify it as you see fit.

1:03:12 So w-we can say today's audience is the HETUG, and they are egg-cellent, something like that. Uh, and then we'll publish that, and so now everybody who has access to this can see that that was a change. Um, my favorite part though is lineage. So I'm a visual person, hence my love of Tableau and maps. Um, and lineage, I think, is where things start to get really powerful.

1:03:39 So we're gonna work our way from right to left here. So our, our demo workbook in Tableau has some associated documentation that we can see here on the right-hand side. Uh, it looks like I forgot to make an owner for this, so I'm gonna hit the plus button, and I'm gonna say the IR office owns this from a data governance perspective. There's different ownership types that we can have, and we can also customize this list.

1:04:01 Uh, so for sake of argument, we're going to leave them as a default technical owner, but we could add other roles. And we could also say that, uh, maybe our VP Academic is definitely not a technical owner, but they are the data governor. So if there's ever a major issue or conflict between terms, um, uh, this role might get involved. So we'll add that as well.

1:04:22 And… Oops.

1:04:24 Uh, so now we've got some owners listed here. I've already pre-tagged this as being in the student domain. We might have other domains like human resources or finance. We might have sub-domains as well, so student is a pretty broad category, for example.

1:04:35 Um, and I haven't tied any glossary terms to this yet, uh, but we'll see those later. So now working our way from right to left, we've got our dashboard, uh, as our starting point, or sorry, our ending point, I guess would be the way to phrase that. And then within our dashboard, we have three different charts, and these all have appropriate documentation as well.

1:04:55 Um, and then we're gonna go further back. So further back than our charts, we have a embedded data source in our Tableau workbook, which is on Tableau Server, effectively.

1:05:07 Um, the embedded data source is the one that you see when you're designing dashboards that connects to a data source that's been shared through Tableau Server or through Tableau Cloud. There's also what's called a published data source one layer deeper than that. They're usually almost identical, but the only difference between them is typically that the embedded data source might have, uh, calculated f- excuse me, calculated fields or similar, um, that are unique to that workbook as opposed to in the published and shared data source.

1:05:35 So they'll look almost the same from a documentation perspective, but there's a subtle difference there.

1:05:41 So we now have our, um, published data source called Banner Course Enrollments, and we can get into more detail of these if we need. So we can see there's a hundred and seventy-seven columns in this data source, so it's, it's big. Um, but we can find ones that are of interest. So, uh, for example, I could look at registered, um, and find it there.

1:06:01 But we're gonna move a little further back now. So the part that gets interesting here is our Tableau Server or Cloud connects to a view in our data warehouse, and this is an Oracle view, but any other database technology would look basically the same.

1:06:16 Um, and then we can start working further back. And this is where having additional tools beyond what's built into Tableau Catalog can be really helpful, uh, 'cause often it would only see that database view that I've highlighted kind of there in the middle.

1:06:29 Um, so let's, let's start working our way back and figure out where all this came from, and then we'll take a couple stops along the way to show a few things that are, that are helpful. Um, so we have a database view, which essentially pre– uh, sorry, pre-joins a variety of data within our data warehouse, and it makes an analytics-ready view. And the purpose of this really is to make our lives easier in Tableau so we don't have to make a whole bunch of joins in each data source that we make.

1:06:55 Um- When we go a layer earlier, I look at fact course section here. And so this is, uh, the de– the measure related side of our course section information. So it does things like, uh, count up how many students are registered in a particular section in a particular term. Uh, so it's our primary aggregation from a metrics perspective. Its close cousin is course section, which looks at dimensional values.

1:07:20 Uh, so what is the course and, uh, things that are about it, for example. But let's go a little further back from fact course section. So how does that get populated? Um, it's a small logo here, but in the middle, we have a workflow, um, using a tool called FME. But you might have, uh, other tools like, um, one of the Microsoft ETL tools or Informatica or something similar.

1:07:41 Um, all of those would, would sort of show up similarly to this in that they transform the data at different steps. So let's go a little further back, and we're gonna go from fact course enrollment, which is at the, um, course sect– Sorry. We're gonna go from fact course section at the course section level to fact course enrollment, which is individual students enrolled in individual course sections, so more detailed.

1:08:04 Um, and then let's try and figure out where does that even come from, 'cause that also looks kind of data warehouse-y to me.

1:08:10 And here we can see that there are ten children of the fact course enrollment. It's showing seven by default. I don't know why it picked seven. It's a lucky number, I guess. Um, so if we need to, we can search for others or we can click show more or show all and show all the children. Conveniently, the table that I'm looking for is SFRSTCR or SFRSTCR if you're a banner speaker.

1:08:33 Um, so this is actually our source for a lot of this information. It contains the registration information from Banner. Uh, of course, if you're a Banner user, you know it's a little more nuanced than that. But for demo purposes, let's pretend everything lives here.

1:08:47 Um, and we have some documentation related to that. It tells us, uh, what, what the data grain is. Uh, so student and CRN, which is a way of identifying course sections, uh, and term.

1:08:59 Uh, so this is, this is our trace all the way back. Uh, so we can do a few other things here though. So I've shown you kind of how the lineage works. So now let's actually take a look at some more detail.

1:09:09 So if we look at this, um, uh, SFRSTCR table, we can see that it is pulled in from our, our database, all of the different fields. Although I'll admit, I did not put descriptions in for every one of these banner fields, uh, for today. But it, it pulls this all in from our database. And if we already had descriptions in the database, it would pull that in as well.

1:09:31 Um, we have table level documentation similar to what we saw with the Tableau stuff, and then we can get back to lineage where it starts with this, and we're gonna work our way forward.

1:09:41 Um, so I'll look at fact course enrollment, where we've done a little better job of documenting things for us. So we can see here that we have a number of different definitions. It knows what a foreign key in our database is. Um, it tells us where a lot of different information comes from. Some of these are populated by Excel spreadsheets through our data warehouse, uh, and so on.

1:10:00 But our key term of interest was registered, so I'll, I'll pick on that one. Um, so we have, uh, another chicken and egg example here. So we can see that, uh, w-we're using both the SSB sect, the course section information, and the SFRSTCR or registration information, uh, within this field, uh, in order to figure out the answer to the question of which came first, the chicken or the egg.

1:10:26 Um, our real definition is actually below here. So I'm gonna remove my April Fools prank so that the next time I give a webinar tomorrow, um, I don't run into quite the same thing, and then I'll update my definition, and now it's been updated in our database, and so it's live.

1:10:41 Uh, so that takes us through some of the basics of that. Oh, I apologize. I missed one step with registration. Um, so the other thing is we can have a glossary term. So if we have the field registered in a whole bunch of places, we can have kind of a centralized documentation of what that means as well. Uh, so we can build that out, and we can see where that glossary term is used, uh, currently just in our data warehouse, but I would eventually expand this to include Tableau and potentially other places.

1:11:10 Okay. So I've got like three minutes left, and I definitely want to give people a chance to ask questions if they'd like. So let me quickly show you the last couple of pieces, and then we will move from there.

1:11:26 Um, so moving back through our chain here and getting back towards Tableau, uh, one of the other things that I, I ought to have mentioned right at the beginning there is as we, as we go through Tableau, we can also link this the other way. So if you're starting in the data governance world and you wanna see what things look like in Tableau, uh, you can click the button, and it'll take you right to the right place.

1:11:46 Now, if you had a functioning Tableau Server today, um, it would link actually to the Tableau Server instead of, uh, my webpage. But, uh, that part's all automatic. The other part that I really enjoy about this is you can ingest data automatically. So you could do it, you know, every day, uh, or you could do it on some kind of reasonable schedule, or you could do it on demand.

1:12:05 Um, so if your, if your data warehouse or other data sources or your dashboards are changing on a regular basis, this will help ke- you keep up with that, and it'll identify places where you haven't got your documentation in order, uh, and that maybe you need to make some improvements. Uh, so that has made our lives a lot easier with this, and, uh, we found it to be really helpful for three key things.

1:12:27 Uh, one is, um, just having documentation period is often an improvement over where a lot of people, people start out. Um, two is, uh, it turns out that this is a really good training tool when you onboard new staff. Uh, so they don't have to ask their colleagues questions that they feel like should be simple. Um, they can just go and, and get the information that they need.

1:12:48 And so it can be really useful for that as well. And then, of course, this helps improve transparency, and it can improve communication between, say, institutional research and IT or other stakeholder groups as well. So we found those to be pretty useful.

1:13:01 Um, so with that, um- I think I've covered a whole bunch of this stuff. Um, the biggest thing here is that, uh, you'll gain time back because you're not chasing definitions. They're pulled in automatically from your various systems. It can talk to all the systems at once, and this helps you reduce the risk associated. Um, the other big one happening right now is with AI tools growing in leaps and bounds, and the usage of them going accordingly, um, it's really important to actually have definitions, documentation, lineage, all this stuff available so that when your teams are using these AI tools, um, they're built on a solid foundation rather than, uh, just getting to a wrong answer faster.

1:13:41 Okay. So, um, your data's already moving through different systems, um, and it's good to start centralizing those and having an understanding of where they fit. Uh, lineage can help make this stuff visible. You need it for AI to do really well, and you can start, uh, very quickly with a pilot through the open source community or by talking to us about the same thing.

1:13:59 If you'd like to see this in more detail, I'm actually doing a webinar tomorrow. You're welcome to join. I'll throw the link in the chat shortly. Um, and it will be recorded 'cause I know not many people can, uh, make a webinar on less than 24 hours notice, but we'd love to have you if you can make it. Um, so with that, I will say thank you so much for having me.

1:14:15 This gives a bit of a sense… And I apologize, I could not see the chat window while I was presenting, so I'm happy to answer questions if they came up.

1:14:23 Roshni Gohil: We have a lot of, "Holy crap, this is amazing," type comments and, um, some, some fun, uh, snarks at your spelling.

1:14:31 Um-

1:14:31 Andrew Drinkwater: Ooh, I tried to keep those Americanized and I might have missed some of them. That's great.

1:14:36 Roshni Gohil: Uh, excellent. Excellent only had the one L.

1:14:40 Andrew Drinkwater: I think that's actually a real typo. We spell it the same way in Canada.

1:14:44 Roshni Gohil: Oh.

1:14:44 You never know with these Ls and whatnot. Um, but that was-

1:14:48 Andrew Drinkwater: That's true

1:14:48 Roshni Gohil: … amazing, so thank you so much. Um, I don't think we had any main questions. A lot of, "I need to watch this re- uh, the, the recording." Um, I… We did put the link to tomorrow's longer webinar on modern data governance. The s- some of this will be sort of a slight refresher for folks who were able to, to get through today's session. Um, but yeah, and we're totally looking forward to it.

1:15:10 I will not be watching live, but I will be catching the recording because work meetings.

1:15:17 But thank you so much for coming back, Andrew. Um, and I'd like to, I'd like to let you know because it is April Fools-ish, um, I had a cup of water on, uh, my desk at some point. I was like, "Holy crap," and I knocked it over, so my pants are now drinking water.

1:15:33 Andrew Drinkwater: Awesome.

1:15:33 Roshni Gohil: And you're welcome.

1:15:34 Andrew Drinkwater: Love it.

1:15:35 Roshni Gohil: So with that-

1:15:36 Andrew Drinkwater: Well done

1:15:36 Roshni Gohil: … thank you all so much. Thank you, uh, Andrew. Um, that was great. And I don't know if there are any more questions in the chat. Uh, take a look. I don't think there are, but we'll see. Um, in the meantime, we do wanna know what you thought of today's meeting, so tell us what you liked, tell us what you didn't like, how you like your eggs cooked, which chicken you'd match, uh, what your, your favorite banner table is, uh, or BAN- banner table abbreviation perhaps, um, and how you pronounce it.

1:16:06 All of those things we wanna know. Uh, as we mentioned earlier, uh, at our official April HETUG, 'cause remember, this is March 32nd, we're doing our March TUG. Um, but at, at the April one on the 29th, Ginny's gonna be presenting, uh, about, uh, her t- Iron Viz top 10 entry with eggscavating the viz, uh, uncovering DataFam artifacts. Um, but we've got some empty nest boxes in our April meeting, so if you're interested in sharing your own data journey or presenting some tips and tricks in between it, please let us know.

1:16:39 Um, or if you wanna present later in the year, we've got, uh, some, some openings in May and June, and we have yet to set up our July through December schedule, but we will soon. Uh, so feel free to chirp or cluck in our ears, uh, by using the, the links we just put in the chat.

1:16:54 Um, and I'm gonna pass this over to Sam.

1:17:00 Sam Homier: Yes. Hi. Sorry, I'm with my puppy Rooster, who is very energetic right now. Um, but he's going to announce some conferences coming up soon.

1:17:10 So there's still time to register for the Research Analytics Summit in Rhode Island at the end of this month. Um, also at the end of this month is the Higher Education, uh, Data World Conference, and then Analytics Day on June 11th. That is free again, so make sure you register for that. And then of course the Tableau Conference, and again, there's that, uh, code if you are in the industry sector.

1:17:35 And then, yeah, that ends our meeting. So thank you so much everybody for coming, and thank you to our speakers for being here, Michael and Andrew. And I hope you guys have fun today.

Leave a comment

Your email address will not be published. Required fields are marked *