Tableau Conference’ish Devs at Desks Livestream Watchalong 2020

0:00I might not accept this invite into the

0:05stream.

0:07Hi, Marcus. How are you doing?

0:14Where are you from? Whereabouts are you

0:18joining us from?

0:27There he is.

0:30Hello, hello, hello.

0:32Hey, okay, cool. Let's switch you on. Let's

0:36switch you back on. There you go.

0:38And let's chop myself out of the view. Let

0:42's see.

0:43Are we live? We are live. Of course we're

0:45live. Yeah.

0:47Hello, hello, everyone. Welcome to second

0:50oneness. Apologies for my lateness.

0:52I was just finishing off a Tableau Doctor

0:54session.

0:56Well, the problem was solved.

0:58She's always good.

1:00Good. Let me just chop myself out the top

1:03of your video feed. Do some real-time

1:05editing in OBS.

1:07It's never a good look.

1:10Right, Zoom. Let's go crop you a little bit

1:13here. You can hear me, of course, Ravi.

1:15Can you hear the stream? Can you hear the

1:17live stream?

1:19Yes, it's not started yet, I believe.

1:23No, the main live stream. Not the main live

1:24stream. I mean our live stream.

1:26Not yet. I'm just helping it open.

1:29That's fine. Cool.

1:31All right, cool. Let's get some cropping

1:33done.

1:34Excellent. We are live. I can see it now.

1:41I don't think I can see the chat. Here we

1:43go. Chat is now open. Wonderful.

1:46Wonderful, wonderful, wonderful.

1:49How are we all doing? How's everyone's

1:50first day of conference been?

1:52I guess for the US it's the second day.

1:55For me it's the first day. Anyone seeing

1:57any good sessions they'd recommend or

1:59anything that they'd like to give shouts

2:02for?

2:03Bueller, Bueller.

2:08Bueller.

2:09Why am I getting the keynote one?

2:25They've got no pre-show today, which is why

2:29it's a little bit not interesting in the

2:32stream.

2:33Wonderful. Here we go. Live stream is loud

2:36apparently.

2:38We all good? Are we good on a sound check?

2:41I think so. I think so.

2:43Excellent. Are you just doing your hair?

2:46No. I'm looking down at the live stream

2:49just reading through the comments.

2:52Wonderful. Drop something in the chat. Give

2:54us a shout out where you're from and we can

2:56see where everyone's tuning in from.

2:59I think if you were in yesterday we were

3:01from...

3:02Where were you? We're both in Essex.

3:06Essex. Essex Central.

3:08The Essex Central podcast.

3:11South East Central, isn't it?

3:13Yeah, exactly.

3:15You no longer have siege night with you

3:18today.

3:19No, absolutely not.

3:22Just trying to get channel one open. Is it

3:25alive? Are we alive?

3:26Channel one has just got a disclaimer. It's

3:29a Salesforce disclaimer. Sign of things to

3:31come. Forward looking statements and all.

3:35Are you on... How are you looking at this?

3:37Full screen on another screen, which is

3:39what's coming into UBS. OBS.

3:42OBS. Wonderful. So we're coming in at equal

3:46stream, equal volumes before the mainstream

3:48kicks in. That's good to hear, Mark.

3:51Give us a shout on in terms of levelling

3:52when the stream starts. I know a few people

3:55yesterday wanted to listen to what was

3:57being said.

3:58Apologies in advance because we are going

4:00to chat.

4:01You should have pure stream.

4:04If you want to hear the stream, jog on.

4:07In the nicest possible way.

4:10We'll time it. I think the good thing about

4:15this is...

4:16Sorry, are you watching this on YouTube or

4:18are you watching it on the TC website?

4:21I'm watching it on the TC website because I

4:24'm taking the livestream from their embedded

4:27live feed, which is going to be better than

4:30YouTube.

4:31That's my assumption.

4:33And it's probably less lag. I'm just

4:35watching it through the channels you're

4:36supposed to be watching.

4:37Here we go. Cool. I'm with you.

4:39If I just exit, this is just a browser

4:41window that I'm just in.

4:43Then I'm just full screening it on the

4:44screen and that's basically what's coming

4:47through.

4:48Wonderful. I'm there as well. Forward

4:50looking statements. We are away. We are

4:52away.

4:53Yeah, brain dates. Good shout out there

4:55from Mark. I'm a big fan, a big purveyor of

4:58all things brain dates.

5:01I've got 17 lined up over the next three

5:03days, which I'm looking forward to.

5:05I've had two. One was actually about the

5:09death of the COE, as in what's the point of

5:12having a COE and is it worth it?

5:14I think you had probably Elsa in there,

5:16right?

5:17No, actually, no. I've got Elsa in a couple

5:19of sessions tomorrow.

5:21This was with Mina Osgin, of course, who we

5:23both know, a State School alumni.

5:26A guy called Ajay who works at Groupon

5:28based in Dublin.

5:30And someone called Sumit who was based at a

5:32company, based out of New York, but he was

5:34based in India.

5:35Really good discussion about the pros and

5:37cons of Ambien's sense of excellence.

5:40And the other brain date I had today was

5:42just a chat with someone who I'd never

5:45spoken to before, which is great. That's

5:48what I love.

5:50I had two today. Very good about video

5:52making. Really good. Some from Wells Fargo

5:56and then I also had some from an insurance

5:59company as well that was interested in how

6:02they get their users excited.

6:03It's touching on that COE point, but it's

6:05also touching on that user education and

6:08how do you get people amped up about

6:10analytics or more importantly, how do you

6:12teach them how to use a dashboard you've

6:14just spent hours passionately building.

6:17So yeah, it's going to be an interesting

6:19few years in terms of that. I think the

6:22level and standard of which COEs have to

6:24work out is constantly increasing.

6:27Before you just had to get them together on

6:28a lunchtime webinar. Now it's like, you

6:31know, gamification, you need a website, you

6:35need Slack, you need Teams, you need Slack,

6:39what do you call it?

6:41Pushing Tableau visas to left, right and

6:43centre. It's a whole sort of involved thing

6:46now.

6:47Maybe we'll come back to that question from

6:51Lantus. But just before this stream starts,

6:54we've got a couple of minutes left. Maybe

6:56let's talk a bit about what we're expecting

6:58and what are the hopes and dreams?

7:01I'm going to call this Devs on Stage. Devs

7:03on Stage doesn't have the same ring.

7:06But yeah, Devs on Stage or Devs at Desks. I

7:09expect this to be finally, if I don't see

7:13something reinventing the way we do

7:15dashboard layout at this Devs on Stage, I

7:19think I will actually just quit Tableau.

7:22Wow, that is a bold statement.

7:25We cannot go another year without having

7:28improvements to the layout system. We just

7:32cannot have that happen.

7:34We need a better layout system, honestly.

7:40Everyone knows it. All the people spending

7:43hours nudging, you know, floating things

7:45one pixel there and they're publishing it

7:48to public realising it's minus two pixels

7:50off and that's just got to disappear.

7:54We've got Prep and Browser now. So they've

7:56spent some love putting something like that

7:58in the browser. Let's get some of that

8:00browser goodness into desktop and let's

8:02make layout much, much better.

8:05Get rid of layout containers, get rid of

8:06tiles and floating. Give us a canvas. Creat

8:09ivity has no boundaries. So stop making us

8:12putting it in containers.

8:16Yeah, I mean, I was happy with the, we got

8:18collections last night, which I was very

8:21happy about. That whetted my appetite. I

8:25just want to be surprised.

8:27You know, when you watch an Apple keynote

8:29these days, you're waiting for the one more

8:31thing that will never come. I really want

8:33to be surprised. I want to see something

8:36like, hey, that's really cool.

8:40Yeah, that's true. This is an inspiring

8:42event. It's supposed to get us excited

8:43about the product we use for the next year.

8:46That's the real reason. This is their

8:48chance to sort of show us things they can't

8:50show us for a few more months and sort of

8:52wow us. You're absolutely right.

8:55So just reading the comments here real

8:57quick. So Carl's just saying no Tim, no

9:00about your quit tablo now. Mark Edwards

9:02doesn't want spell check. He's just like I

9:05don't care. Just learn to spell. Carl's big

9:10request is prep public.

9:12Yes. Yeah.

9:15I think that'd be an interesting one. It

9:17gets people to use prep a bit more, maybe

9:19brings it into the self serve environment.

9:21I think I'm a big fan of, I really want

9:23public to get a lot more love. And I think

9:25that's coming. I think that is on the way.

9:30So yeah.

9:33I think one of the things when when prep

9:34first came out, I always before I'd heard

9:36anything about it in my mind, I could only

9:38draw parallels with Power BI, which have

9:40this sort of query builder in part of the

9:42project, right? There's one whole data prep

9:45solution inside of Power BI.

9:47And so I always imagined prep as being, you

9:48know, that connection window where you

9:50connect to data and you do joins and blends

9:52. That to me is where I imagined prep to

9:54live. I never thought it would become a

9:57separate product.

9:59Exactly. I thought, no, no, they're gonna

10:01beef up desktop and make it incredible. And

10:03so the separation is actually, you know,

10:06for quite some time, it's thrown me off. So

10:08I'm hoping for that sort of a bit closer

10:10integration. Yeah, exactly. Exactly. Let's

10:13just make them one thing. Prep public is an

10:14opportunity to really sort of showcase that

10:17right. Do your data prep hop straight into

10:19Tableau build a viz boom end to end done,

10:22you know. And I think you go for it. Yeah.

10:27I was just gonna say cause yourselves what

10:28love do I want for Tableau public and I

10:30think my love for Tableau public is to make

10:31it easier to start like give me give me

10:34public in the browser. But let me start

10:36there and like, and let me copy and paste

10:39data and let me connect to web data source.

10:42Like, there's a lot of guys.

10:44There's some easy wins you could do with

10:46public data sources, right? I think Tableau

10:48could probably ask the community, what are

10:50your top 100 data sources that are public

10:53and just have those in Tableau public right

10:55from the get go, right? Climate change,

10:57diversity metrics, World Health

10:59Organization data, just do the hard work

11:01and make those standard data sources in

11:03Tableau public and boom, people can start

11:06telling stories like

11:08that. It can help you create sources. Cool.

11:11We're under was in the chat. Hey,

11:15amazing. Amazing. Nice to see.

11:18So we're starting we've got cock grieve and

11:21Keisha. Okay, good. I'm chatting a bit

11:23about deficit desk. So let's go. Good.

11:26Where is this? I'm looking at the way you

11:29were you seeing this? I can't see this in

11:31the stream.

11:34I'm in the stream. It's just started. I'm

11:36on channel one. Right? Am I not on channel

11:38one? This is 50 seconds for me.

11:41Interesting. He just come on. This is

11:46bizarre. All right. Timmy behind. How can I

11:49be hired? I got here before you.

11:54Okay. Here we go. I'm going to refresh the

12:04page just in case I get a much more up to

12:04date stream. I bet you that's exactly what

12:04will happen. Yeah, of course. Yeah.

12:04This model predicts avocado prices based on

12:09month and cyclical yearly trends. If you

12:12got quite well, yeah, but I know Tableau

12:16can do better. So introducing model

12:20selection, you can specify

12:23other models. Apply and notice the

12:28difference. No spoilers. But what about

12:32prices this coming winter? Well, now

12:34extending our axes and making predictions

12:38for future values is so easy.

12:42Let's open the dates context menu and

12:47choose the new option show future value.

12:54Why that's why we can pick from a list of

12:54pre specified time ranges or select a

12:54scientist and see how smooth this is.

12:57We can extend our axes by one. Of course,

13:09these values can be used in downstream

13:10table counts or export it like any other

13:11data in Tableau for further analysis.

13:14Yeah, it's not what I'm looking at avocado

13:16prices. Inspect the local housing market.

13:19Since my lease is up this coming year, here

13:21's a map of Seattle broken down by zip code,

13:24right colored by average monthly rent.

13:28I want to bring in more information into

13:31this view. But layering multiple location

13:34data is not possible in Tableau. Until

13:38today, introducing map layers, you can now

13:41layer as many geometries as you want in the

13:45same view.

13:48I cannot let's bring it house. This is good

13:50. This is good. This is some of the

13:53properties I'm interested in. See this.

13:56This is like a little state. It's like

13:58feature.

13:59Drop zone. I released my who's the dad that

14:02was a T years ago.

14:05Mark's marks. Yeah, it was the mobile

14:19mobile. Mark type change our color interest

14:19our border. Yeah.

14:20And our size. So this is good. And you know

14:22what, let's also rename our layer so easily

14:24loves it because I know the key is gonna

14:26say, third layer phrase dragon draw. And as

14:35a proper Seattleite, I can't

14:37consider it absolute chaos. So let's bring

14:39in stories, multiple ways we accomplish our

14:42locations into the view. Everything is

14:44chaos. Reasoning, right? So you can always

14:47create chaos. It's and resize. Let's see

14:50how he's using this just like that.

14:53The maps with three layers, the rainy

14:56Seattle skies, the projects, the map layers

15:01easily. It was a project. You can probably

15:05go infinite, for instance, interacting with

15:08marks across layers can be very distracting

15:11.

15:12But with map layers, we can disable

15:14selection with this new option. Okay, now,

15:18this is the key continues to update as I

15:21change and coding directly from the marks

15:25guard. And check this out.

15:29Currently, coffee shop locations are

15:31layered on top of housemarks and hide some

15:34houses near downtown. All right, well, map

15:38layers can fix this too. We cannot reorder

15:41our layers and bring coffee shops down. Now

15:45, watch the visit as I release.

15:49Here we go. Everything is now clearly this

15:58is not just for maps, right? Oh, wait a

16:06minute. So I want to compare individual

16:06prices to zip code averages. This requires

16:06level of detail expressions, also known as

16:09LODs. Since we need to calculate how these

16:12fill at the aggregate and at the row level.

16:16Now we know everyone's think disabling

16:18enabling stuff on the most gorgeous. It's

16:22hard to know where to start with closest

16:24syntax error. Right? The regular flow of

16:27analysis.

16:29Well, no more. Early next year, tableau

16:32writes your LODs for you. Introducing easy

16:36LODs. Check this out. I'll select zip code

16:40and monthly rent. Right click and choose L

16:44OD calculation.

16:47Everyone just wrote an LOD from scratch.

16:52Right now we can of course change our

16:58aggregation. So my question is why did he

17:00say zip code? How does it know what I want

17:02to subtract this from our individual

17:03monthly rent values?

17:05This way, we know exactly what houses are

17:08under or overvalued. Now let's rename our

17:12calc appropriately.

17:15Hey, I mean, look, I think I think this is

17:17great. It takes it is drop it on the house

17:21layer. Right color show option, then look

17:24at that step away being a bunch of marks

17:26near downtown or orange.

17:29I need to check down what you want to put

17:31in. So as your party, what are you looking

17:35at real estate data to figure out your next

17:38virtual home visit, inspecting, or

17:41analyzing regional demographics, you could

17:44probably see them adding options in a menu

17:46to say, exclude and easy LODs open up a

17:50whole

17:51world of possibilities for analytics in

17:56your organization. We cannot wait to see

17:59what you come up with.

18:01Good morning, Data fam. My name is Azar and

18:05I'm a software engineer in tech.

18:10Today, I'm excited to show you some new pre

18:16pping features. Let's get started. Here I

18:22have the popular collection trays around

18:24Seattle, but it doesn't tell me what part

18:26they are located in.

18:27Luckily, I have the part boundaries as a

18:30shape file. I'm within the new spatial

18:34support in Apple.

18:37I think I wrote mark yesterday because of

18:39the tabular music.

18:40Whitney Houston.

18:43With support for spatial data, you can now

18:47connect any spatial data you have and work

18:50with it just like any other data.

18:53Yeah. Oh, the zip file, not even the shape

18:55file. The zip file.

18:56Looking at the spatial data here, I see

18:59that the part boundaries are in the same

19:02place.

19:04I'm wondering what he has to update in the

19:06book. If you haven't checked it out, go

19:08check out Tableau prep up and running.

19:11With spatial data coming in different forms

19:13, it's hard to find their relationship

19:16without doing a complex math.

19:19But, being making it simple for you, by

19:24also bringing spatial calculations to prep.

19:30You can now perform any spatial operation

19:34you want, such as buffer, make line, or

19:39make point.

19:41In this case, I want to map the trays to

19:47the parts.

19:49So, first, I use make point to convert the

19:55trace lat long to point.

19:59This is my point about transition. Let's

20:01call it geometry.

20:01Just have it in the same browser window and

20:02it will be perfect.

20:04Then, I use spatial join.

20:09Jonathan Drummey posted yesterday all the

20:11installers has got from Tableau before.

20:14And the big expansion from when it went 64.

20:18Yeah, that's a vision itself, right?

20:21And there it is. 47 trays and the parts

20:23they are located in.

20:25That's nice. They overlap.

20:28Working with spatial data in Tableau prep.

20:31It's not a map, it's just a…

20:33Next, I want to calculate the trail ratings

20:35.

20:36So, obviously, I'm thinking, like, can I do

20:38this spatial analysis with my total data?

20:41Can I lat long to do the transformation I

20:43usually do in…

20:45I really like to keep my workspace

20:47organized.

20:48That is one Achilles here, right? Lots of

20:50tools assume you're working with geography

20:52when you could just work in space in

20:53general.

20:55I want to be able to hide the already clean

20:57columns without filtering them out from my

20:59data.

21:00Oh, Francois is telling us that visualizing

21:01it on the map will come next.

21:02Now, in Tableau prep, you can.

21:05It's great having, like, something like

21:07that.

21:08I was about to say!

21:10You can also hide fields from…

21:12Francois, you're more welcome. Like,

21:14honestly, do this more often.

21:15Or select multiple columns and hide them

21:19together.

21:21That's so cool!

21:24These hidden fields will show up in the

21:26profile part of the data column.

21:27And you can search for…

21:29That is for that.

21:31If you want to. Beautiful.

21:34Now, I can focus my data cleaning on all

21:37these four columns without the fear of

21:40losing the hidden fields.

21:42Okay. I already finished data cleaning. I

21:46calculated the ratings.

21:48My flow is complete.

21:50But it looks more complex.

21:53I want to simplify it before I share it

21:55with others.

21:57I want to hide some details to bring

21:59attention to the most important part of the

22:02flow.

22:03With group step, I can do exactly that.

22:09I created a group for the spatial mapping.

22:12I'm number one for the rating calculation.

22:17Let me also change the description.

22:21I like the way she does it. I like the way

22:23they demo the features. It's really

22:24important, actually.

22:25When you explain it to customers, it's

22:28better to see how the dev has explained it

22:31to you.

22:32I can still expand the groups to see if

22:34they change.

22:35Yes, exactly. How did you actually design

22:38this to be used versus how am I maligning

22:39it?

22:40Simple and clear my flow loops.

22:42Now that I have two groups in place.

22:45And that's data preparation in my flow prep

22:49.

22:50Simple, fast, and flexible.

22:54Hi, I'm Ryan Atala and I lead the Ask Data

22:59team at Tableau.

23:02We've been working on some incredible

23:04improvements.

23:06We're calling Ask Data 2.0, our biggest

23:10update since we originally launched Ask

23:12Data.

23:13I'm going to show you an early preview of

23:16the fully redesigned user interface.

23:19Which combines the speed and flexibility of

23:21search with the power and intelligence of

23:23natural language processing.

23:25Here's how it works.

23:28As you type, Ask Data gives you search

23:30results word by word.

23:32Giving you instant feedback about what's in

23:34your data and what Ask Data can do.

23:36This is cool. I've seen a rough demo of

23:38this.

23:38Ask Data will automatically choose the best

23:40interpretation of each word in your query.

23:42Or, if you prefer, you can tell Ask Data

23:45exactly what you mean by choosing a result

23:46from the dropdown.

23:48Because I think that autocomplete to show

23:50you as you type is really cool.

23:51Ask Data learns from your selections to

23:53improve the quality of the experience and

23:55save you time as you use the feature.

23:57Here's that same query after some questions

24:00about San Jose.

24:02Note, this is the most popular result of

24:04the previous couple of months.

24:06This is particularly useful in cases where

24:08you have the same value of carrying in

24:10multiple places in the data.

24:13Let's say I'm interested in all the results

24:16from Kelly.

24:18Of course, there are multiple Kelly's in my

24:20data.

24:21I'm interested in Kelly Williams.

24:24But I see that she's both an employee and,

24:27if I look at the bottom of the list, a

24:28customer.

24:29The new Ask Data UI makes it easy to see

24:31that and choose exactly the one that I want

24:33.

24:34Transparency is key to effective analytics.

24:38And the new Ask Data UI makes it easy to

24:40see exactly how your query is being

24:41interpreted.

24:43Here's how.

24:45As you type, Ask Data emphasizes the words

24:50that it understands and grays out the ones

24:54you've done.

24:56And what's more, if you hover over an

24:58understood word, Ask Data highlights the

25:01corresponding analytical operation,

25:03showing you exactly the impact that it has

25:05on your result.

25:07So, if you ask a question Ask Data doesn't

25:09quite know how to understand, you can see

25:11exactly why and fix it.

25:13Take this example.

25:15Which industry has the most profit?

25:18Yours is here as well.

25:20But also, you can orient to a new data

25:22source.

25:23Can it connect to Alexa yet?

25:26You could probably build a browser plugin

25:29so you can talk to Alexa and it should.

25:32Yeah, exactly.

25:34With something in between.

25:36I've always wanted an Alexa plugin for the

25:39server.

25:40It goes "boom, the server is down."

25:43Now Ask Data has been trained to interpret

25:46industry and will answer questions about

25:49industry even if it's never seen them

25:51before.

25:52This is good.

25:54There you have it.

25:58A new sneak preview of the completely

25:59redesigned Ask Data user interface.

26:02But that is not all.

26:05We're also introducing governance tools

26:06that let you fine tune Ask Data experiences

26:10to address specific business use cases.

26:12And we're increasing accessibility so you

26:14can bring Ask Data to every user in your

26:16organization.

26:18The new Ask Data is going to be fantastic.

26:21We're making asking questions about data as

26:24easy as a Google search.

26:26With 2.0, Ask Data is easier, smarter,

26:30faster, and easier to govern at scale.

26:34I can't wait to bring it to you all.

26:36Thank you.

26:38Hey y'all, I'm Dorey, a software engineer

26:42on the Metrics team.

26:43Broadcasting to you live-ish from my desk

26:46here in Austin, Texas.

26:49There are a ton of great new features in

26:51Tableau Server.

26:53I'm excited to show you a few of them.

26:55We've really tried to focus on the onboard

26:56ing experience recently.

26:58We want your users to be successful from

27:01day one.

27:03We want you to start with a very exciting

27:06checkbox.

27:07Yes, the grant roll on sign in checkbox is

27:10new.

27:11While it's tiny, it is mighty.

27:14One click can keep your users from having

27:15to pester your server administrator.

27:18Just check the box.

27:20And when your users first sign in, they'll

27:22automatically get a role assigned to them.

27:27To the user, this looks like a normal sign

27:30in.

27:32Sorry?

27:33You're quieter?

27:35Let's not turn you up.

27:38I've got an audio mixer for us today, so it

27:41's a lot better.

27:43It's a lot better.

27:45So your newest users can get a preview of

27:47the sweet visits they'll have access to.

27:50They'll be even more tempted to dive right

27:53in.

27:54Once you've signed in, you've got some cool

27:56new concepts coming your way in Tableau

27:58Server.

28:00I want a welcome email.

28:02I want to customize welcome emails.

28:05But when I want to go find those shared

28:07visits weeks later, I end up digging

28:10through old emails.

28:11Having to ask a colleague to share that

28:13same dashboard yet again.

28:16I like that they take advantage of this

28:17conference to remind people what they just

28:20released.

28:21So people just miss it.

28:23Of course.

28:25We're just upgrading from 2019/03 to 2020.

28:29It's way easier to track down design news.

28:31I can even see which custom views they

28:33share.

28:34In this case, a blue cone.

28:37Speaking of custom views, we've given them

28:41their own section on the work page.

28:44Now authors can see what custom views exist

28:46on our workbook and can change the owner or

28:50delete each custom view.

28:52Basically, you can now see and understand

28:54your custom views.

28:57Now, we all know that Tableau is the best

28:59platform for data and analytics.

29:02But I still sometimes view that full force.

29:06You know what? That's okay.

29:09Tableau has made my life easier by bringing

29:11export to Excel to the browser.

29:14I can now see export as a crosstab.

29:18You can select an Excel format and Tableau

29:20will maintain more of your table formatting

29:23.

29:25This is another one. It kind of drives me

29:27crazy.

29:27It's great. Excel users can stop

29:30complaining.

29:32Helping me with my best Excel lifestyle.

29:35I know Ellen did collections earlier, but I

29:38'm so excited.

29:39I just have to say collections are awesome.

29:42I think Mark's right.

29:44You don't know you need it.

29:46Without moving business around between

29:48projects.

29:49Here, I've got a metric in my monitoring

29:52project that I want to share with you.

29:54I want to share with the executives and my

29:56working team.

29:57I select it, choose add to collections, and

30:00assign it to both collections.

30:02This was really good yesterday when Ellie

30:04Fields showed this off.

30:06Some really good thought gone into this.

30:09Which I really like.

30:11It's funny. Before collections.

30:14I've always recommended tags as a way of

30:16curating data.

30:18The amount of times I've had the

30:20conversation, "Can I certify a workbook?"

30:23This is going to make a big difference when

30:24it arrives.

30:25Keep all your notifications from across the

30:27world.

30:28Talking data alerts, subscriptions, sharing

30:30, commenting.

30:31All organized and actionable.

30:34Looks like Ron's left a comment that it's

30:36about time to wrap up this demo.

30:37I guess I've got to get back to work.

30:39That's it for me from Texas.

30:42Thanks so much for watching, and have a

30:44great rest of the conference.

30:46Ish.

30:48Well said.

30:51Hello, my virtual friends.

30:53My name is Miranda Osterhelm, and I'm a

30:55Tableau PM who gets really excited by

30:58features that make life for admins easier.

31:00Those features that deliver flexibility

31:02with control and security to keep my

31:04deployment safe.

31:06One of those awesome new features is

31:08changing the way extra refresh scheduling

31:10is handled in Tableau Server and online.

31:13I have a dataset here that contains all the

31:15contents of a grocery delivery I received.

31:19During these very interesting last few

31:21months, I've gotten super into the food,

31:25probably at the coping mechanism, and

31:27paired it with a very anxious need for

31:29control and planning.

31:31So I use this dataset to make sure I use

31:33all my produce before it goes bad.

31:36Previously, I was only able to schedule a

31:38refresh of this extract on a schedule pre

31:41configured by my server admin.

31:45But we've heard from many of you that you

31:47want the ability to take control of config

31:49uring that extract refresh schedule since

31:52you know best when it's time to update your

31:54data.

31:55Introducing custom schedules.

31:59This feature brings us the ability to

32:01configure an extract refresh schedule that

32:03fits my exact needs.

32:05Let me show you.

32:08I get a schedule of refresh for the first,

32:10the fifteenth, and the twenty-first of each

32:13month at twelve thirty-five.

32:16Where you're at right now, I know I'm not

32:19planning with the breakfast of data that's

32:23refreshed on a schedule that has the

32:24flexibility that real life often needs.

32:27I love this bespoke way of scheduling.

32:31This is great because custom schedules

32:34helps me prevent spoiled produce and stale

32:37food. And the only thing worse than stale

32:39food is stale server content.

32:41Hey my server admins, you know it's a

32:43daunting task to keep content on the server

32:47relevant so your users don't have to sift

32:49through outdated workbooks to find the dash

32:51boards that they need.

32:53And we all know, we don't want musty

32:55content taking up question free space on

32:58our servers.

33:00With the new stale content admin view, I

33:02can identify content, workbooks, and data

33:05sources that haven't been accessed by my

33:07users over a time period threshold that I

33:10set.

33:11Let me show you. I can filter down to some

33:14stale workbooks and select some of the ones

33:17that are taking up a good amount of space

33:20on my server.

33:21Because those are good candidates for

33:22closer inspection.

33:25In fact, let's go ahead and tag these work

33:28books as stale. And then at any time, I can

33:31always filter down for the workbook I've

33:34identified using that dashboard.

33:36Inspect them and toss them as necessary to

33:38prevent them from cluttering up my site.

33:42That covers stale content. Now let's talk

33:45about stale data.

33:48You clearly didn't watch my videos, Ravi.

33:55Too busy being a zen, right?

33:55Bitch!

33:58No time for a sample fake.

34:04Oh gosh, there's your shelves.

34:15Dom doesn't want his workbooks to go stale.

34:19I'm guessing there's a lot of... I remember

34:22the Tilt server. There's a lot in there.

34:25Ravi, since you've been gone, we've

34:27completely overhauled it. You're talking

34:30about the past.

34:32You're talking about the past.

34:35You're no longer part of us.

34:38You're lost to us.

34:41No more. Now we can do it all from the web.

34:44See? Here I can update my credentials.

34:47Bridge is always like a... You don't

34:50understand it until you use it.

34:52I've just set this up for a client. It's

34:55really good.

34:56I think it's one of these things where it's

34:58going to be more and more prevalent as the

35:00move to cloud happens.

35:02I think the last six months I've seen a lot

35:04of people move towards the cloud.

35:07Which allows Tableau Online to select the

35:09best possible bridge to send any request

35:11for job to.

35:12Or office infrastructure is not good.

35:15You suddenly leave the office and

35:16everything is slow.

35:18Make life as a Tableau admin so much easier

35:21.

35:21And keep my end users secure and self-

35:23service.

35:24Which is about the only thing that can get

35:26me more excited than fresh produce.

35:28Smart Bridge is good.

35:30Thanks for your time y'all. Enjoy your

35:32Tableau conference.

35:34Thank you guys. Allowing something to be

35:37smarter than you to choose the right

35:39schedule.

35:40Wow! Keisha, that was quite a jam-packed

35:43session.

35:44Which was your favorite demo?

35:46Come on Andy. There's no way I could be as

35:48quiet as you like. I love them all.

35:50The demos and desks were all pretty

35:52inspiring.

35:54But before we go, I want to give a quick

35:56shout out to our amazing dev team.

36:00So let's recap. What did we go through? Can

36:04you remember what the first thing was? It

36:04was Filippos first.

36:05Filippos. We missed the beginning because I

36:06was watching a lag stream.

36:08So you probably heard the beginning of that

36:10. But models, right? Models.

36:13Yeah, so they're like really forecasting

36:14models. Giving a bit more control.

36:17And it looks better. I think that's what I

36:19think.

36:20I think it looks a bit better. Which is

36:22great.

36:26I'm always a bit skeptical on what's

36:28happening with giving people too much

36:32control.

36:34It's good to have the option, right?

36:38It's good to take your learning journey as

36:40far as you get and discover you've got the

36:42availability to see it.

36:43It just pushes that sort of edge further

36:45out.

36:46Then we had layers.

36:48Map layers. Map layers.

36:50We're not sure if it's map layers or just

36:52map layers.

36:53Good point. Good point.

36:55Visit the Beyond Tableau page of the event

36:57website for the link to the server.

36:59There is so much coming up.

37:02What do you think to that? What do you

37:05think to map layers?

37:05To help you stay focused and keep you energ

37:07ized, we have a brand new caffeinated

37:09partner who promises to keep you motivated.

37:11General layers. I think there's other

37:13creative use cases for layers in general.

37:16Like just beyond maps. And I think that's

37:18actually more powerful features.

37:20I'd love it to be more than just map layers

37:21.

37:23What you don't want that to become is

37:24another way of doing a de-allaxes chart.

37:27You don't want that to dress it up.

37:30I have, I have, I think.

37:32I'll turn it right down.

37:34I think you're absolutely right.

37:39I think there are use cases.

37:43Imagine if this allows you to be more

37:45dynamic with adding and removing layers

37:48based on a data direction.

37:51So I think the stream is over so I might as

37:54well just go back to this view here.

37:58Yeah.

38:00No, absolutely. So I think that's my

38:02thoughts on map layers. I'm intrigued.

38:05I mean, it's similar to SaaActions. I was

38:07like, yeah, this is cool, but I don't know

38:08what, what could be possible.

38:10And I'll leave that to the community to

38:11blow my mind away. Mark Reed if you're out

38:13there.

38:14Yeah. This is it. This is it. This is it.

38:18There are certain people in the community

38:19that just take these features and just tell

38:20us how they should be used. Right?

38:22Yeah. No, absolutely. Absolutely.

38:25So we had, what do we have there? Then we

38:27moved on to ask data, was it?

38:29Yes, absolutely. No, no. Ask data. What was

38:33after, just after, um, uh, thing. Let me

38:36scroll back in the chat. That's a hack.

38:38We talked a little bit about prep. Yeah. So

38:43prep.

38:44Oh yeah, of course. Yeah. There was a prep

38:46strategy.

38:47We had the prep scene come in.

38:48Yeah, yeah, exactly. Um, so we had, well,

38:51prep in the browser was yesterday. That's a

38:53really sort of a big, big feature.

38:55Quality of life improvements, hiding

38:57columns that you don't need anymore, which

38:59is again, a big one.

39:01I hope that's linked to sort of not using

39:03those fields when it's doing that sort of,

39:05you know, let me look through your data

39:07kind of search.

39:09Uh, progress bar that it always does that I

39:11think frustrates a few of us. Um, and then

39:14after that we had some, um, the calculation

39:17.

39:18So, um, Ava was doing something with

39:20calculations. I think she was doing a level

39:23of detail. No, not level of detail

39:24calculations.

39:25What was she doing? Um, I need to have, I

39:28should have taken notes, right?

39:30But yeah, I think the, the, the, the main

39:32one for me was the grouped steps. I think

39:34that was very good.

39:36I really liked the way that they, they came

39:38through. Um, so it was, it was quite good

39:41to see them and yeah, no, it's funny

39:44watching back to this.

39:47Um, the, like, it was just feature, feature

39:49, feature, so many good, so many good things

39:50coming through this.

39:52Um, I obviously professed to quit Tableau

39:55if they didn't do anything about layout

39:57containers.

39:58So guys, I guess that's it.

40:04Dashboarding didn't get any love. Spatial

40:05Calcim Prep. That was a big thing. Thank

40:07you.

40:08Um, yeah, I think that, that, that, that'll

40:10be, that'll be a really big game changer as

40:12Mark said earlier in the chat.

40:14Uh, I think what's going to be very cool is

40:15, uh, when they add in that map layer

40:17interface, especially if they're bringing

40:19like, you know, the, the native Mapbox maps

40:21and different layers we can see there.

40:24So you can do football analysis, right?

40:26Yeah, exactly. And I mean, we've just

40:29started this process of moving towards

40:30Python and suddenly we're getting pulled

40:32back into prep.

40:33Now I think it'll be cool to be able to

40:34give a few people, you know, um, that, that

40:37level of, um, access to things like the

40:39tracking data.

40:41Right.

40:42If you're able to project, but it's good to

40:44, good to dig into that a bit more.

40:46Exactly.

40:47Um, grand. So we had that and then we had,

40:49um, ooh, uh, uh, collections, collections

40:53got a recap.

40:56Collections from yesterday. Um, so if

40:58anyone who missed that, uh, basically a

41:00home for all your data visualizations,

41:02right? Which isn't just, uh, as Francois

41:04said, isn't just a bunch of links on a

41:05SharePoint page or something like that,

41:07which is good.

41:08It's good. It brings, brings things

41:09together, especially being able to put

41:11metrics alongside dashboards will make that

41:13really sort of a good space.

41:15Hopefully in the future, we get a little

41:16bit of control of how that page looks right

41:17.

41:18Then you can really start to see sort of a

41:19landing page comes to life.

41:21I mean you can embed it.

41:24We completely skipped the R state there. I

41:26'll say it got some bit of love with the

41:28subtleties.

41:29The integrated search function is going to

41:31make it easier. And I think again, like I

41:33said, I'm intrigued to see it in practice

41:35in terms of speed and the understanding of

41:38how that, how that will look together.

41:41I'm really, I think it's a good step in the

41:43right direction, especially with the

41:45context.

41:47And then we went, went on to server custom

41:50views, listing way, really happy with that

41:53one. Really, really happy with that one.

41:54Uh, so we can, we can dig into that one as

41:56well, as well as, uh, custom schedules.

41:59Just makes it easy to do some housekeeping,

42:00right?

42:01I'm happy with both of those features. I

42:03think, um, custom schedules is going to be

42:05super useful in terms of like, you know, fi

42:07finding out what's, what's happening, uh,

42:10finding out, um, the different elements of,

42:15of what people are looking at.

42:16What people are most interested in. Like we

42:17have a bunch of custom views and I think

42:19managing it will become a lot easier.

42:21Uh, and we also reach out to people that

42:22are creating them. Right. Um, quite handy.

42:25And then alongside that, obviously, um, we,

42:27we talked about custom schedules, the

42:29controversial server admin topic.

42:31Yeah. Custom schedules, who gets the option

42:34to, uh, create those?

42:36It'll be site admins, but I think you made

42:39a good point.

42:40Francois said, this is key for online,

42:42right? So where you get a set of schedules,

42:44right? So being able to do custom schedules

42:46is, is a, is a good thing there because you

42:48don't need to run something every hour, but

42:51sometimes you'll choose that because that's

42:53the only option you have to get something

42:54in at a certain window.

42:56So that's what you end up running. So could

42:58, could bring some performance improvements

42:59for everyone in the long term, right?

43:02Yeah, exactly. Exactly. I think the other

43:04thing was, uh, it's going to be superb for

43:06Tableau Online. I think it's good. It's

43:08good to see online getting a bit more love.

43:10And I think it's a really good, um, I quite

43:12like online as a concept because no, not

43:15every company is going to have an IT team

43:18or it makes it more accessible for us, a

43:20small organization to not have to worry

43:23about that sort of thing.

43:26And I think, you know, maybe this helps it

43:27become more scalable as well. Yeah. When

43:30you have multiple data sources that you,

43:31you might want to know more, a different

43:33schedule, you know, so things like that.

43:36It's good. Stale content and style content

43:41tagging. That's actually in 2020.3, which

43:43is good. Um, so yeah, it's pretty much it.

43:46Um, that was, that was pretty much it.

43:50We didn't get a re we didn't get, um, low

43:52level security and getting a home as a sort

43:55of recap again, and this in the session.

43:59I mean, I'm intrigued. Are there any

44:01sessions? Have you looked at whether

44:03I did watch one? I did watch one. It was

44:06interesting. I think, I think the real key

44:08benefit is having it centralized, right?

44:10Because you always had this issue about,

44:12you know, taking extracts.

44:15And then when you take the extract with

44:16something like database row level security,

44:18it captures just the security for the

44:19person who took the extract.

44:21So bringing a centralized home on server

44:23where you can build those makes it slightly

44:25easier. And it sort of separates the author

44:28ing from the security, which makes sense

44:30because on Tableau server, there's a really

44:31good security model, right? So that is

44:34probably a better home.

44:36I'm not sure where that sits. There's a ski

44:37there. Yeah.

44:39So here's my second thing, right? So I

44:40think I didn't really use user filters too

44:42much when I was in the information lab, but

44:44we've thought we use them quite a lot here.

44:47Obviously, there's a lot of medical data

44:49that has to have security applied to it.

44:51One of the things we noticed this summer.

44:53So in terms of the structure, the U16s and

44:5515s were previously combined and now split.

45:00What we found was we had to re-extract the

45:02data source, remove the user filter, change

45:05the user filter after regenerating to a

45:07live data source to tick the performance

45:10things and then bring it through again.

45:13It didn't just give us the option when the

45:15extract ran. I think that's because the way

45:17that user filters operate is it creates

45:19static list for you to then match up to

45:21your different groups of users.

45:25So hopefully this maybe moves that away

45:27from that sort of thing. And again, I think

45:30we always talk about this at the

45:31information level. We did.

45:34Talking like he's still there. I hope

45:37everyone caught that. He thinks he still

45:40works for us.

45:42The relevant security never gets enough

45:45love. There's not enough content out there

45:49about user filters and relevant security

45:52and the different elements. In part, I

45:55think because people are reticent to share

45:56the different things they're doing in their

45:59data sources, their operations.

46:01But perhaps also because it's kind of a

46:04tricky concept to explain.

46:08It's a hard one. Yeah.

46:11Maybe this makes it easier to explain.

46:12For me, I think it's the groundwork. I

46:15think there's an opportunity to add this

46:16really tough challenge where you have a

46:18single business unit. Let's say I'm working

46:21for a fast moving consumer. Good. I sell ch

46:23ocolates.

46:24I have a marketing, sales, finance,

46:27distribution, stock and inventory team.

46:31Right?

46:33How does it all work for, drum roll, one

46:35brand? How do I give that team access to

46:38the data that they need to do to do their

46:41jobs and simultaneously doing the exact

46:43same thing for another brand?

46:46Then do the same thing for all their senior

46:48managers. Then do the same thing for the

46:50executive. Then do the same thing for the

46:51CEO.

46:53Have that hierarchy beautifully flexible so

46:56that I don't have to keep creating projects

46:59to separate the finance people from the

47:02sales people and groups and is member of

47:05the finance centre to manage that.

47:09It's trying to, and you almost want that

47:12centralised thing to then also start

47:14including permissions for content.

47:19And suddenly you've got a really rich

47:21environment where you can control privacy

47:23settings and permissions and also imperson

47:26ate hopefully.

47:28I spoke to a dev a couple of months ago or

47:31a month or so ago about what I don't like

47:35about data management add-on. It doesn't

47:39allow me to manage my data.

47:42Right. It doesn't do what it says in the

47:44tin.

47:46It allows you to look at lineage. It allows

47:47you to catalogue your data. But does it

47:49allow me to add in comments to my

47:50calculated fields? Does it allow me to add

47:52in the rich metadata of adding folders?

47:55It's getting there. It's getting there. You

47:57can tag fields.

47:59Yeah, but you can't do everything you want

48:02for a TDSX. So TDSX being my packaged data

48:05source.

48:06And you still have to create those and then

48:08publish them as a TDSX with the rich

48:11folders, with the calculated fields, etc.

48:15I'd love to be able to do that within an

48:17interface. And when I saw that yesterday, I

48:19was like, "Oh, damn, they work quite fast."

48:22But then it was relevant security.

48:24It's a similar sort of feel and look where

48:27you're thinking like, "Hey, if this becomes

48:30a norm, suddenly this interface really

48:33allows you to..."

48:36It's kind of similar to Power BI, right?

48:39And their DAX interface where you could

48:40start doing prep.

48:43You mentioned their name. How dare you?

48:45On your channel.

48:47How dare you? Oh, God. I get fed up with

48:50the Power BI people going, "Oh, why are you

48:53making videos about Tableau?" And I'm like

48:57...

48:57It's Tableau Tim.

48:59Get with the program, honestly. No, but you

49:02make a good point. I think the data

49:03management add on is a new product.

49:07When I give an example, a car maker

49:08releases a new car, the first model is

49:10always like, "Okay, I don't really get it."

49:13Second model, they really sort of

49:14understand their customer.

49:15Third model, they nail it. Exactly. And so

49:17this is the data management add on.

49:19The iPad wasn't good until the iPad Air.

49:21We're getting there. And I think this is

49:23the start of something. I think that

49:24eventually I think you'll get to where you

49:26're talking about.

49:27But in order to do that, you have to lay

49:28the groundwork. And a lot of this stuff is

49:30the groundwork. And it's not as sort of

49:31sexy, but it will get there.

49:34You've got a similar story with prep. I've

49:36always been like, "I can see what prep can

49:38do and what it could be, but it just isn't

49:42."

49:43It's not there yet.

49:45But it's one of these things that's a

49:46chicken and egg scenario. How much alpha

49:47testing and beta testing do you want to do?

49:50Where we just chuck it out there and be

49:51like, "Right, everyone else figure out the

49:53contours and then tell us what you want."

49:57The more mature the product gets, the more

49:58you have to do that because for every

50:00incremental feature, there's potential to

50:02break a whole load more features.

50:04So when Tableau is back in version 7, you

50:06add a feature, you're not going to break

50:08that much stuff because it's fairly new.

50:11You redesign the backend. Again, you have

50:13to redesign a whole bunch of features that

50:15can break stuff.

50:17As we move more and more into this more

50:19fully featured mature product, you stand

50:22the chance of pissing off a lot of the

50:24people you've brought behind you.

50:27I think you do have to take more caution

50:28before you put things out.

50:30You can see it in Tableau's demos. They

50:32really take care in how they demo their

50:34product. If you do trainers, training.

50:37There's so much effort that goes into the

50:39script.

50:40Yeah.

50:43Carl makes a good point here. We're saying

50:44that about prep because I've had other

50:45options and I've looked at other things.

50:47That's absolutely true. If you have nothing

50:50. I make the exact same point when people

50:53complain about web edit.

50:55Well, yeah, but if you teach web edit like

50:56you're teaching desktop and you're telling

50:58them what they're missing.

51:00It's like if you're trying to sell a car

51:02and be like, "Well, it's not like a Porsche

51:04or it's not like a Land Rover, but it

51:07drives."

51:09You're not selling it properly and it's all

51:10about delivery. I think Carl's absolutely

51:12right. If you don't have anything and you

51:16see prep, it changes your life.

51:17It can really get you, to coin a phrase, up

51:19and running.

51:21Hey!

51:23It's one of these things where it has to go

51:26through that motion. As you say, when we

51:29see new features and we see new different

51:31elements to it.

51:34It comes back to this whole concept of,

51:37well, we talk about, we've mentioned was it

51:41map layers, spatial things in prep today.

51:45Immediately, both of us, I was like, "Yeah,

51:47show me your map." It's like, "We're not

51:49there yet."

51:50Just added in this spatial engine, like, "

51:52Jesus Christ, give us a break."

51:55Because, like you say, the second you make

51:57that step, other options start to pop open

51:59and you start thinking about different

52:02things.

52:03Correct.

52:04Wowzers. This has been a fully featured

52:07little thing. I think it was so much more

52:10compressed today than usual.

52:12I think you get the fanfare of conference.

52:15You spend 20 minutes just getting into the

52:17venue and that is the hype.

52:19That is most of the hype, actually. Then

52:21you have the moment on stage and then I

52:24think when you're there in an arena, it

52:26just feels more atmospheric.

52:30I really hope next year we can do this in

52:32person. Exactly.

52:34There wasn't a pause to wait for the clap.

52:37There wasn't a "Let me show you" pause.

52:38There wasn't a high five between them. They

52:40could have had pineapples, Francois. Where

52:42was the pineapple? Come on.

52:44I think he's gone now.

52:46You'll catch this later. Absolutely. I've

52:49just got the gallery up on the stream

52:50instead rather than just having a session

52:52playing in the background.

52:55These are some of the faces you'd see. You

52:57had the t-shirt cannons, they brought those

53:00light sticks out last year which is great.

53:04There's so much atmosphere actually just

53:07being there itself. We can't recreate in

53:10the stream. The stream is good but it can't

53:12recreate these things. Maybe next time we

53:13'll have effects.

53:15I think we've all been to it quite nicely.

53:18I'll get a stream deck.

53:21I didn't even use this.

53:24Could have got this out. I completely

53:27forgot it was on my desk.

53:29What's in the back of it? Is it a thumbs

53:32down?

53:32No, it's just the thing. I just did this.

53:36Fair enough. That would have been good.

53:38Didn't even use this.

53:40Next stream.

53:42I'll do a stream deck and some effects.

53:46That's pretty much it.

53:47I think we can call it there. Thank you

53:49everyone for joining. I think we're an hour

53:52.

53:53Hopefully this was fun and useful. We tried

53:56this mainly for us to recreate the sessions

54:00.

54:01The atmosphere. This is purely a vehicle

54:04for me and Ravi to share some banter

54:06basically.

54:07If you don't want to sit there listening in

54:09silence and just want to hear us chatting

54:12about NH.

54:13Can I swear on that?

54:17I don't know. I don't really have any rules

54:19against that. It depends what I marked it

54:21as on YouTube. Is it safe for kids or not?

54:24Excellent. I've got a brain date in 15

54:27minutes. I'm going to try and find some.

54:29Same here. I've got a brain date. Several

54:32brain dates actually.

54:33Thank you everyone. Thanks for tuning in.

54:36Hope you enjoyed it.

54:37We got there with the audio but I don't

54:39sound like Tupac Shakur today.

54:44We will hopefully do this sometime soon. We

54:45're not doing this for Iron Viz just because

54:47Iron Viz has its own commentary.

54:49We kind of want to watch it. We don't want

54:51to talk over the commentary there. That's

54:53part of its own thing.

54:54Hopefully stay tuned. We're going to do a

54:56podcast new season coming out very soon.

55:00Look out for that and we'll get back on

55:02track as we typically do.

55:04We've got a fair few comments just saying

55:05thank you. Thanks to Karl and Mark for

55:06coming into both sessions.

55:09I don't know what the song of TC20 has been

55:13. There's one to grow on.

55:16Whitney Houston so far. That's the only

55:19thing that keeps ringing my bell.

55:20I've got a copyright strike from yesterday

55:23's stream. I have to share the revenue from

55:25the stream as if there's any adverts on

55:27there.

55:28Meraki would like to connect with us

55:30offline. Yeah, absolutely. Both me and Tim

55:32are on Slack.

55:34We're chatting next week Meraki. Looking

55:37forward to that. I've got stuff lined up

55:39for you.

55:40Again, like I said, great stuff on the prep

55:42development and things like that.

55:45Let's call it a wrap. Thanks everyone for

55:48joining. Take care.

55:50Thank you. Cheers.

55:52I have to wait.