Tableau at Dreamforce and 24-3 updates to Tableau - S5 E6: Byte

0:00Ravi, we're back.

0:02We're back from paternity leave.

0:05Summer vacation.

0:07Summer vacation.

0:08Yeah, because I I can't take it on a vacation because my summers are busy.

0:12Yeah.

0:14Let's count it as the football season's underway.

0:16We've had two international breaks already.

0:18Yeah.

0:19Chambers League started, the new format.

0:21Yeah.

0:22And you've got a new member of your family.

0:24Yeah, absolutely.

0:25And the Premier League starts as well.

0:27And yeah, football season.

0:28Yeah, yeah, no, so much going on, honestly.

0:30Um on on top of all of that, we've had Dream Force.

0:34We've had

0:35New releases from Tableau.

0:37There's lots of energy in the analytics space, I think.

0:40Like it feels it feels exciting for the first time again, doesn't it?

0:43Like Tactoba

0:45Yeah, Techtober.

0:46I think it's maybe just that actually.

0:48October is when all the announcements get um kind of laid out.

0:51So it's good.

0:53And you sound better than ever.

0:55We're here now live with a new mic kindly donated by

0:59TN Media.

1:02I think otherwise known as Tableau Tim.

1:05Otherwise known as Tableau Tim.

1:06I think it was on a on a run that I was listening back to our last podcast, which was two months ago now.

1:10Right.

1:11I was like, Tim, this is there's so much echo, you're so kind.

1:14as to um how how good my mic actually is.

1:18Um unless I'm right against it.

1:20So But we do have to keep your mic game on point.

1:23Every time you look away it goes quiet.

1:25So

1:25Uh keep keep keeping it.

1:28Yeah.

1:29But no, it's it's a good little upgrade for the podcast.

1:32I think it's about time to be briefly honest.

1:34Like I think um

1:35Yeah, we we're gonna do this more often.

1:37It's a bit of an investment um and I think it makes sense.

1:40So we have a lot of topics to talk about.

1:42I'll do quick

1:44Uh like what is it?

1:4510,000 foot view, they like to call it in AWS land.

1:49Dreamforce, all the stuff that got announced there, agents, Einstein got a proper showcase and a tableau keynote.

1:55Ah, agents versus co-pilots will come onto that.

1:58Uh as you would like to say, mind the gap.

2:01Mind the gap.

2:02Mind the gap.

2:04With the classic Salesforce slide about the gap between, you know, everyday analysts and the rest of the business.

2:10And then there was a couple announcements since Dreamforce that kind of really speak to the fact that Tableau is becoming a platform.

2:18We heard some announcements from DBT and Tableau partnering together, but there's also a couple of other things we will sort of dig into there.

2:25And then lastly, um, new releases.

2:27There's been lots of new tableau.

2:29Um Einstein has been getting updated monthly, Pulse has been getting updated monthly, but 24-3 is like the big

2:36Statement release, the last one of the year is not a release for server, but it is a release for cloud.

2:41So lots of things there that are

2:43coming out for desktop and cloud.

2:45So yeah, lots to get into.

2:47Lots to get into.

2:48And DFM Europe is only a month away.

2:51Of course, of course.

2:52My first here's a crazy, here's a crazy fact.

2:55I've never been to a Tableau conference as Tableau Tim.

2:59This is true.

3:00This is true.

3:01You information lab, Tim.

3:03Yes, exactly.

3:04Exactly.

3:05I I've not been a to s to a conference since twenty nineteen.

3:08I started my channel after the Las Vegas conference.

3:13I went to New York with uh Bree and uh back then we weren't married, which now my now my wife and yeah, it was crazy and then COVID.

3:22Yeah.

3:25So I've not been to a tableau conference.

3:29I've been to a tableau event virtually.

3:31You weren't even a visionary then.

3:33Uh I've not yeah, true, yeah.

3:36True, very true, yeah.

3:38You were just a dude.

3:39I was just a dude.

3:40I was just uh what what what do you call us when you're not

3:43Uh a summer.

3:46A partner or a customer, yeah.

3:50Member of the data fam.

3:51Yeah, exactly.

3:52So I'm looking forward to it.

3:53I am gen I've taken time off like either side of either side of this conference.

3:57I I really want to enjoy it.

3:59For me, I think it's just a great opportunity to meet people yourself, obviously, and tons of other awesome people that can make it

4:05to Europe, um including the Tableau devs, Tableau product managers, Tableau Leadership.

4:11I mean have you seen where it is, the location?

4:13It's literally

4:15next door to the data school information lab.

4:18Ah man.

4:19So it's behind St.

4:20Paul's Cathedral.

4:21So it's a proper walk down memory lane.

4:24It is gonna be a full on

4:26Walk down memory lane at Cannon Street, coming off like all the oh maybe I can get some of my old bad habits in and go get me some uh Leon on the way in.

4:36That Leon I don't think is there anymore.

4:38Is it not?

4:39Is it did it get closed down, the one next to data school?

4:41I think so.

4:42That must have been like a post-COVID thing.

4:44Well there'll be a fifty pretz that I can lean on instead of.

4:47Borito Joe's is no longer a stand, it's a shop.

4:50Alright.

4:50Yeah.

4:51Anyway

4:51It's a shop.

4:54Wow.

4:54Okay.

4:55You know what?

4:56It's a little secret.

4:58I hated Burita Joe

4:59I never liked the thing.

5:01I never liked the place.

5:02I went there twice and I just did not get the hype.

5:05You know, I'm a Shake Shack person.

5:07That's that's where you'll find me.

5:08Nice uh greasy food.

5:10But hey, there you go

5:12Toby.

5:12Let's get onto it.

5:14Dreamforce.

5:14Dreamforce.

5:15What did you think?

5:16What did you take away?

5:17It was the big event, right?

5:18Like I think

5:19There there's a lot of things there was the event prior to Dreamforce where we heard about the new Tableau experience, which is Tableau Einstein, which I think we covered in depth.

5:28in our last video.

5:29Um but Dreamforce was the one where they this was coming to life.

5:33Yeah.

5:34Um my my favorite moment was uh

5:38the the the gag between tableau pulse and tableau plus by the chief marketing officer.

5:43Former chief marketing officer.

5:45Former, yes, yes correct.

5:47Emacs has left the building.

5:49She's now at contentful

5:51Which is a content company helps people write content.

5:53Anyway, I'm max.

5:55Yeah, no, um but yeah, Tableau Pulse, Tableau Plus.

5:57Um even happens to the best of us.

5:59Um but yeah the biggest biggest change obviously is uh agent force for Tableau.

6:05Uh so goodbye to Einstein Copilot, hello to

6:08Agent force for Tableau powered by Tableau Einstein on the Einstein Analytics platform.

6:15Secured by the Einstein Trust, which you need a Tableau Plus.

6:20uh s uh Sierra number to to to Akagadia.

6:23I think I think you know when I posted about that on LinkedIn

6:27I wrote that sentence and I was a yes, I thought I'd drag it out a little bit and just, you know, just go all the way with that sentence to just highlight the silliness sometimes of um

6:38I want to say tableau product naming, but I think I have to call it Salesforce naming.

6:44There is a sort of big search to always give everything a name, but I I think you could just be massively simple.

6:50I kind of said to um

6:52someone from product marketing that take a leaf out of Apple.

6:56The feature can be named simply, but you don't need to keep shoving like Einstein and Tableau in front of them.

7:02They can just be named after the things they do.

7:04But anyway, we'll get into that.

7:06So I think the thing about this dream force was that really what it felt like

7:12A pivot.

7:12And to be fair, and I think this is a really important context to take, to be fair, the whole industry is pivoting away from copilots to agents.

7:20So this isn't just like

7:21something that Mark Benioff has sort of come in and thought there is a general mood in the industry with AI that everything we've seen so far is great, but

7:32Things need more agency.

7:34These these AI capabilities need to be able to do things on your behalf.

7:38And so agents have become the moniker that you know a lot of startups but also big, large tech companies are using.

7:45That's why

7:46you know, Gemini is is is going to be an agent.

7:49You know, Siri is going to be an agent on your phone in in many ways.

7:52And so the first time I heard about an agent was when I was deciding what to do with my dissertation.

7:59Um

8:00And it was uh yeah, my usual behavioral science, blah, blah, blah.

8:03Right.

8:04And that you could do a simul that there was a paper that did a simulation which of the the village or werewolf problem of you need or vampire, whatever you want to call it.

8:11And it was a simulation where you create agents.

8:14To you gave it a program and you basically said you have a finite amount of villages, but and everything would as every day passes, you need to give an offering.

8:23Like there is an equilibrium workout.

8:26And then the machine would go out and work out what the equilibrium was.

8:29Which is basically what a complex AI agent will do, right?

8:33Like I think the most interesting thing about this is

8:37Everything that's happening around AI is very well documented, right?

8:41Like we are seeing the conversation happen in front of us and it's

8:45historical whereas I think definitely with the the internet like the World Wide Web Yeah it's you you have to go looking for it even early early innovations in

8:55social media in the tech space, you can't really find it.

8:57Whereas this sort of thing is the revolution is very much happening online.

9:01Exactly.

9:03Which is where like you most are comfortable with the pace it's going at because you can see the pace it's going at

9:09Yeah, and in a way, um, I think the progress you see is more tangible because it's kind of happening in front of you.

9:16They're all building it out in public.

9:18I mean, we're gonna talk a bit about that with Einstein in a second, but

9:21Because they're building it in public, there is a sense that you can keep uh dare I use the pun, a pulse on what's going on.

9:30So, um, yeah, I I think that's absolutely right.

9:33Um, I do think though, I do think though, there is there is room for a little bit of what I would say wiggle room.

9:39And I think this is

9:40This is something I don't fully understand natively within Salesforce.

9:43So we have this agent force thing, right?

9:45And um I mean I'll be fairly blunt.

9:49Even product managers at Tableau weren't on dial the day

9:55this was announced, right?

9:56So it it feels to me like something got decided at Salesforce.

10:01It got branded as we are not going to talk about copilots anymore.

10:06Anything that says co-pilot, replace it, like find and replace it with agent force.

10:12And so everything became went from being a copilot to being an agent.

10:16But it doesn't feel like anyone looked at the experiences that people were using.

10:20And actually tested that question whether agent was a fitting term for the feature that they were going to replace copilot.

10:28Because I think copilot, in many sense, did actually make broad sense

10:33Uh when they initially taught of it.

10:34Like it made sense.

10:35Like you're kind of wingman or wingwoman that kind of gets you to where you need to be.

10:40It's not doing the work for you.

10:42You can instruct it, you drive it, you tell it what to do, right?

10:45But on some level, isn't an agent doing some of the work for you, right?

10:49Like I think isn't that the slight difference here?

10:53And this is where I think there's a big difference between the Salesforce platform as a whole and why agents I think work there as a context and Tableau, it doesn't work because in Salesforce

11:03There are many activities that are essentially chained.

11:06Salesforce kept on talking about flows, um, you know, experiences, and those experiences are basically stitching together different parts of the Salesforce ecosystem.

11:15So there it does make a sense

11:17to talk about agents because you can ask an agent to to look at certain scenarios, handle certain types of interactions with customers, and have specific outcomes based on those interactions.

11:28An agent can go and do those things.

11:30But in the realm of Tableau, if you take that same theory, what you would expect from an agent in Tableau would be something that's monitoring your metrics, monitoring your dashboards

11:42understanding the metrics you really care about and surfacing them to you and telling you a story and pointing you to the things that need your attention.

11:50That to me is an agent.

11:51That's sort of where pulse is hopefully going to end up, but it's not there yet.

11:55But you're building in public, right?

11:57So it's okay because now you you by saying you're building and developing with the customer, and you know you don't you want to make sure it's

12:03fitting to the journey we're on together.

12:06Like you you can then end up releasing something that's not fully finished because you're not saying this is the polished product.

12:12This is a the start of something really big that was going to change everything

12:16And that's where I think, you know, in my video, I said there is a world where both can coexist.

12:23Like on your way to building the agent, you can still have copilots.

12:27It's totally fine.

12:28Like a developer when they're coding something in VS Code thinks of the autocomplete as a copilot, helping them write code.

12:36They still decide what they do.

12:38So I do think there was an opportunity missed to say, yeah, we're working towards agents, but we're not quite there yet, at least on this feature set.

12:45So what you have here in prep is a copilot.

12:48What you have here in Tableau is a copilot.

12:50Yeah.

12:50What you have in pulse and what you have in the whole of Tableau, that is an agent.

12:55And agents can work with co-part.

12:57You know, there is a whole world there that I think missed an opportunity.

13:01But I get from a marketing perspective the

13:03The cleanliness, the sort of um what's the word?

13:06Um there's a word, um it's tidy.

13:10It's very neat.

13:10Tidy, yeah.

13:11Very neat.

13:12Um

13:13It's the kind of word you'd use for like a dentist.

13:15Um what's the word?

13:16Sanitized.

13:18Sanitized, that's the one.

13:21You got me there.

13:22So yeah, it's a very sanitized way of dealing with the problem and probably wanted to move off the co-pilot naming.

13:27Because that's synonymous with Microsoft, someone they, you know, generally would like to be seen to be competing with.

13:32And whether or not they do directly is a whole nother thing, but um nonetheless, yes.

13:36So

13:37So now put putting the the hat on of someone who's a purchase decision maker customer, right?

13:44Yeah, yeah.

13:45What do you want to do?

13:45What am I buying?

13:47Yeah.

13:48Do you know what I mean?

13:49Like if if you if you're sat here

13:51Yeah, as a decision maker, what you end up having is what what am I buying?

13:56What is does my license get me and how much am I paying for what?

13:59And uh and yes, you might get a package, a deal, something that's bespoke created for you.

14:05But ultimately, it might not be the right thing, or you might get upsold something, and suddenly you've got something that's relatively sticky, and yeah, it it becomes a bit a lot more ambiguous and and I I don't think

14:19The average consumer likes ambiguity.

14:22I don't like ambiguity.

14:24Yeah, no one does.

14:25No one does.

14:26Um

14:27It's also quite I mean if I dig into what you've just said there, well, you're getting nothing unless you have Table A plus number one.

14:34And cloud and cloud.

14:36Yeah, exactly.

14:37Um

14:38So those two things are pretty important.

14:40And number two, um, there's only one pricing to you, so like everyone you basically Tableau Plus is like

14:49Like they like to say wall-to-wall tableau, you know, in many sense.

14:53Because of all the other stuff you get, it doesn't really make sense to do it in a small way.

14:57Um and then lastly

15:01I think this is like a what's the word?

15:03Um there is actually a a philosophical problem with AI in an analytics product.

15:09And it's simply this.

15:11Before AI

15:13With analytics products, it was already tough enough to make sure that what you were seeing was correct, right?

15:19Like it was already tough enough to deal with this, what I would call governance and validation problem that you know data sources create.

15:27And Tableau, we're all about building the tools to give you that visibility and trust.

15:31Yeah.

15:31Like freaking data management whole SKU exists entirely to ask this, answer this one problem.

15:39How do I build organization?

15:41Yeah, yeah.

15:42How do I build how do I build trust, validation, and governance into my platform?

15:49And data management was supposed to give you all of that, the lineage and all that.

15:53And here you are, just like a snap, like a Thanos snap.

15:56Boom.

15:57You've introduced AI.

15:58That goes out the window because the thing isn't accurate.

16:01Like the thing

16:02Thing generates mistakes.

16:04And the the worst thing is to validate those mistakes, I mean, so far, um, in different places is different experiences.

16:12In Tableau Prep, I actually think it's really, really good because it's very focused on calculations.

16:16Now it's very easy for it to get right and you can see the outcome very easily in prep.

16:20Natural fit.

16:21In Tableau Desktop, you know

16:24even just using one of its own prompts, it just it just falls off a cliff almost straight away.

16:29And it I think is there the scope is too broad.

16:32And so

16:33You know, you introduce this new problem, which is if you're a new analyst who this is supposed to help and you're gonna use this, you need your work peer reviewed by someone, right?

16:42You need your work

16:43peer reviewed by an experienced analyst, you're not solving the problem.

16:46You're just you're just creating a new one in many ways.

16:49And then on the other hand, if you're an advanced user, you don't use it because

16:52Like you you you try the most complex things you already know how to do and it can't do it yet.

16:57It tells you, oh this is not supported.

16:59It'll even do things like say, I can't write regex.

17:02Regex is not supported in Tableau.

17:04Hello

17:04calculation window has a reference thing that says regix is there.

17:09So you know it's it's just a really weird paradigm in a product.

17:13I just find it strange.

17:14I think the the tricky part again with desktop especially is people don't write calcs for the r reasons Tableau think they write calcs.

17:23Yeah.

17:23Yeah.

17:24For example.

17:25Yeah, yeah, 100%

17:26I've j I've found my first ever work or use case where I might have to do the map players hack to do it advanced chart type.

17:34And I've wanted to avoid this.

17:36I wanted this to like

17:38Have a splash flash in the pan and then disappear.

17:41But here I am.

17:42My two options are buy or pay someone to create a Viz extension for me.

17:48Right.

17:48Or use Map Players.

17:50So here alright, well so we here we we we're there.

17:52We I'm now career looking at map players to solve this problem I've got.

17:57Uh

17:59But then but how how is Tableau to know if I then say, okay, cool, I want to create a complex chart type, but I'll instead of using let's say

18:08A scatter plot and a bar.

18:11I want to use map players based on this.

18:15Let's say it's so smart.

18:17I can upload a sketch based on this sketch.

18:19Help me backward energy this

18:20It won't know what to do because this is very much a data creativity problem.

18:26It's not even a data wrangling problem.

18:27It's a creative data problem where you have to understand everything Tableau's doing, everything you want to do.

18:35And then create the steps in between.

18:37Which is why, again, another thing I've done recently is speak to someone who I think is one either in my team or someone who is one of our creators.

18:49Was using Chat GPT to create there or tweak the fixed LODs and I'm like, come on.

18:55It's not it's gonna create a really complex fixed LOD and it doesn't really understand how to do it.

19:00Yeah.

19:00So just first understand it.

19:02Because it's not like SQL that's more like homogeneous and there's almost like the entire of SAC Overflow solving SQL problems.

19:10Um it's still a niche skill, right.

19:14Yeah, exactly.

19:16And you know, don't get me wrong, I think these tools do surface knowledge to people in a way that I don't think that's ever been possible in the product.

19:23So much of

19:24Tableau is buried inside of Tableau.

19:26It's not, you know, the documentation doesn't surface these things.

19:30Uh we'll come to this later when we talk about 24-3 and spatial

19:33um parameters like I went down a massive rabbit hole there and I just there was just stuff I didn't know was the case and it's not in the documentation either.

19:41So, you know, yeah, absolutely.

19:43It's really, really, really tough.

19:45Um, but, you know, if if I sort of use that as a transition to the infamous slide

19:51that we always see with everyone's context this this is the slide that Salesforce or Tableau always use have done systematically at virtually every keynote

20:04for the last now three and a half years.

20:06It shows a gap.

20:08And on the left you have your uh Tableau community members, uh your visionaries, your ambassadors, and on the right

20:15Um sorry, I've explained it the wrong way around.

20:17On the right you have your, you know, experts and on the left you have your business users and analysts and they talk about the chasm and then they place the thing they're pitching in the middle as bringing a closing this gap basically

20:28Basically, such a you know metaphor, visual metaphor for people.

20:31But we had a rainbow this time.

20:33Oh, we had a rainbow, which means uh things are getting better, hopefully.

20:38So, yes, no, uh you know, to to sort of close that gap, I think

20:43There is a realization that these stores have some way to go and they can't just answer the questions.

20:49They've also got to solve the problems and then do the work, right?

20:52Like

20:53all three of those things need to need to fit in place.

20:55And so I think we are starting to see a bit of that in actual tabloids and there's an acknowledgement that like, right

21:01If we're gonna do this properly, we better build a freaking awesome semantic layer because none of this works without it.

21:06All right.

21:07And you know, all of that stuff starting to sort of come together.

21:10What you could do, Tim, is create a new semantic layer, or you could buy or partner with one, such as DBT.

21:17Exactly, exactly.

21:18So you

21:20Yeah, great segue, and let's give some context.

21:22So Tableau and DBT announced a partnership.

21:27And this is off the back of Coales, Coales being DBT's um annual conference.

21:33And um Tableau being um uh a a technology partner of DBT also had a blog post about this.

21:40Uh let me just let me just pull it up.

21:42Um

21:43But I think this is quite um interesting um because w I think what it's a it's a sign of is this thing that Tableau have been saying for some time.

21:53Which is that, you know, Tableau is a platform, right?

21:56And they've been saying that a lot, but what it's meant in the past is that Tableau itself is building platform like features uh and not really mean

22:06Tableau itself is becoming a platform that other things can plug into.

22:10And since VisQL data service, we've started to see the flip of this narrative, right?

22:16Tableau genuinely becoming a platform.

22:18Bringing people in.

22:19And that's what kind of led to this announcement.

22:22Yeah, so so you've already got an existing relationship with Databricks, which was announced I think either earlier this year or last year.

22:27You've got a

22:28a strategic relationship with Microsoft to an extent, which obviously serviced some of the newer features, integration with Teams, which we'll talk about in a bit.

22:36Um

22:38But in addition, you've got like all new features seem to come with an API, right?

22:43Yeah, exactly.

22:44So new features in Tableau have an API or something allowing it to be extensible to developers

22:51And I think the the data dev ambassadors in particular are gonna have a great time sort of evangelizing, exploring all of these new APIs and trying to educate and make sure that

23:01There's enough documentation and and sharing around the the cool stuff that's being done here.

23:06Exactly.

23:08I mean And I think this developer ecosystem, again thinking about exchange, thinking about Viz extensions, will just continue to grow and grow.

23:16And yeah, like I said, the it will just become a new big user base alongside your traditional dashboard designer and your tr your traditional server admin.

23:27This is now a significant portion of uh of the Tableau customer base.

23:33100%.

23:34Many ways Tableau's become the probably the more agnostic analytics platform compared to Power BI, which it

23:41Power BI really to get the most out of it requires you to lean into Azure and some of the other stuff, right?

23:46Um, but yeah, to touch on what was actually announced by DBT and Tableau, just as a quick highlight

23:52the ability to export DBT models and metrics straight into Tableau, create data sources without having to do do that stuff in Tableau, amazing.

24:01Lineage and data help all the way through.

24:03Um the interesting thing about some of this is some of it is already available in server.

24:07There's actually been a DBT extension on the Tableau extension gallery, but this is specifically about bringing it to cloud, which doesn't have that capability yet.

24:15Yeah, because that integration needs to be done by both of them working together.

24:19Um build visualizations import uh based on important DBT data models.

24:24So starting a workbook

24:26on a model in DBT rather than a connection.

24:29That's like pretty awesome.

24:31Um pulse metrics ex again being done directly connected to

24:36DBT model so you write your DBT metrics and then you ping the um API and pulse just generates those metrics automatically.

24:44And then better collaboration between people.

24:47So getting DBT into apps like Slack and Teams, we'll come to that in a second.

24:52Um, so people have a better experience throughout the platform.

24:55So really it just literally feels like someone just opened the guts of Tableau and said, hey, what can we do here?

25:00And Tabla's just like

25:01Anything.

25:03Yeah.

25:03Have it all.

25:05And that's really exciting.

25:08It's actually a very useful

25:10Direction to go in in terms of collaboration in terms of extensibility in terms of where the world's going in this space

25:17Well, you know, 2015, 2017, 2018, it's quite crowded.

25:21Like there's a lot of people doing a lot of things doing about the same thing.

25:24Yeah.

25:24And suddenly you have to map against them.

25:26But I think each each product in the Garden of Magic Quadrant.

25:31um had has its own thing now, right?

25:34Like so it's almost like you've got a really good broad charge of, well actually if you don't want to do that, this probably isn't the right tool for you, and that's okay because

25:44Um you have choices.

25:46Yeah, there's another way with something else.

25:49And you can bring your own stack to that as well.

25:52Yeah.

25:53So I think this is great.

25:54I think this this move moves moves Tableau on for sure.

25:57Yeah, and even the partnership with Microsoft to get teams um the same playing playing field as Slack.

26:05Um obviously Microsoft Teams is just not as great and I think I think I like I keep saying this, people laugh, but I genuinely think there is not many people who love Teams, if that makes sense.

26:17They just have to use Teams because

26:19It's kind of hoisted on their companies through the Microsoft partnerships, right?

26:22So it's also just not as stable.

26:25And I'm not saying Slack is better.

26:26Slack also has its problems.

26:28But the interesting thing here is because you have to use it for work.

26:31It's actually good now that Tableau is able to hook into it because it feels like this is a much bigger user base of this kind of application.

26:40And so in a weird way, Tableau will get probably more feedback from Microsoft Teams users about this kind of capability than they will get from Slack users because Slack is just not as embedded

26:51Ubiquitous enterprise uh companies as Slack.

26:55You have to be more or less a startup to be using Slack, you know, like have a have a certain opinion about these kinds of chat apps to use Slack.

27:13This is the problem you have when you're the first mover that everyone uses.

27:17Everyone use has e everyone's first computer is a Windows machine.

27:21Right.

27:22Therefore you don't love Microsoft Office.

27:25Right.

27:26But when you discover Google Docs, for example, you're like, oh, this is super cool.

27:29Yeah.

27:29Right.

27:31In a similar vein.

27:32No one picks teams.

27:35Like when you when you're like, oh, you know, let's hang out with friends.

27:37I'm not like, yeah, let's create some teams channels and hang out.

27:40Teams, yeah.

27:41You you might have a friend Slack channel.

27:43I know people that have had Slack channels just for just for pays.

27:47Um you'd use Discord, but like it's not a fun, different platform experience, right?

27:54It's very much a

27:56This is this is just it's just teams.

27:59It's just Skype shifted in to what a new product, right?

28:04I mean there was a world once upon a time where you had

28:06Teams Yammer and Skype for business under your saying Yammer.

28:11Gosh.

28:12Um Microsoft existence.

28:14But yeah, no, it's it's

28:15Or competing for it's all competing in the same space, which is bizarre.

28:19But no one loves Teams, you're right.

28:22But this this this truly continues Tableau's mission to bring analytics to where your users are

28:28Right, which is exactly what we're doing here with Databricks.

28:32It's exactly what we're doing here with DBT.

28:34Yeah.

28:35Let's not try and Databricks.

28:37Yeah.

28:37Yeah.

28:38Let's stop trying to create a tool that does notebooks.

28:41Let's stop trying to create a tool that is has a semantic layer that you can create models on.

28:46No.

28:47just let's just open up and eat where we're eating.

28:51Which I think we we see the ro we we will see the benefits from

28:55Because product manager and developer time will then be simply spent on developing table tableau features.

29:02Yeah.

29:03Leave the other stuff to the other companies you've partnered with and just start connecting people.

29:09Yeah, exactly.

29:10Yeah, and it's um it's an interesting one because I do feel generally the industry is gravitating towards

29:21Well, I think enterprise companies are gravitating towards more open more open platforms, i.

29:26e.

29:27, platforms where you can you can pick and mix

29:31What you want to do where.

29:32So if you want to do your metadata management in the platform, the platform can do it.

29:36But if you don't want to, you can go get something else, right?

29:38And that seems to be almost becoming the case with everything.

29:41You've got Viz extensions, you've got um uh table extensions, which are to do with the connections, Ravi, remember.

29:47And then we've got t uh tableau tables, which are

29:51coming up in in a bit.

29:53At pretty much every point of the platform now it seems like there is an external alternative um that allows you to sort of lean into whatever you want to give companies really the flexibility.

30:02I guess

30:03In a funny way, Tableau just want to make the product stick.

30:05So if that means you can get your your mitts into it and you know tangle yourself up in it, great.

30:11Like Tableau loves that.

30:13And

30:14I think this exists in sports technologies.

30:16When we look at sports technologies in in my line of work, one of the things we look at is, okay, cool, how how does it operate?

30:22Does it have an API that we can push data out of?

30:25And also where we need to, can we pull data in?

30:28And then when we're entering data, can we manipulate it?

30:32It has to do all three in order for it to be a sticky platform, as you said.

30:36So it's quite interesting.

30:42Exactly.

30:42Exactly.

30:43Shall we talk about tables?

30:45Yes.

30:45Yes.

30:46Twenty four three.

30:47Shall we start from the top?

30:49So twenty-four three just came out.

30:51Kinda came out really fast.

30:53I was not expecting it to come out in October.

30:54I thought more November.

30:56So I feel like now we do know the official dates that Tableau's going to be targeting for next year.

31:02They're very sort of equidistant, starting from about uh March.

31:06So you get March, April, May, June, July, August, um, September, October.

31:11And in that you have conference as well, which kind of sort of throws things kilter.

31:15But 24-3, final release of the year, um cloud release and desktop and prep and all that stuff.

31:21No server release.

31:22No server, nothing for me.

31:25We only had one update, that was the previous one, 24-2

31:28Um smooth upgrade by the way.

31:30J just as a server admin shout out.

31:35Yeah.

31:36Nice.

31:36Good.

31:37So 204T.

31:38So you're on 204T.

31:40Yeah?

31:40Yeah.

31:41You can use this nice new table extension because it's backwards compatible, Ravi.

31:45So

31:46You're not gonna miss out.

31:47This is actually quite a nice thing.

31:48And we'll come back.

31:49That's why we like this extensions.

31:51Yeah, yeah.

31:52We'll come back.

31:52Like let's let's sort of lean into this.

31:54So

31:5524.

31:563, I think biggest feature in the release, obviously it's going to be a desktop feature, table viz extension.

32:03What is it?

32:04It's essentially not the thing I thought was wrong, and it's not

32:09Anyway, not a table extension, which is a way to bring in um R and Python as a data source

32:18on top of your data inside of the connection window, inside of the data model or the joins, whatever.

32:24Which technically extends your table.

32:27Extends your table of data.

32:29Yes, exactly.

32:30It's like you can run a computation on a column and then have it sit alongside your data, basically.

32:35But you can do that in live query time, whatever.

32:39Table Viz extension, Tableau tables.

32:42I don't know why they use two different names.

32:45These bring, and I called it Excel-like capabilities.

32:50um two tables.

32:52So you can essentially add as many columns as you need, bring measures, dimensions in, format them, get them to work really nicely.

33:00Um I'm not going to go into it because I think you just have to see it to understand, you know, how it's different to a traditional table.

33:06But it very much answers the thing that

33:09You know, most end users want with these tables when they get them, which is the ability to manipulate them like Excel.

33:14So search them, filter them, all that jazz.

33:17Um now.

33:18And export SV.

33:20And export well, they export as Excel.

33:23They're not an SCSV.

33:24But but this is this, I think, nails the crux of why it exists.

33:30Yeah it's a hey Excel users, here you go.

33:33Yeah, exactly.

33:34Exactly.

33:35For those of you who love Excel and just want something familiar, here you go.

33:38And actually this brings me on to a very important point that I feel like I want to get off my chest

33:43I've had a couple of people respond to my video and just say, well, just use Excel then.

33:48And my gut response has been, it's a bit like saying to a car enthusiast,

33:52who takes a car to a dynamic to get a few more horsepower, just use a horse.

33:56It's like that's not what they were asking.

33:58Just add a horse to your car.

34:02It's like

34:03They were asking for this functionality within Tableau, hence the phrase Excel alike.

34:08Excel is a v Excel is an experience

34:11As well as a product.

34:12So an Excel-like thing.

34:14Exactly.

34:15Excel-like means can I do the things that I've learnt in Excel in the product?

34:19And that's essentially what this Viz extension does deliver us.

34:23I already we we've already had an internal discussion that there's so many use cases for this.

34:27Now, in order to do this sort of thing in Tableau right now, you have to use parameters, you have to use filters.

34:37You have LODs.

34:39Use LODs.

34:40Exactly.

34:41It's so unperformant.

34:44And bam, we've got physic extensions.

34:46Yeah.

34:47That make things performant again.

34:49Make data performant again.

34:51There's a slogan there.

34:53Bring back the cue.

34:54performance again.

34:55I don't know what that is.

34:56It's not MAGA, but it's not MAG.

35:00Sorry to bring part into it.

35:02Yeah.

35:02So close to the election.

35:03But you know what I will I will go out on a limb here and say

35:07Viz extensions, I think, are better than native extent native chart types.

35:13100%.

35:14They are just that when you're using them and experiencing them, they are just built different.

35:19I don't know how else to say that.

35:20They're sharp.

35:22Yes, yes, they really are.

35:24They're sharp.

35:25They're responsive.

35:26Like the tool tips are just crisper.

35:30And it what it what it sounds like is that they're easier to make than the native charts.

35:36Because for how long have people been asking for this kind of functionality native in Tableau?

35:41And Tableau's response has been actually like

35:44Well take a hell no to that.

35:46Instead, we'll put a whole nother thing that allows us to put whatever we want in here built on web technology instead.

35:53So it it also feels like a depart de like departing this world of VisQL, right?

36:03It feels like it's departed in this world.

36:05And instead, instead, and this is where you kind of have to bear with me here, this feels like a culmination in technologies.

36:14I don't think it's a coincidence that Viz extensions came out after VizQL data service.

36:19These extensions are being driven by VizQL data service off into the web world.

36:26And then the web capabilities.

36:28which are modern, being developed by thousands and thousands of people in the world today, are just running off with things.

36:34And that's why these are crisp and native and lovely.

36:37But they're still formatted to look like Tableau.

36:39That's easy to do

36:41And suddenly you've got this world, this cube-like world, where when you want exactly the thing you want, you can get it because the Viz extension will just do it for you.

36:51Yeah.

36:52I think there's so much potential here.

36:55And and most importantly, you want to get an update to that Viz extension, it will just update itself because it's hosted in the cloud.

37:03And it will just it will just update.

37:05It just gets better.

37:07The bugs are not tied to the release, which is even better.

37:10Like they can just get fixed.

37:12Now, there are a couple of issues.

37:13So the extension isn't perfect.

37:15There's a few things.

37:16And, you know, again I did another LinkedIn post where I was like, listen, we don't have to load these extensions with all the burden of everything that's ever come in the last decade.

37:25like printing them out, doing subscriptions, um, being able to get one-to-one PDF exports, being able to do page line all of these things have come out of the woodwork.

37:36And in many ways, I think it goes back to something that people will say is just because you can doesn't mean you should.

37:41Tables aren't the solution to everything.

37:43If you're using this to like just

37:50Sorry.

37:51There are other options available to people to to like go get this.

37:55This is just meant to bring an experience to the product.

37:58And don't forget, these work.

38:00on a dashboard.

38:01You can use dashboard parameter actions set actions on these to get them to do the awesome things that you're already possible.

38:07So these are just even better than tables in many ways.

38:10They are

38:10There are tables on you know steroids with the capabilities that Tableau has.

38:15So use those functionalities to get around these new problems, but it will be more performant than anything that came before it because this base capability has been built in.

38:24Now, security.

38:26Oh man, I can't tell you how many people say, oh, my company's not gonna take a look at this.

38:30And I think this goes back to conference, right?

38:32This goes back to

38:33Tamascoli's grateful stood up on stage and said, Look what I can do with a daisy.

38:38My kittens.

38:41And I think it's just it sort of escalated from there because the server community is very small.

38:45But understandably, Tabber haven't done a good job of actually answering the question.

38:49in lots of different ways.

38:50And this extension is network enabled, unlike the Sankey chart, which is Sandbox.

38:56So I think there is a little bit missing there to explain why is the Sankey chart built by Tableau Sandbox.

39:02But this table extension isn't sandbox.

39:05Like what is the difference between the two?

39:08And I think there is something there.

39:09Probably Tableau can't share that story.

39:11But there's something there that needs a little bit more exploration.

39:15Additionally, why is it the developers of other extensions can't build sandbox extensions yet?

39:20That's like a really important question to ask.

39:22They want to be able to.

39:23I've talked to Tristan.

39:24He says he'd love to be able to, but he can't at the moment.

39:27It's not a possibility.

39:29So I think answering these questions will get start to get people over the hill.

39:33But final, final point

39:35If you use Tableau Cloud, your Tableau's your customer, this extension is built by Tableau.

39:43I think the burden of getting this through your team should be zero because you're already using the platform.

39:50It's hosted in US East One.

39:52So if you are even on that Tableau pod, there shouldn't be any issue with data portability.

39:57I guess if you're in Europe or some other country and you use a Tableau Pod for that exact reason

40:02then maybe you can't use this extension.

40:04But for anyone who's new SE where this extension's hosted, also on Tableau Cloud, I just don't see the problem here.

40:10Like

40:12Especially with the trusted data layer.

40:14Exactly, exactly.

40:15It's all on the same platform.

40:17So, you know, I do think we're starting to get over the hill here.

40:20We're starting to overcome some of these barriers, but Tableau could do a bit more.

40:24to you know start to answer these questions more more more generally yeah and I think uh just just just on that point for a generic user

40:36Who just wants to be like, yeah, but I just want to be able to drive this in and put it out and start using it.

40:41Yeah.

40:42I don't think they see the steps required

40:44From an infospect perspective and all these things.

40:46Which is why cloud again, as as I think you said back way back in 2019 or 20, even early days of your channel.

40:53Like cloud is gonna be the single as cloud and way too early.

40:58I said about way too early.

40:59It's still not fully happened.

41:02It's not fully there yet, but it will be because that's the direction that they're pushing and pulling it, right?

41:07Yeah.

41:07Yeah, it's getting there.

41:08Yeah.

41:09The backward compatibility, the fact that I can use this Viz extension today is amazing.

41:15Right.

41:16Otherwise I'll be like, well, yeah, cool, but it's not for me.

41:19Like just imagine in the past, because the native extensions were dependent on releases, that wouldn't be the case.

41:26And it I think it's so easy to miss that point.

41:30Game changer, game changer in every way.

41:33I I I I would be happy if in 25-1 Tabot just said, boom, all charts of his exemptions.

41:44deal with it.

41:44Like I would I would.

41:49If the native ones were sandboxed, why would that be an issue?

41:55So they they actually are simply Viz extensions, but I don't know it.

42:01Yeah.

42:03They just run natively.

42:04They come pre-packaged with your product.

42:07Tableau loads up a piece of JavaScript.

42:09Ignore what I just said.

42:10If it's faster, I don't care.

42:11Make data performant again.

42:19Ultim ultimate ultimately, right?

42:20Like because again the the thing that hamstrings a lot of what Tableau does

42:26at speed is the fact it's VizQL.

42:28Yes.

42:29Like Render Man, okay, great, amazing.

42:33Like

42:33Yeah.

42:37And it still is really, really good.

42:39Right.

42:39It's still incredibly good as a

42:42experience.

42:43However, you peel back that hood and you do a performance recording and you're trying to work out why

42:51One of my dashboards within a eight no fifth thirteen page workbook is always bugging out

42:58Yeah.

42:58And I don't know whether it's a server-side problem or a dashboard problem.

43:02And I'm trying to investigate.

43:04And then you realize the depth of which a query has to occur.

43:10in the in the actual dashboard in order to cre in in order to visualize the one, two, three, four, five chart sheets I've got on this dashboard.

43:19Yeah.

43:20It's crazy.

43:21Where do you start?

43:22Where do you start?

43:23In a vi in a is it in a Viz Extension world, it's either the Viz Extension hasn't got the right data or there's too much on detail.

43:30But I've still not seen a Viz Extension be slow yet.

43:34Yes.

43:35Yeah.

43:36Touch wood.

43:39And I think additionally, um because the work isn't being done to compute in memory to visualize.

43:46The work is the same thing.

43:51The Marks Pain defines the aggregation and what should be generated as a table.

43:57The way to think of it is in the background there's a secret cross tab being built.

44:01That crosstab is then being shipped off to the extension for visualization.

44:05And at that point web technology takes over.

44:08Now reason but that's fast.

44:10But that is the same thing that happens with VizQL.

44:12You get a temporary b uh table

44:15Stored, but it's not a materialized crosstab that exists, which is why again it comes down to this is the new cube to an extent, right?

44:24Viscule

44:25Still has to then render tiles.

44:27It doesn't, it's not JavaScript.

44:29True.

44:29It renders tiles.

44:31It checks to see if that calculation to see if it can do it on device in browser or on server.

44:37If it can't do it on device.

44:39It ships that off to the server.

44:41The server renders the tiles and sends it back to you.

44:45By this time, the spinner's done four laps around the world and your verse is still not loaded

44:52So these Viz extensions don't need to do that.

44:55They get the data and they just process them there and then.

44:58Um I talked

45:00to um uh Jonathan Drummy a while back about this concept of tableau's frame rate, right?

45:06And how many of charts

45:08are like three frames per second.

45:11If you actually try and track an update, it's just so slow.

45:15But these Viz extensions, they feel like the ones built on whatever they're built on now.

45:20Like if you have a high refresh rate screen, let's say 60 FPS above, like you can literally see that the chart, like the thing tracks your mouse like faster than your eyes can

45:31perceive it's it's just another it's just another experience.

45:34I know maybe we're we're going into this too too deep.

45:37We love extensions is the short is the short bit

45:41We can't wait for more.

45:43An another example of a platform that's fully extensible.

45:47Yeah, exactly.

45:48I can't wait for someone like Matbox to drop a Matbox Viz extension.

45:52that just brings Mapbox into Tableau and just lets you kind of set up your map inside of the Tableau V-Pane.

45:59That would be awesome, right?

46:00Like

46:00That's the kind of stuff I'm thinking.

46:02Just give it a few more years and you know, you know Uber when used to go to their website used to have that map of the world, right?

46:08That was very sort of

46:09uh stylistic.

46:10Imagine like an Uber map uh as a Viz extension for visualizing, I don't know, Uber taxi rides or something like natively built for tablet.

46:20I think the best Viz Extensions will never grace the Extension Gallery, because they'll be built for internal use

46:27to do awesome things that we will just never see.

46:30And I I genuinely think that is actually what's gonna be the best physics engines will never be something anyone will see publicly.

46:38They'll sit within a company doing incredible things and you'll see a tablet dashboard and you'll be like, how on earth is that being done?

46:44And it'll be this, it'll be this.

46:46Completely agree.

46:47Completely agree.

46:49Yeah.

46:49Um and I hope to be a beneficiary of that world.

46:55Absolutely.

46:56Just make it happen.

46:57You'll be there.

46:57You'll be there.

46:58You'll be there.

46:59Um any other parts of this release?

47:01Obviously, there's cloud manager.

47:03I think that's quite an interesting concept.

47:05It's not gonna pull me to cloud, Tim.

47:06Don't worry.

47:07It's not gonna pull you to cloud.

47:08But I do want to touch on something here, which is I think there's um

47:12There's this idea that one Tableau Cloud website is one customer.

47:18And actually there's so many instances where where

47:22Actually that's not the case.

47:24Um and the way to think of this is that look Tableau Cloud website is a site in the Tableau Server terminology.

47:32And the site in the Tableau Server terminology allows you to separate resources, separate space, and separate access.

47:40So things like authentication.

47:42It's essentially like a firewall between two spaces on the same Tableau ecosystem.

47:47And so you can use this for lots of different devices, for example, regulation, compliance, um uh cost centers in some cases, right?

47:56But then you can also do it to split business units.

47:59You can do it to split uh different operational functions within one organization.

48:04You could split companies that are owned by a parent company using this mechanism.

48:09There's so many more

48:10And so I can't really give more details than this, but I want you to try and guess, Rabbi.

48:17How many companies do you think have more than

48:24Let's say ten Tableau Cloud instances.

48:30Not sites instances.

48:33Before this feature there would have been instances, yeah.

48:40Over a hundred.

48:41Okay, I I've known friend wrestlers, that's a big number or a small number.

48:45Right.

48:46To give you some context.

48:49That is off by a country mile and in the wrong direction.

48:54It's really way more than that.

48:56It's way more than that.

48:57And I can't I can't give this metric, but

49:01Though there is a scenario where we're talking about triple figures numbers of customers that have, on average, well above 30 cloud instances.

49:13To one organizer.

49:14Well above.

49:15Like well above.

49:16Now, if I think about my time working in the consulting world and seeing the number of departments and mini departments and instances where you're just sat there like

49:26Why is this not one thing?

49:27And then Tableau's sitting there like, ha ha ha ha, just counting the money they're making of all these different departments and their various budgets where they've just spun off Tableau site after Tableau site after Tableau site.

49:38Yeah

49:39So this feature, although it's pretty much a nothing feature to most people because we're not admins, we're not or whatever, if you're listening to this and going, what's the big odds about this?

49:48Well, what you should do is you and if you if you're using Table Cloud, you should go tell your admin and make sure they know about this.

49:54Because typically it's not even your admins for cloud who know this.

49:58It's the in IT teams.

50:00in your organizations that are out there buying separate cloud instances for separate teams.

50:05And they'll also be the ones complaining to Tableau about why on earth am I having 20 different invoices.

50:11Can I just group them into one?

50:13And with this, the answer is yes, you can, because Cloud Manager allows you to manage your licenses under one roof and distribute those licenses to different cloud accounts.

50:23Which can also be different regions under the same uh organization account.

50:29It's a bit like in Snowflake where you have

50:31an organization at the very top is a new concept, then you have accounts that belong to an organization.

50:37It's very similar to that.

50:38So um

50:41Very dry feature, but I also just wanted to make sure I sort of drive it.

50:45Give it some give it some low I I was

50:48I was blown away by the number of companies that have more than 50 cloud instances to one company.

50:56It's actually wild.

50:57And that's not even close to the actual number.

50:59I'm deliberately keeping distance from the number for for various reasons.

51:03So it's definitely a um uh yeah, uh a very interesting thing.

51:08Now

51:09Other things to note, Tableau Pulse got a whole bootleg of releases.

51:14I'm not even going to go into each and every one of them.

51:17Tableau Pulse, uh Tableau Cloud went to Hyperforce.

51:21I still don't really know what this is other than to say I think it's just Salesforce's cloud instance of

51:26That's exactly what I think.

51:27Everything.

51:28Um so instead of Tableau cloud sitting somewhere else, it now sits on the same cloud as MuleSoft, blah blah blah, trailhead, whatever.

51:35So it's all under one roof.

51:38What else was there?

51:40Obviously, Microsoft um spatial parameters.

51:43Spatial parameters.

51:44Oh yeah, yeah.

51:45With spatial parameters based on WKT uh

51:51uh spatial data thing well known text like who came up with that name um

51:58Interesting feature, so many details in there that I just didn't know about, and now I've done the video and I've I've I've I've been, you know

52:06The maps team has flooded me with lots of detail.

52:09I wish they'd given me before I'd done the video.

52:13It's like this is wrong.

52:14Like for example, Ravi, what's a compact list?

52:18A small list.

52:20What what is it in Tableau?

52:22Oh right, it's it's where you're No, I don't know what you're talking about

52:25You said compact list.

52:27It's a drop-down list.

52:28A compact list is a drop-down list.

52:30It's another name for a drop-down list.

52:32Okay.

52:32When you're mapping, it's called a compact list.

52:34And I was doing a video and I was literally hovering over the term compact list and I was saying, why do we not have a drop-down list?

52:42And I'm hovering over compact list.

52:44So someone from the uh architecture team sort of got in touch and said, Actually we do, it's uh called a compact list.

52:49And I was like, I just replied, Well, why is it called the compact list?

52:52And

52:52They just replied, Well, it's always been that way.

52:55And I'm like, I didn't even know that.

52:59If you open Tableau, you go try and d like create a parameter or something with anything

53:06I I I can't remember the exact context, but it's always been compactless for spatial.

53:10It's just so bizarre.

53:12So so bizarre.

53:13Anyway.

53:13But I also remember that Tableau didn't have an apply button.

53:16And I think Andy Krubel definitely takes

53:19Credit for that being like f for the multi for select there there was an apply it's like put an apply button in.

53:27I put the apply button in every multi-select, so thank you, Andy.

53:31But it it's small things like that where you like

53:33These are just legacy features that people are like didn't realize people wanted that.

53:38Yeah, yeah, yeah.

53:39Now, there is one th final thing with this thing.

53:42Do not load

53:44really complex polygons into it because you will you'll just freeze your computer.

53:49And this feels like an a a trap waiting to catch people out.

53:53But in essence, there's two ways you can use this.

53:57Using parameter actions and using a list of list of spatial objects.

54:03Thank you, Rabbi.

54:03A compact list.

54:06If you use a compact list and you try and load into that list polygons from shapefiles, you are likely to probably freeze Tableau because what it has to do is take each and every point on that polygon and process that.

54:20Into the parameter, which takes forever if you've got, let's say, the polygon for the United Kingdom, which I don't know how many data points it has.

54:29So what you should do is you should generalize the polygon in something like altrix.

54:33Bring it down from thousands of tens of thousands of points down to maybe uh a hundred, that will massively make that faster and make it easier to do.

54:41Then it will work.

54:42Or use the um the parameter action.

54:45So something else feeds into that parameter.

54:47It's just feeding one item in.

54:50I've got a third option.

54:51Yeah, go for it.

54:52Physics tension.

54:54Yeah.

54:56Yeah.

54:57Spizz extension for everything.

54:59Well the the you say this and you know I am surely make the most sense for his biz extension.

55:08Well, super tables by InfoTopics, right?

55:11That to me is a good example of an extension that has its own little world of capabilities.

55:17Built into the extension that don't need anything else like parameter actions or whatever.

55:22So there is nothing to say that you couldn't build an extension where within that chart

55:27You could build in custom interactions, right?

55:30Like a correct network chart or like something like where you click and drag and it changes the relationship.

55:35You know, all that kind of stuff is technically possible.

55:37So yeah, Viz extensions could absolutely be the answer.

55:41Yeah, there's lots of interesting things um that I've spoken to a few people about.

55:46Um where once you start peeling back the onion, you start to realize how much interesting things

55:53you can end up doing with a bit of big ascension.

55:56And one such thing um would be like a dialogue box, just creating a custom filter control set.

56:04That is just your own way of creating parameters.

56:07Because that if your entire dashboard is physics tensions, can talk to every single one of them, but in in the same ecosystem.

56:15And that is Baller.

56:17Yeah.

56:18Yeah.

56:19Yeah.

56:19And and if you don't like the way that Tableau does filtering or sorting or parameters, right?

56:24Like I wish I could do an actual multi-select parameter.

56:27Build it as a visit extension.

56:29Yeah.

56:30That's essentially the power that Tableau have handed people.

56:32Like instead of instead of complaining about it, build it.

56:35Like if you really need it and you value it, build it.

56:37Become a data dev.

56:39Yeah.

56:39I actually think there's not enough of them.

56:41Um Correct.

56:42Yeah, and this is why like I think this is this is about to become

56:47Uh not a problem for Tableau, but you can tell when they pitch a lot of these ideas

56:55they haven't got the same sort of richness as as they get when it comes about pitching other things that customers are coming to them flooding with.

57:04So

57:04And the kind of companies that would be building these kind of awesome things are not really bought into Tableau.

57:10They kind of, you know, they kind of take not a good look it's it's a classic, you know

57:15Salesforce tableau dunking you get from like Silicon Valley and like I will never touch since Salesforce and like, you know

57:23Like, you know, people laughing at like, oh my god, you look at how slow it is.

57:27It's not even built in the open web.

57:32You know, this kind of stuff.

57:34And I was like, yeah, but it is it is enterprise software.

57:38It's gotta go through hundred and fifty different things, like

57:43authentication and it's got to be built to a certain standard that you would expect.

57:47So yeah, this is just software.

57:49This is just how things work.

57:50Sorry.

57:51But anyway.

57:52It's just software.

57:53That's just how things are.

57:54What a place to end it.

57:55What a place to end it.

57:56Absolutely.

57:57Absolutely.

57:57I think it's been an interesting couple of months.

57:59So much has happened.

58:00I'm looking forward to the end of the year.

58:04Um

58:05Which also means, Ravi, we need to actually get on the case with our little survey that we're gonna do, right?

58:11Um and it won't be a survey per se, like a big survey, but we are gonna try and reach out to

58:16Um, some let's say established members of the community, but also new members of the community, try and gauge their pain on different parts of the product so we can build I think a picture of what 2024 has been like.

58:27Um so we can kind of get ahead of the Gartner report and actually give people some sort of perspective on what it's like to use the product rather than what it's like to purchase the product.

58:36For sure.

58:37If that makes sense.

58:37So yeah, we're gonna be getting on that and sort of putting more stuff out.

58:40So look at that.

58:46Indeed.

58:46So look out for us there.

58:48If you are attending, just come and say hello.

58:51Absolutely.

58:52Absolutely.

58:53I don't know what we're gonna actually do around this podcast, but we'll we'll do something.

58:57Roving mics.

58:58We'll just do a five minute walk around the the the the conference floor together.

59:02We can just have a chat after the keynote.

59:04Let's just like do like um that that's so much easier.

59:08I'll bring my my

59:10Just clip a mask and we can just straight off the literally immediate hot takes.

59:15Ummediate hot takes after the conference.

59:19As we're walking out, we can just talk

59:21Yeah.

59:21Just capture what we say.

59:22And then we can do some uh post-conference edit.

59:24That's probably what we should just do the whole conference.

59:27Just record little snippets and then piece it together into a podcast.

59:30Yeah.

59:31No, I I think that's uh that's uh I I'm I'm the sort of guy that hates roving mics.

59:36Uh 'cause it's kinda like you're not getting the you're not capturing the essence, you're just asking questions.

59:40Yeah, when I say moving mics, I'm talking about these um Oh no, oh sorry, lapel mics, fine.

59:47Yeah, exactly.

59:47I thought you meant like, hello, I am Tim.

59:52Clip it in, switch it on, let's just talk.

59:57Yeah, good.

59:58Good.

59:58Good.

59:59Right.

59:59We'll call it there.

60:01See you soon, Robbie.

60:02Take care.