0:00Hey, it's Tim here. I wanted to ask you a
0:01question, how should the Tableau release
0:03features?
0:04Should they release them when they are
0:06feature complete? Or should they release
0:07them in bits as
0:08they have been doing for the last few years
0:10, until they get to a point where the
0:12feature is more
0:13robust? My response to this might be a
0:15little bit unpopular. To get into that, let
0:18's get stuck in.
0:18Now to go through this, I'm going to show
0:20you the Tableau release navigate, it's
0:23actually the best
0:23single place to find out everything new in
0:26Tableau as a link to resources and terms of
0:28documentation, but also just as a list, it
0:30's everything is here that I can think of.
0:32And it
0:33sometimes has a bit of lag, they're sort of
0:34a little bit slow to update it. But
0:36nonetheless,
0:36I think it's an honest reflection of what's
0:38actually going on with features. And one of
0:40the
0:40nice things you can do is you can choose a
0:42product, you can filter based on the
0:44version. And I've
0:45actually done a video showing you all of
0:47this capability already. You can enable and
0:49disable
0:50different add ons. And you can even see the
0:52different names to them as they've changed
0:53over
0:54time. And you can also see features that
0:56have been changed, deprecated, new or
0:58retired. So you can
0:59actually just go and get a laser focus view
1:01on things that are changing in your
1:03platform, but
1:03you might not be aware of anyway, how are
1:05we going to use this? Well, I want to come
1:07in here and
1:07search for a specific feature. And we're
1:09going to see the history of that feature.
1:11So let's go ahead
1:12and type in virtual connections. This has
1:15been a area that Tableau have been updating
1:18over time.
1:19And you can see sort of my point, if we
1:21actually go back, you can see that the
1:24first set of
1:24capabilities really go all the way back to
1:27let's say 2022.1. So they've been adding
1:31new virtual
1:32connectors support as we've been going on,
1:34what I might do is just remove my phrase,
1:36because I think
1:36I just needed to say virtual. And if I say
1:40connections, which is the terminology they
1:44've
1:45used, I might get a better representative.
1:47Yeah, there we go. That's a little bit
1:48better, I get a
1:49few more hits. So it's a little bit of a
1:51reflection of what's going on. So you can
1:54see here clearly
1:54that depending on the product, you've got
1:57lots of versions spanning lots of different
1:59times. And so
2:00my general point here is this, should Table
2:03au in the case of Tableau cloud, really have
2:06not
2:06released anything for virtual connections
2:09until we got to let's say, June 2023, when
2:12virtual
2:12connector support for a whole bunch of
2:14additional connectors came in, came in, I
2:16think in 23.3,
2:17they're saying that all the connects are
2:19going to be available, or is it actually
2:21good that we got
2:22the features sort of over time, and as of
2:24those features have sort of been brought in
2:27, we've been
2:27able to build up an understanding of how
2:29the feature works and how it all sets in.
2:32And if you
2:33look at this sort of duration, you can say
2:35that it's broadly over the space of a year
2:37and a half,
2:37virtual connections have started to become
2:39a much bigger deal and actually go back a
2:41little bit
2:42further. I think because this data source
2:43is not thorough enough, it's not actually
2:45going back to
2:462022.1, which is when these actually
2:48initially started to come in, you've even
2:50got some from 21.4
2:52where you could use virtual connections in
2:54your flow. So the feature set goes back and
2:57forth,
2:57this is a little bit tainted. But you get
2:59my point, those features are sporadic over
3:01time.
3:02And if you're on Tableau server, obviously,
3:03you get them every other release that was a
3:05change
3:05that Tableau also made in the past. And so
3:08here's my point and general sort of issue
3:11with this,
3:11this is great, because I think it means
3:13that if you're building something and a
3:15specific feature
3:15helps you in your specific setup, then you
3:18can actually sort of get a capability that
3:20adds
3:21enhancements to whatever solution you're
3:22trying to build. The problem though, and
3:24this is my theory,
3:26shoot me down if I'm wrong. I think
3:28businesses don't update as fast as Tableau
3:30's own release
3:31cadence. You see, in reality, if you're
3:33using Tableau server, which is one group of
3:35customers,
3:36you upgrade server probably once a year to
3:38a net new version. By that I mean, you're
3:40going from
3:40let's say 23.1 to 23.4, maybe, and that
3:44might be your sort of jump. But the patches
3:47in between those
3:48get fixed. So you want to do patches,
3:50because those fix problems and bugs inside
3:52of your software.
3:53But in reality, between those versions,
3:55what ends up happening is whatever you had
3:58to start with,
3:59is what you sort of use to form your
4:01business practice. That's what you do to
4:03train teams.
4:04If you're running a server with any sort of
4:05decent scale, then you're going to also
4:07have
4:07to communicate these changes, figure out
4:09best practices, test them, make sure they
4:11're rigorously
4:12gone through all the checks that you'd have
4:14to do inside of your own business. And of
4:16course,
4:17Tableau is not the only tool. Tableau then
4:19interfaces with other tools. And you have
4:20to
4:21figure out all the kinks that are related
4:22to that. By the time you go through that
4:24whole process,
4:24I think easily four to five months rolls by
4:27. And then you get to sort of halfway
4:29through your
4:29release cycle. And you're already thinking
4:31about the next release. And then that
4:32release comes
4:33around. And so if one of these features isn
4:35't sort of doing exactly what you want it to
4:37do,
4:37it ends up sort of just missing that
4:39release cycle altogether. And then what
4:41ends up happening is you
4:42actually wait until the next major release
4:43until it gets to a point where you can
4:45actually use it,
4:46then that's when people really sort of get
4:48involved. And if I go look at another set
4:50of
4:50features, if I look at something, let's
4:52look at something like ask data, which has
4:55had a lot of
4:55sort of energy and emphasis over time, this
4:57is actually a great example, because you
4:58can see here
4:59in desktop, at least, it's been going on
5:01for some time, I mean, I can't even, I have
5:04to move my sort
5:06of mouse and cursor out the way here. So if
5:08I just scroll down here, you can see the
5:10ass data just
5:11had a ton of trickle, trickle, trickle,
5:14sort of sets of features all the way
5:16through. So it's
5:17really desktop and server. In this
5:19particular case, I don't know where the
5:21cloud information
5:22has gone from this data set just seems to
5:24be a mess. But I think the story is fairly
5:26accurate
5:26here for desktop. So we start in 2019. And
5:30we go all the way to this evening
5:32chronological
5:332022 21.4. That doesn't seem right. And it
5:36definitely goes on a lot more recently,
5:38because
5:39our status also had enhancement more
5:40recently as well, I can definitely tell you
5:42that here we go
5:4322.3, our state lenses appear in lineage.
5:46So the features sort of been growing up
5:48over time. But
5:49what would have happened here is that you
5:52wouldn't have been able to really kind of
5:54deploy this and
5:55be happy and get sort of comfortable up
5:57until a certain point. And I can't say what
6:00that is for
6:01our state. But I guarantee that when our
6:04state of lenses came about a lot of people
6:06who were using
6:07our state, if anyone was using it, had to
6:10sort of change their whole workflow and get
6:13that all set
6:13up. And again, if you're building practices
6:15around this stuff, it just takes you time
6:17and releases.
6:18And you might be waiting on some of these
6:20enhancements that come later on to sort of
6:22come about. If we go and look at something
6:24like the data model, not many items in here
6:29,
6:30you can see here, you can swap the root
6:31table, and it doesn't go back far enough to
6:33sort of give us
6:33the full picture. And everything here is a
6:36net new feature. But if we if we just look
6:39at that as a
6:39sort of feature set, it's interesting,
6:41because data model, I think, got released
6:43over maybe two
6:44years and sort of all the big update. And
6:46there's still some things that we've seen
6:48at conference
6:48that are yet to come. And should something
6:51like that have come about maybe in year
6:53long sort of
6:54phases. So in year one, you get a big bunch
6:56of features, then in year two, you get
6:58another set
6:58of features, or is actually good that every
7:00quarter we're getting some sort of
7:01enhancement.
7:02And so my, my opinion that I think is
7:05slightly unpopular is I think tableau has
7:07gotten to the
7:08scale and size as a piece of software,
7:11where it just needs to launch feature
7:14complete features.
7:15Actually, I think more damage is done when
7:17the features are released, and they're not
7:20quite
7:20hitting the mark. And there's some sort of
7:22additional tweak or additional perk. And
7:24that perk
7:25or tweak has to make it through the whole
7:26development cycle to sort of make it
7:28through into
7:29software. And by then businesses have moved
7:31on, they find new ways of working, or it's
7:32no longer
7:33sort of a timely thing for them to
7:35investigate. Also, I think if you think
7:37about this, if tableau
7:39release features that are more feature
7:41complete, what it actually means is that
7:42when those features
7:43come out, there's more of a bigger deal
7:45around that thing. I used to remember way
7:47back when I've
7:48actually been doing some research about the
7:5020 years of tableau. And I've got some
7:52really amazing
7:53content, some old release documentation
7:55from all the way back to version number one
7:58. And what's
7:59fascinating is looking back over those
8:01releases going even as far back as let's
8:03say 2020. The
8:05releases really had big deal features in
8:07their features that were sort of statement
8:09features
8:10that the releases used to go by,
8:12specifically one or two features. And now
8:14it just feels like there's
8:15there's lots of small things, but there's
8:17no real sort of big hit, as I like to call
8:19them things
8:20that really get people excited. Instead,
8:22you have these general sort of movements
8:24and directions,
8:24and they're sort of themes rather than
8:26specific things that people are excited
8:28about. So
8:29I think what that does is also means is
8:30less excitement around the features. And so
8:33I don't
8:33know, what do you think? Should tableau
8:36release features when they are done? Should
8:38they release
8:38them in bigger chunks? Or should they
8:41release them as they are every single
8:42quarter when they
8:43have something ready, ship them straight
8:45away? That's what I want to know from you.
8:47Let me know
8:47in the comments below and I'll catch you in
8:50the next one.
8:51Bye.
9:01[ Silence ]
Tableau | Tool strategy | Industry trends