Thursday, November 10, 2016

TC16: My Favourite Tableau Conference 2016 sessions



Well, TC16 -- my very first Tableau conference -- is now officially over.

As a bit of an introvert, I frankly found the sheer size of the conference -- and the crazy Data Night Out party -- a little bit intimidating. Before now, the biggest conference I'd ever attended was NICAR at 1,000 attendees. This conference had 13,000 people.

But I also learned a lot and got to meet a lot of people who, before now, I only knew through Twitter.

I also attended some great sessions, which I thought I'd note here since Tableau is going to make recordings of all the sessions available in the coming days.

I'm also hoping others might share what their favourite sessions were -- either on Twitter or in the comments -- so I have a cheat sheet when I start making my way through the hundreds of recordings.

So, in no particular order, here's a list of my favourite sessions, with a couple of notes on each:

50 Tips in 50 Minute with Andy Kriebel and Jeffrey Shaffer.

Exactly what it says on the tin: a load of great Tableau tips in rapid-fire succession. I love sessions like this as, even if one tip isn't helpful to you, the next one will be. This one will require re-watching to catch some of the specifics of how to implement each tip. But I easily came away from this session with dozens of time-saving tricks and ways to make my work in Tableau better.


I was hoping to make the similar-titled Rapid Fire Tips & Tricks with Daniel Hom and Dustin Smith but it was the only session of the conference I was turned away from because of lack of space. Will definitely be watching that one on video.


The Visual Design Tricks Behind Great Dashboard
s with Andy Cotgreave

I was told Andy Cotgreave's sessions were not to be missed and that was good advice. This was a great conceptual talk about how to think about ways to make your Tableau Dashboards more engaging and easier to read.

A lot of Andy's advice in the talk is similar to what I tell my students (like making sure your title actually says something interesting). But there was also a lot of advice that hadn't occurred to me that I can put into use. And he had a fun Few-McCandless data viz continuum with Alberto Cairo right in the sweet spot.



Visualizing Survey Data 2.0 with Steve Wexler

Steve Wexler has published a heap of great resources on how to visualize survey data in Tableau and this talk had some really useful updates on some new tricks he's developed -- including using a "Dual Pivot" to allow you to visualize demographic data more quickly and how to deal with situations where you have too few respondents to a given question. Great stuff.


Sealed with a KISS-Embracing Simplicity in Data Visualization with Chris Love

At a conference where a lot of people were showing off all sorts of intricate, complicated graphics, Chris Love's talk was a helpful reminder that the simplest charts can sometimes be the most effective. In one of the talk's more powerful moments, Chris took a beautiful, but hard to read, Sankey diagram and remade it live as a series of simple bar charts that actually told the story of the data much more clearly.


I've asked Chris to please do more of these "Simple Makeovers", starting with a complicated Guardian chart he showed during his talk. He seems game, which I think would help more people see the value of keeping things simple. UPDATE (Nov. 13): True to his word, Chris has already updated the Guardian chart as a small multiple hex map! And then as a second version, too!

Advanced Mark Types: Going Beyond Bars and Lines with Ben Neville and Kevin Taylor

This is kind of the anti-talk to Chris Love's presentation. Ben and Kevin went through several cool chart types -- like lollipop charts and hex tile maps -- that aren't in Tableau's built-in "Show Me" menu but can be created with a bit of fiddling in Tableau. Most of the chart types were actually useful, rather than just being show-offy -- and, in some cases, they looked pretty easy to implement. I know I'm planning to use lollipop charts a lot more in my work now.


Cross Database Joins: The Unexpected Solution to Tough Analytic Problems with Alex Ross and Bethany Lyons

This was the one and only "Jedi" session I attended and, I'll confess, I only went because I got turned away from the Rapid Fire Tips session I mentioned earlier.

The material in this session went by really quickly and a lot of it was over my head. But it's a testament to Bethany Lyons' infectious enthusiasm that this session made me want to learn more about how I could "create more data" use cross-database joins to solve gnarly data problems. And she did such a good job of explaining what she was doing that I feel I actually got the conceptual gist of this talk even if I'll have to re-watch the talk in slo-mo to get all the steps.

In the TC16 preview podcast I mentioned in an earlier post, pretty much everyone was raving about Bethany as their favourite speaker and now I can see why. (Alex Ross also did a great job summarizing the key concepts.)

Unfortunately, there were a lot of sessions I was hoping to make but didn't, because I was tied up in hands-on training, like Busting the DataViz Myths with Matt Francis, Data Journalism: Creating Awesome News Graphics in Tableau with Robert Kosara and New Ways to Visualize Time with Andy Cotgreave.

Speaking of hands-on training, I was really impressed with the calibre of the instructors in all the hands-on training sessions I attended and the quality of the materials (including, for each session, a "web workbook" that you can refer back to at your own speed with all the problems and solutions in it).

So those are my favourite sessions from TC16. What are yours? Please let me know in the comments below or by sending me a note on Twitter.

Note: Tableau is paying for some of my conference-related expenses.

Wednesday, November 9, 2016

My 5 favourite new Tableau features

So I was a bit distracted with a certain political event last night, so didn't get a chance to post this yesterday as I'd originally planned.

But I wanted to write up a quick post on some of the cool new features coming to Tableau.

There were two big keynotes yesterday: Tableau Vision which laid out some of Tableau's big plans over the next three years, and Developers on Stage, in which Tableau's developers introduced some cool new features coming in the next few months.

Some of the features announced during the Tableau Vision talk were pretty cool, but it's hard to know how long it will take before they show up in a release. So I'll spend most of this post talking about the features that are coming soon to Tableau (in some cases, it sounds like, as early as Version 10.2).

Here are my Top 5 favourite new features.

#1: Support for Shape Files

I've been wanting Tableau to support spatial files, like SHP and KML files, since I started using the product six years ago and I'm so glad it's finally here. There are some hacky workarounds now for getting spatial data into Tableau. But they're pretty clunky, complicated to use and prone to error. Being able to just connect to a SHP file, like you can now to an Excel or CSV file, will be a huge improvement and make Tableau a much more powerful, and popular, mapping tool.

#2: PDF Connector

Getting data out of PDFs can be a huge headache and, until recently, was almost impossible. I remember not too long ago spending hours fiddling with command-line tools like pdftotext to try to get data into a spreadsheet.

Luckily, the tools available to extract data from PDFs have gotten a lot better in just the past few years, chief among them Tabula. But while Tabula is well known among data journalists, most people I come across haven't heard of it (or the nearly as good Cometdocs).

So building PDF extraction right into Tableau will, I think, make that data a lot more accessible to a lot more people and -- if it works as good as Tabula -- it will be quicker to do it right in Tableau than having to fire up another tool.



#3 Expressive Text Editor

Since I started using Tableau, I've wanted the ability to simply highlight a word in a text box and hyperlink it to a web page. Like with Shape files, there are hacky workarounds now: you can create a little Sheet and then use a URL Action to make clicking on it open a web page. But it's not very natural. And it doesn't allow you to, say, have a normal title and description on your Dashboard and hyperlink a single word. That's now coming to Tableau, along with a bunch of other features that will provide more flexibility to any text on a page (including captions and tooltips), such as dropping images into a text box and kerning text.


#4: Better Dashboard Formatting

A number of little improvements to Dashboard creation were announced yesterday, including the ability to add margins and padding to a Dashboard and to evenly distribute Sheets with a single click. It doesn't sound like a big deal, but as the features were demoed it became clear that, along with the Expressive Text Editor, these features should make it a lot easier to make cleaner, nicer looking Dashboards with a lot less finicky formatting.



#5: Tooltip Selection

Another neat feature that could be quite powerful: you'll be able to put links in a Tableau tooltip that, when a user clicks on them, will highlight related elements on your viz. This could allow for some neat discoverability without having to always put in separate filters or highlighters.

So those are my five favourite features coming to Tableau soon.

But, as mentioned, in the morning Tableau Vision keynote, the company also unveiled some of the things they're working on over the next two or three years. I'm not sure how excited to get about these features, as they could still be a ways out. And while I wasn't at last year's conference, it sounds like some features that were previewed then still haven't arrived (like charts in tooltips), so these things can take time

Still, some of what was shown at Tableau Vision was pretty cool and I think it's worth a mention.

Project Maestro is probably the thing I'm most excited about. Tableau has already taken a lot of data prep you had to do outside Tableau with OpenRefine or Alteryx and baked it right into the product with things like Pivots, Unions and Cross-Database Joins.

Maestro seems to be an attempt to go even further in that direction, with really powerful tools for reshaping your data. Maestro will also include the ability to use visual cues to join datasets together and be alerted to mis-matches in your data, such as highlighting join errors in red so you can quickly correct them.


Also interesting: Selection Summaries, which will give you little pop-up visualizations based on the marks you've hovered over or selected. This seems like a pretty cool way to get a drill-down view of your data quickly without having to build out an entirely separate Sheet on your Dashboard.

The Tableau Vision keynote also showed off some of the work they're doing on Natural Language Processing. The idea here is that you could ask a question of a Dashboard much like you'd ask a question of Siri on your iPhone. You could type "Show me the most expensive houses in Vancouver" in a text box and the map would interpret that query and change the view of the data.

If/when this works, it would be pretty neat. I'm a bit skeptical NLP could be smart enough to be reliable. But, already in the demo, Tableau has built in little sliders below the query box that show the user how their text query has been translated into data filters. That seems to me to be a pretty good way of being transparent about how the feature is working.


My only complaint about some of these new features is that they didn't come sooner (especially support for Shape Files and hyperlinking text). But I'm glad they're here now and they will make my personal experience using Tableau a lot more enjoyable and powerful.

Note: Tableau is paying for some of my conference-related expenses.

Thursday, November 3, 2016

TC16 Preview: What I'm looking forward to at my first Tableau conference



Next week I'll be boarding a flight to Austin, Texas to attend my very first Tableau Conference.

In many ways, I can't believe it's taken me this long to attend.

To say I'm a fan of Tableau is a bit of an understatement.


While it's hard to pin down the specific date, as near as I can tell I've been making charts and maps in Tableau for more than six years (I think this was one of my first, which accompanied this 2010 Vancouver Sun series). Since then, I've created more than 100 Tableau Public charts. I do regular Tableau training workshops and Tableau is a big part of my Data Visualization course at Kwantlen Polytechnic University.

So why have I never been to "conference" before?

In part, it's because my conference schedule was already pretty maxed out.

With two young children, I don't like to be on the road too much.

And, as a data journalist, over the past few years I've tried to make it to both the NICAR data journalism conference and Tapestry, the great little data storytelling conference put on each year by Tableau the day before NICAR begins.

Tableau often hosts several training workshops at NICAR, and a lot of the Tableau Public folks attend, so going to NICAR felt like I was getting my Tableau fix without having to go to a whole separate conference.

However, since leaving The Vancouver Sun last year, Tableau has become an even bigger part of my life. I'm still doing data journalism and teaching the next generation of data journalists at Kwantlen. But I'm also doing a lot more non-journalism data visualization consulting and training -- and almost all of that has been using Tableau.

Plus, with NICAR and Tapestry in Florida this year -- about as far from Vancouver as you can get and still be in North America -- I'm probably not going to make it to them this year.

So it felt like it was finally time to see what Tableau Conference is all about.

As I prepare for the conference, the one thing I still have trouble getting my head around is how freaking big Tableau Conference is.

Tapestry is probably my favourite conference (and I'm not just saying that because I got to give a short talk at it in 2015). It attracts an amazing mix of academics, journalists and data viz experts and -- at only about 100 people -- you get a real chance to mingle and get to know folks.

Tapestry 2016. Photo: Robert Kosara
In contrast to Tapestry, I've always found the NICAR conference to be a tad intimidating, with its 1,000+ attendees -- especially when everyone shows up for the lightning talks.

NICAR 2016. Photo: Olah Weld Bessid
Which is the long, roundabout way of saying that the number of attendees at the Tapestry Conference -- which was about 10,000 last year -- is difficult for me to even comprehend. I mean, this is a picture of the keynote last year in Vegas:

Tableau 2015. Photo: Alexander Thamm

Or to put it in terms Tableau users might understand...



At 10 times the size of NICAR, the Tableau Conference also seems to have (at least) 10 times the number of sessions. Luckily, the conference app does a great job of letting you filter down to specific session types, topics and difficulty levels.



Once you sign in, you can also use the app to pre-register for hands-on training sessions, which is super helpful. (I also discovered you can quite easily de-register for hands-on sessions, which is nice if you're still undecided but want to lock down some sessions just in case.)

One thing that is both great and frustrating about the hands-on training sessions at the Tableau Conference is that they appear to each be 2.5 hours long.

For the most part, I think that's great. The hands-on training sessions at NICAR, with some exceptions, were usually just an hour long, the same length as panel discussions. In most cases, that was only enough time to get a taste of how to do something. In contrast, 2.5 hours is a good chunk of time: enough to learn a concept in some real depth. Heck, my classes at Kwantlen are 3 hours long and we can cover a lot in one of them.

What's frustrating about that length, though, is that going to one hands-on training session basically means missing out on two breakout sessions -- or, essentially, giving up your whole morning or afternoon. It also means that, even if you want to go all-in on training, you can really only make it to five hands-on training sessions the whole conference (two a day on Tuesday and Wednesday and then the morning on Thursday).

On second thought, frustrating is probably an exaggeration. I'm really just complaining that there isn't enough time to do everything I want to do. I realize it's not fair to expect conference organizers to be able to bend space and time.

One cool thing about Tableau Conference is how many sessions are repeated, so if you can't make it at one time you can go to another. This avoids the situation I've often felt at NICAR where two sessions I really want to attend are on at the exact same time (that's how I missed Scott Klein's amazing history lecture at NICAR one year; thankfully he did it again at Tapestry).

So what am I planning to go to?

Well, a lot is in flux at the moment because I'm really conflicted about how much hands-on training I want to do compared to the breakout sessions. I love to learn new things, and there are several areas of Tableau I'd like to know more about (like LOD expressions). But I also think a schedule of nothing but hands-on training will be too much for my brain to handle, so I think I'll probably aim to spend half my time in hands-on training and the other half in breakouts.

For hands-on training, I've currently got my eye on:
As for breakout sessions, these are the ones I'm most interested in so far:
And, yes, I realize I can't make it to all of those. Luckily, it looks like all of the breakout sessions will be recorded so I can hopefully catch up later on those I miss.

The other thing I'm looking forward to in Austin is the food. I'm used to going to NICAR, where you're on your own for lunch. So the scale of food choices available at the Tableau Conference is a bit mind blowing.



If you've been to Tableau Conference before, and have any tips for a newbie, I'd appreciate any advice. Just drop me a line on Twitter (@chadskelton) or post a comment below. (As an aside, the Tableau Wannabe Podcast just posted a great TC16 preview episode that's worth a listen.)

I'm planning to write some shorter posts next week once the conference gets underway, so please check back for that.

And please say Hi if you bump into me in person next week!

Note: Tableau is paying for some of my conference-related expenses.

Tuesday, October 11, 2016

How nonprofits can use Tableau: Chad Skelton's TechSoup webinar

Last month, I hosted a one-hour webinar in partnership with TechSoup Canada about how nonprofits can make use of Tableau (and get free licenses!).

You can watch a recording of that webinar below.




If you're a nonprofit in need of some Tableau training or consulting, you can drop me a line at cskeltondata@gmail.com


Friday, October 7, 2016

How to think like a data journalist: Chad Skelton at Information+

Credit: Information+ Conference
Back in June, I had the pleasure of speaking at Information+, an interdisciplinary conference on information design and information visualization organized by Isabel Meirelles and Katherine Gillieson.

You can watch my talk -- on "How to think like a data journalist" -- below. If you don't have time to watch the whole thing, folks on Twitter did a pretty good job of summarizing the key points.

You can check out videos of all the talks on the Information+ website.

There were so many great talks at the conference, it's hard to pick out my favourites.

But I particularly recommend you check out Tamara Munzner's keynote, Robert Kosara on pie charts, Gregor Aisch on creating graphics for The New York Times, Lena Groeger on ProPublica's "meat charts", Andy Kirk on doing data viz training, Scott Murray on designing online learning, Heather Krause on communicating data globally and Johanna Fulda on the state of data journalism in Canada.

Chad Skelton from Information+ Conference on Vimeo.

And thanks to Alberto Cairo and Heather Krause for saying nice things about the talk after I posted it on my blog.



Tuesday, June 21, 2016

My next public Tableau Training workshop is Sept. 24-25, 2016

My next two-day public Tableau training workshop will be held on Saturday, Sept. 24th and Sunday, Sept. 25th at SFU Harbour Centre. You can buy tickets here or by clicking the button below:

Eventbrite - Tableau Training: Telling stories with data

Here are some testimonials from folks who attended my earlier training sessions.

If you can't make this workshop but would like to be alerted when the next one is scheduled, just add your name here.

If you have several people at your organization who need training in Tableau, I'm also available for onsite training.

Thursday, May 19, 2016

Is your commute really so bad? An interactive map for BCBusiness magazine



My latest data journalism project just went live on BC Business magazine's website: An interactive map that lets you compare your commute to neighbourhoods across Metro Vancouver. You just choose your commute time from the dropdown menu at the top and the map changes colour to show you which areas have a shorter or longer commute than you. You can also hover over any neighbourhood to get more info on how people in that area get to work.




The map accompanies a feature story on commuting by my friend Frances Bula.