Book recommendation: Storytelling with Data

I want to recommend a very useful book: Storytelling with Data, by Cole Nussbaumer Knaflic. It’s available from the Seattle Public Library. (I wish I could credit where I heard about it, all I know is I put it on my ‘to read’ list about 5 months ago.)

Storytelling with Data book cover image

What I really appreciated about it is that is chock-full of highly practical suggestions that can be readily implemented.

It’s also a nice balance of viz authoring advice (much of which will be a refresher if you’ve read Tufte, Few, Yau, Cairo, etc) along with very concrete suggestions for how to present data in a way that can drive action – thinking about the audience, what story you’re telling them, and what you want them to leave with.

Showing the steps involved, whether in building / improving a viz, or in telling a story using data, was tremendously helpful to me – showed how to put principles into practice.

Cole gave a related 15-minute talk at the Tapestry conference – one of my colleagues was there, and their review was:

Brilliantly relevant to our customers.  Cole did a skit of a typical boring business presentation of data findings.  She then itemized a clear set of storytelling techniques: repetition, pictures, narrative arc (plot, rising action, climax, falling action, ending).  She then retold the story.  Might have been the best talk of the day.

Here’s the talk:

I highly recommend the book to anyone who wants to communicate with data.


The Lost Macalester I-Fries Recipe, Reconstructed

Aka Italian Fries, as served at the Grill in Macalester College‘s old Student Union (circa 1990s, before the remodel and adding an extra “e” to create the Grille, which seems WAY too healthy to ever serve such a concoction). I got a tip on ingredients list relayed from a friend of a former Grill student worker. Came up with quantities & directions through several (mostly delicious but occasionally ridiculously cheesy) trials.



  • Pizza crust, 12″ (Boboli original is the closest I found, feel free to use your preferred pizza dough – it needs to have some body to be like the originals)
  • 4 oz shredded mozzarella
  • 4 oz creamy Italian salad dressing (I use Field’s)
  • Parmesan cheese
  • Olive oil
  • Garlic salt
  • Pepper
  • Marinara sauce


  1. Preheat oven to 450
  2. Brush outer ring of pizza dough (crust) w/olive oil
  3. Pour 4oz dressing in middle of pizza, brush out to meet crust/oil band. Feel free to go lighter depending on how “juicy” you liked your i-fries.
  4. Sprinkle 4 oz mozzarella evenly. (It might look like it’s not enough, but it is.)
  5. Grate Parmesan generously over the pizza
  6. Garlic salt: 3-6 shakes
  7. Pepper:  3-6 grinds
  8. Bake at 450 until cheese starts to get a little browning – 8-12 minutes.
  9. Take out, let rest for a minute or two so all the cheese doesn’t slide off, slice into 1″ wide strips, rotate 90 degrees and cut all the strips in half
  10. Serve with marinara sauce for dipping

What Mac memories do i-fries bring back for you?


Finding Free Images for Presentations

Ready to move beyond bullet points for your presentations?

People are visual. If you’re sharing data, try creating a data visualization, instead of a table of numbers. (See Tufte for a start on this.)

And if you’re conveying an idea, a single bold image can be more compelling than a slide full of text. (Check out Presentation Zen and the companion website for much more.)

Here are some of my favorite places to search for images to use in presentations.

  1. Flickr search (need another license? More Flickr CC search links here.)
  2. Google image search supports usage rights, too. (Google’s help article)
  3. Creative Commons search (and as of July 2017, a new CC search is in beta) has pointers and a front end for searching additional resources that support usage rights.

These searches help you find images you have permission to re-use. (The Creative Commons license is a very widely used way of indicating these permissions.)

Always pay attention to re-use rights and respect the creator’s intent – don’t just take any image you find.  If the re-use requires attribution, be a mensch and include the credit in your preso.

Building a Tableau dataviz of Tesla Model 3 reservation data

Twice as many Americans as Germans want self-driving cars.

That’s one of the first things I learned from asking the next set of questions.

About 400,000 people have put down $1000 each to reserve Tesla’s upcoming Model 3 – the all-electric, potentially self-driving automobile. Some of them even stood in lines that stretched for blocks to put down $1000 on a car they hadn’t even seen pictures of yet. (I’m one of them.)

Teslas can all be custom-ordered. Tesla has a great configuration site where you can pick out every detail – if you’re ordering a Model S or X. But so far, we Model 3 reservation holders don’t even have the opportunity to tell Tesla what color we want.

At least, not through Tesla’s site. to the rescue! This crowd-sourced data lets anyone self-report the options they want to choose for their Model 3. Just register and fill out a survey based off previous official statements, and maybe some rumors and tea-leaf reading, too.

Now, Teslanomics’ Ben Sullins – formerly the Chief Data officer at Pluralsight – has uncracked that dataset with a Tableau Public viz:

We interrupt this blog post to try out different ways around the problem of embedding Tableau Public vizzes on WordPress. Seeing a couple different suggestions using Tableau’s new support search.

Tried using the “logical line breaks” method John Barr described on the Tableau forums.. It did not work, even when I copy/pasted from John’s example. Though clearly it works on his blog.

Ugh. The regular embed code works just fine on WordPress in general – it looks like it’s a problem with specific WordPress themes. Bleah. Ok, for now I’m just going to cheat, strip out most of the embed, and use only the static image. Click through to Tableau Public to interact. There are other approaches I can use, I’ll come back to those…

Alright, back to Tesla datavizzing!

Pretty cool, huh? Seeing this dataviz made me want to ask some more questions, though. What about the rest of the world’s orders? Do we all want the same thing? Let’s see:

Model 3 Options v2

I plan to come back to this blog post and update it with a few details about how I made this viz, and keep updating it. Already I’ve tried out a few techniques and features I haven’t previously, and learned a thing or two.

Sharpening the saw: program manager reading list


One thing that works very well for me is reading, especially lessons learned and case studies from practitioners.9780596517717-228x300

A book I frequently turn back to is Making Things Happen by Scott Berkun (he was a Microsoft program manager who shipped things like Visual Basic for Excel, and Internet Explorer 4 – the first really good version of IE). It has been a few years since I last read it, I should do a re-read soon. Scott’s blog is also excellent.

Over break, I read a book recommended by a coworker: The Membership Economy. This is one of those business books tha510x2zsjzkl-_sx329_bo1204203200_t has a few excellent insights which then get spun out to fill a book. Worth checking out from the library and flipping through for an hour or two. Or find a good summary online. It may help crystallize things by providing a shared vocabulary for discussions with my colleagues, but I don’t think it is as insightful (or as data-driven!) as Effortless Experience.

I’m currently reading Kathy Sierra’s Badass: Making Users Awesome. I love it so far; it has the 5199hst-ofl-_sx332_bo1204203200_great design of Kathy’s Head First books, and I love the message and mindset. Highly recommended.

For this year, the topic I’m thinking a lot about is how to work effectively in teams. The book I’m reading now is Beautiful Teams from O’Reilly, with dozens of chapters contributed by a wide range of thinkers and practitioners, from programming gucatrus Grady Booch and Scott Ambler to sci-fi writer Cory Doctorow.

Another book in my queue is from Bob Lewis, longtime Infoworld columnist (and IT consultant). He has some great, short books like Bare Bones Project Management: What you can’t NOT do. (And a followup on Bare Bones Change Management that is also very helpful for 51veo882b2bl-_sx331_bo1204203200_1anyone making changes.) The book I’m going to read next is his latest: The Cognitive Enterprise. I haven’t read it yet, so I can’t tell you if I recommend it or not! But now that I’ve told you about it, it’ll motivate me to actually follow through and pick it up off my to-be-read pile – aka tsundoku.


Tsundoku illustrated by the daughter of redditor Wemedge

Tableau Easter Egg: The simplest, fastest way to get Tableau Desktop’s version, build number, and architecture

Of course, you can SEE the version and build number of Tableau Desktop from the Help menu, with Help > About.

About Tableau

But usually you want to give this information to someone else – for example, pasting it into an email or support case form.

You could use your eyeballs and fingers, and transcribe the sixteen digits the old-fashioned way.

But I just learned there’s an even easier way to do it.

When the “About” dialog is open, just Copy!

There’s no need to select anything. Just hit Ctrl-C, or right-click > Copy – and you’ll get the version and build right on your clipboard for easy pasting, like so:

9.0.0 (9000.15.0325.1651) 64-bit


Googling Tableau’s Release Notes

Tableau releases maintenance updates on a regular basis – about monthly. Since Tableau connects to dozens of kinds of datasources, which are themselves constantly updated, it’s always a good idea to stay on the latest maintenance release for your Tableau version. (E.g. if your company is running Tableau Server 8.2, you should upgrade to the most recent 8.2.x – 8.2.8 as of today.)

But in environments with intensive change control requirements, it can be very useful to be able to find specific, relevant reasons for upgrading.

Two frequently asked questions in this regard are:

1. How can I find a listing of issues fixed between whatever release I’m currently using vs. the latest maintenance release?

2. How can I find which release had a fix for a specific issue?

The Tableau support website enables both these scenarios.

The release notes for maintenance releases are available here:

Every version’s release notes are available using this URL scheme:

Let’s say you’re running 8.2.1, and the current 8.2 release is 8.2.8. To find all the fixes between, say, version 8.2.1 and 8.2.8, you can look at:

and so on. They’re now all linked on the releases page, so you can just open them in new tabs manually, or you can write a script to download them.

But, what if you don’t know when an issue was fixed, and you don’t want to have to search a whole bunch of individual release note pages? Then, you can use Google’s site search feature, and specify the following site base:

For example, if you wanted to search for release notes related to HP’s Vertica, you could do:

Screen Shot 2015-02-26 at 12.35.06 PM

Or if you wanted to scope it to only show Vertica release notes from version 8.2.x releases, add /8.2 to the end of the site parameter:

Screen Shot 2015-02-26 at 12.37.00 PM

Here, let me Google that for you.

Enjoy! Leave a comment to let me know if you find this tip helpful, or have other use cases for Tableau’s release notes.

TabPal: Custom Tableau Color Palette Generator

Color is tremendously powerful in creating data visualizations.

Tableau allows you to create custom color palettes. And usually people do this by:

  1. Find an image with the colors you want to use
  2. Use an eyedropper color picker tool; select each color in the image.
  3. Get the color code.
  4. Open your preferences.tps file and write a line of XML by hand that includes the color code.
  5. Repeat for each color.

But that’s old-school. One of my colleagues created a web app that automatically creates custom Tableau color palettes from any image.

  1. Find the URL of an image with the colors you want to use
  2. Paste it into

Tabpal identifies the dominant colors and spits out custom palette information ready to add to your Tableau prefs. Done!

For example, say I wanted to make a viz about the history of Apple, using the color scheme of Apple’s old rainbow apple logo.

Apple’s original rainbow logo, from

I fed it to TabPal. Here’s what I get:

tabpal output

TabPal grabs the top five colors in the image, identifies them, and creates the color-palette XML I need.

But, we’re not quite done yet. The Apple rainbow has six colors, so I used a web color picker (I like Colorzilla) to identify the blue color’s hex code — it’s #118ECE — and added it to the list TabPal generated, then added it to my Tableau preferences file.

color palette prefs

Now I can make my Apple viz in six colors. Moof!

apple color palette

Interested in learning more? Check out this webinar on Best Practices for Using Color in Data Visualization by Tableau Research Scientist Maureen Stone. She’s wicked smart!

Best Music of 2014 meta post

I like to listen to music while I’m working.

In fact, I like to listen to music all the time.

Especially good music.

Here’s some:

For anyone interested, I put together a 6+ hours Spotify (yeah yeah) playlist of my favorite tracks from 2014 (and still working on it).
posted by Lutoslawski at 1:06 PM on December 12 [4 favorites]


NPR Music has put together a massive playlist of their best of 2014, Songs We Love.

That link above launches the groovy in-browser app, which lets you listen on shuffle, choose a genre, and gives you artist info, as well as other functionality. You can view the entire list here if you are more eye-curious than ear-curious.

In a somewhat complete Spotify playlist