A Look Back at the #MakeoverMonday Project

A Look Back at the #MakeoverMonday Project

It’s about time I write a post about #MakeoverMonday…

First of all, what is #MakeoverMonday?

It’s a community data project started by Andy Kriebel, the Head Coach of the Information Lab’s Data School and Andy Cotgreave, a Senior Technical Evangelist at Tableau.

Each week, the Andy’s select an existing visualization (some good/some bad) and provide a cleaned data set to the community.  The idea is to learn, practice, and improve your skills by visualizing a foreign data set each week.

The project continues to grow with over 500 contributors creating over 3,000 visualizations over the course of a year.

Check our their website if you want to learn more.

Here are the general guidelines for the project

Click to Enlarge
Click to Enlarge

Over the year, there were a few bumps along the way interpreting these guidelines.  Some folks viewed them more as hard rules.  This would ultimately be impossible to control as the project grew so large, so fast, and with skill levels ranging from people totally new to Tableau to regular participation by several Tableau Zen Masters.

My approach to #MakeoverMonday


I will be the first to admit that I view the guidelines as just that, guidelines.  To me, this project is about improving your own skills, learning along the way, trying new approaches, and most importantly having fun.  We all live in time-boxes at work.  If you want to spend a longer time exploring the data, looking for an interesting story, or trying a new technique, I certainly won’t be mad at you.

One of the most important parts of being an analyst is knowing the data.  When someone says, “I did this in ten minutes”, I immediately assume they barely had time to look at the data and understand it.  I probably won’t look much closer or put much stock in the analysis.  I fundamentally think knowing the data and providing accurate analysis is more important than speed-vizzing just for the sake of speed.

Therefore, I approach each makeover the same way.  I spend a little time looking at the original visualizations, reading any available stories, and researching things I might not understand before I even download the data.

Then, I look at what fields were provided and start exploring the data for some basic things.  How many records?  What data types are there?  Is the data over time?  Are there geographic fields?  Am I comparing something?  This doesn’t take a ton of time to do and almost 100% of these exploratory views are throw away.  I think Andy Kriebel calls this “Failing Fast”!  I am all about that!

All of the above is to ultimately get to, WHAT AM I TRYING TO SAY?

How I decide on design/style

A few things come into play when I think about design.

I believe there is a balance between beauty and function with data visualization.  You could have the most insightful dashboard in the world, but if it is painful to use or downright unattractive, I guarantee people will not use it.  I think designing a dashboard that is beautiful and functional is a key to success.  It shows you took some time to think about the user experience.  How will someone use this dashboard?  What questions will they answer?  Is it effective? Are the interactions intuitive?  Is it fun?

Backgrounds…this is such a personal preference…  So here is some shocking news: I think dark backgrounds are awesome!  I think they have a modern appeal.  I think they can effectively show data without overwhelming given the proper use of color and technique.  I also totally understand others can feel exactly the opposite.  While I offer a mix of design styles at work, the bulk of those visualizations are of the lighter variety.  So for my personal stuff, I lean towards darker backgrounds.  Like most things, I think you should be good at both.  Practice makes perfect.

watch-out-weve-got-a-fancy-pants-over-hereDesign Complexity.  This primarily depends on two factors: 1) If I found something interesting in the data and 2) If I have time to do it.  I am not afraid to do something technically complex whether it be advanced calculations, a nonstandard visualization, or floating dashboard design if it makes sense and results in a better user experience.  So while the technical makeup of some of my dashboards is high, I’d like to think most of my visualizations are easy to understand.

Some of my makeovers are just single worksheets and some have several elements.  You can achieve pretty much anything in Tableau if you are creative.  My main message here is just because something is “fancy” doesn’t mean it is hard to do.  It could just be some simple formatting at play or use of some imagery.  I often start with a good image and use colors from that image for inspiration.

What I got out of participating in #MakeoverMonday

I am one of the few people who completed every single makeover this year.  I started in week 7, but I went back and completed the weeks I originally missed.

I challenged myself to improve my skills this year by studying what people are doing in this industry.  I can think of no better way than being a part of a project like this.  Where else do you get to see over 60 visualizations (on average) of the same data set?  This is an amazing learning opportunity to see how others see data, how they visualized it, their use of chart type, color, design, interactivity, storytelling, etc.  It is always good to expand your data literacy by working with unfamiliar data.  And, bonus time…Andy K. does all the data prep for you.

I became a part of the larger community.  A community where everyone is so willing to help, provide constructive feedback, be a sounding board for ideas, and the chance to interact with other data geeks.  A great friend of mine always says, “Your vibe attracts your tribe”.  This is so true.  I now have an endless list of people I can (and do) easily reach out to if I need help.  Being a part of this community afforded me the opportunity to start this blog with Pooja Gandhi, add a new co-worker in Rody Zakovich, and to be named a Tableau Ambassador.  I don’t think any of these things would have happened without #MakeoverMonday.  The community was also on full display at the Tableau Conference where Twitter data friends suddenly become real friends over data and beers.  If I ever travel somewhere, I am sure I now know someone close by to meet-up with!

Expanded Portfolio of Work.  I was already a Tableau Public junkie before #MakeoverMonday, but now about half of my public visualizations are makeovers.  The more of these challenges you complete, the more work you will have in the future to show a prospective employer.  As I write this, I have 114 workbooks on Tableau Public.  An employer can get a pretty good idea of your skill level just from looking at a few of these.  If you had a choice of two equal candidates and one had over a hundred workbooks while the other had two, which would you choose?  There have already been #MakeoverMonday hires (including my favorite zen master) as the result of the increased visibility this project offers.

Finally, a BIG THANK YOU to the Andy’s!

I appreciate all of the work that goes into sourcing/cleaning the data, pinning the tweets, interacting with the community.  This has been an amazing learning opportunity for all those involved and a real community success.

So this was my look back at #MakeoverMonday.  I am now looking forward to the next year (and hoping it continues)!

Check out all 52 of my Makeovers here.

Pooja’s are on the left and mine are on the right.

Leave a Reply