When I published this week’s data set, I knew it provided an excellent challenge.



Just because a viz is already good, that doesn’t mean you can’t do a makeover of it. Makeovers, to me, are a way to take a visualisation, look it from a critical perspective, examine and explore the data, and come up with your own data story. Many times, the best vizzes pose the best challenges. In this case, we got a great data set too!



This week’s viz review was on Wednesday, giving people plenty of time to submit vizzes for us to review. We got feedback last week from Robert Crocker to review each viz with the categories of design, story telling and best practices in mind. While we think that worked, it didn’t give us enough time to get through all of the vizzes submitted. We’ll continue to tweak the format and will do our best to pick new people as much as possible. If you have suggestions, we’re more than open to them.

Overall, we really enjoy this format and we think it provides a much better platform for feedback than Twitter. We also don’t want this to become a competition. Just because your viz doesn’t get reviewed during a session, that doesn’t mean you’re not benefiting from the comments we give to others.

All of our webinars are now on a dedicated Webinars page if you missed any of them.



One of the best outcomes of the viz reviews is people are tweaking their vizzes based on the feedback. Eva and I love seeing people continue to improve and aspire to do better work. No one was a better example of that this week than Sarah Bartlett. Let’s quickly look at the evolution of her viz.

Sarah started by trying something new from a design perspective which looked really good.

The only feedback I had was on the title. To me it wasn’t clear if she was referring to the mother’s age or the gestation age. Sarah quickly clarified and proposed we review it on the webinar, which we did.

Eva and I had quite a good discussion about the viz and Sarah took the feedback on and nearly completely redesigned it. This is dedication to her craft that we should all emulate if we want to get better.

Really amazing work Sarah!



Two vizzes in particular this week caught my eye for their design aesthetics. First, there’s this design by Daniel Caroli that is a great example of using ALL of the data, yet keeping it easy to understand. Box plots can be tough for people to comprehend and Daniel makes it work really well.

Much like Daniel, Steve Wood chose to viz A LOT of data by showing each State in each month in each year in a super small format. What makes this work so well for me is that Steve highlights the top and bottom States and uses the rest of the States for context.

If you want to read more about how and why Steve build the viz the way he did, head over to his blog.

Ok, everything wasn’t great. There are a few lessons learned that deserve some attention so we can learn each and every week.



Really, it’s inevitable when you provide geographical data that people will created filled maps. People LOVE maps. Fantastic! Yet too often they aren’t effective.

Filled maps work well when:

  1. People are allowed to compare one area to another.
  2. They identify geographical clusters of similar values.
  3. The filled areas are about the same size.

As an example, consider the two maps at the bottom of Matt Francis’ viz. Matt effectively uses color to show regional patterns.

On the contrary, filled maps fail when:

  1. Areas are not uniform – Consider tile maps or hex maps instead
  2. Color variances can be difficult to distinguish – Consider a dot map sized by the measure instead

There are positives and negatives for every chart type. I’m merely offering a few alternatives for consideration.



The more and more people practice data visualisation, the more you’ll notice how few colors they use, how much white space they allow and how they communicate data clearly by removing clutter.

Cole Nussbaumer has a great video about decluttering which you can watch here. In it, she provided 5 tips:

  1. Leverage how people see – Consider the power of the visual system when designing your work
  2. Employ visual order – How should you sequence your visual story? What should go where for best impact?
  3. Create clear contrast – Context is king. When you are displaying a chart or a number, ask yourself “compared to what?” That will help you design more clearly.
  4. Don’t over-complicate – Use simple visuals that are easy to interpret quickly.
  5. Strip down & build up – Take elements out of your viz. Did it change the story? if not, keep going. If so, you know you’ve gone one step too far. This is particularly true for colors; the fewer, the better.

This really, really annoys me. We provide a link to the original visualisation for a reason. The article will have the context for the story the author was trying to tell. They might have specific insights written in the article, but not shown in the viz. How can you include those into your story?

Too often, we see people blindly approaching the data set with no conception of what the data means or where it came from. If someone gave you a project at work, would you blindly start working on it? Of course not. So why would you approach a Makeover Monday differently?



Author: LM-7
Link: Tableau Public

What I like:

  • Beautiful tile map design
  • Shows patterns across geography and time so well
  • Great interactivity
  • Allowing the user to customize the display
  • Including the totals on the right and bottom for context
  • Clean, easy to understand design

Author: Staticum
Link: Tableau Public

What I like:

  • So pleasing to look at because everywhere you see orange it means something
  • Comparing each year to a common baseline makes it easy to compare the charts going down
  • Using % change vs 2003 makes all of the States relative to each other
  • Using the map for highlighting a State
  • Using divider lines between the charts
  • Including summary text to the right of each chart; this way you see the chart first, then the insight is recapped for you
  • Makes a strong analytical case for a connection between increased mother’s age and decreased baby birth weight

Author: Klaus Schulte
Link: Tableau Public

What I like:

  • He iterated on feedback from the webinar and made his viz better
  • Minimal use of color
  • Good explanations of how to interpret the viz
  • Allowing the user to pick a location so they can customize the view
  • Including a call to action at the bottom
  • Making analysis looks easy

Author: Kizley Benedict
Link: Tableau Public

What I like:

  • Stunning design
  • Grouping the States together by region makes regional trends easier to see
  • Consistent use of color and using it only to call attention
  • Great boxes around each section for framing
  • Massive numbers with summaries next to them
  • Sorting States from highest to lowest change
  • His chart choice was later mimicked (I assume) by Michael Mixon’s tile map

Author: Mike Cisneros
Link: Tableau Public

What I like:

  • How does Mike do it each and every week? Amazing work!
  • Terrific analytical work turned into a great data story
  • Have both sets of lines chart work from the inside out (he calls it a spider chart)
  • Excellent use of BANs
  • Cool cartogram!
  • Great use of text to add to the visualisation
  • I sure hope I can be this good as a designer some day!