Can you believe that with this blog post, the Makeover Monday year 2017 is coming to an end? No, me neither, and thankfully we can look forward to another 52 datasets starting Monday…

For the last week of the year I chose Christmas data after Andy made me do it. I had initially picked a really cute data set but don’t worry it will come in the new year. So this time we looked at Christmas tree sales in the US. Are you buying real or fake trees?

The data set itself was simple and it allowed for quick makeovers which was most important because I didn’t want people to spend hours doing Makeover Monday when they’d rather be celebrating Christmas with their families.

For those who didn’t celebrate Christmas and instead maybe worked, it was hopefully a nice way to exercise your dataviz muscles before the end of the year. And once again like every week I’ve been surprised by the creativity people brought to their visualizations.


Two things really stood out for me this week:

Firstly, I really like the simplicity a lot of people used to approach this data. Andy and I have often written about simple being best and that there’s nothing wrong with a simple bar chart if it tells a good story with the data.

This week a lot of people seemed to embrace this and I was excited to see so many clean and neat bar charts.

One really good example comes from Alicia who shows the decline in sales of real trees with this horizontal bar chart:


Secondly, what I also liked was seeing people approach this very non-business topic with a ‘business-y’ dashboard.

Makeover Monday is a great way to use random data sets, a fresh one every week, to practice your skills of visualizing data for the business environment. And why not use Christmas tree data to create a dashboard that could equally apply to store sales, staff performance or warehouse inventory?

For his makeover this week, Sean Miller created a simple dashboard with a line chart looking at the overall sales as well as a barbell chart looking at the difference between real and fake trees. This dashboard could easily apply to business data and is great way to represent this topic.

I also have a couple of lessons for you so we can all take away some ideas to improve our data visualization skills in the new year.

As it’s the week of Christmas I will not focus on adding bling to dashboards or unnecessary icons. Those lessons have been covered enough in the past few weeks.

Instead I want to focus on the following two:




For clarity and ease of understanding it is really important to be specific with your title and descriptions so that people don’t misinterpret your viz.

Sometimes we use titles that are funny, witty or controversial because we want to get peoples’ attention. This can backfire if we are stating something in the title or questioning something and it actually doesn’t relate directly to the data set or is in contradiction to the data.

This week’s data set included the number of Christmas trees sold over the years in the US. This does not mean the number of households that have a tree or the dollars people spent or the number of trees that already exist. It simply means the number of fake and real trees sold per year.

If we use additional data to enhance our visualizations and analysis that’s fine and we have to clearly state this so that others who participate know how we got the additional information that we are presenting. Otherwise, Let’s be careful to clearly state what the data shows and avoid ambiguity about the content of our visualizations.

Words really matter so my suggestion for you is to read the title of your visualization back to yourself out loud and ask yourself whether it makes sense. Does it accurately represent and reflect what you have analyzed and what you are showing in your visualizations?

I know this can sound nit picky, but it is really important for me that we all get better at this and that we are very clear with the words we choose because back at work it really matters to not be misunderstood and to be taken seriously, so let’s be deliberate and precise with our language.




I promised earlier I wouldn’t critique people for using icons and images in the data visualizations this week. I will stick to that promise but one thing I can’t look past is colors.

A Christmas topic, of course I should expect lots of greens and lots of reds and baubles and trees and stars. I can live with the icons and images that’s fine and in the spirit of Christmas I will not say another word about them.

I do want to remind people, though, that they are making life harder for themselves and their audience by using lots of colors in a visualization or color combinations that are not ideal. Especially when it comes to backgrounds.

Do you really need your visualization to have a colored background? I suggest in most of the cases you don’t. Using a red background and then green color on the background for the Christmas spirit is going to make your visualization very, very difficult to read. And not just for people of color blindness.

If you really want to use a background color, why not try a light gray or even a dark gray but something neutral that makes your other colors stand out?

Or maybe just stick to a white background, it often works best and you can still create a really impactful visualization.

I have seen a number of visualizations this week that used to different shades of green for real versus fake trees. I really like this approach because we don’t need to use two very different colors like green and red or green and orange to depict real versus fake trees. Two distinct shades of the same color can work just as well and look a little little bit more harmonious and, in my opinion, more polished in the final product.


Bad colour combinations

Better but do you need different colours for the lines?

Maybe two shades of one colour will be enough?

Light grey background

Darker grey requires adjustments to the viz colours

Almost black