After last week’s Olympic medal focus, this week we shifted gears completely and brought you data about the value of pharmaceutical drug exports by country from 2013-2017.

What I first noticed last week and then again this week was that despite the data being at the country level, very few people created maps. That was great to see because as we’ve pointed out in the past: just because you have geographical data doesn’t mean you need a map.

So well done on focusing more on the data and what it tells us rather than moving quickly to the visualization aspect.

Maps can be a great tool for context and interactivity and to get the attention from your audience but they can also take up space without adding real value to your data story.


If you want to read a bit more about maps, check out these recaps:


Viz Inspiration


What I really enjoyed seeing this week was how people helped and took inspiration from each other to create their visualizations.
In particular what happened was that Sarah posted a viz showing a bar chart resembling a pill shape.

Andy then ‘had an idea’ (this type of comment from him can be dangerous but usually turns into a very useful lesson and blog post for everyone 🙂 ) and turned the bar chart into a full pill-shape and this design was used by a number of others as they practiced the technique.

Sarah then used Andy’s tips to refine her original viz and the design also influenced the design choices of Jevon and Michael





When you first tackle a new dataset for analysis, there are a few steps you should go through before actually visualizing any day story. And you can use visual analysis to learn more about the quality of the data in front of you.

What do I mean by that?

When you first connect to your data in your tool of choice, it’s always a good idea to familiarize yourself with the dataset before you try and find a story.

  • What are the different fields in your data?
  • What are the members of your dimensions?
  • What are the ranges in your measures?
  • Is data missing?
  • Are there any geographical data points that are ambiguous?

and so on…


In this week’s dataset there were some countries which didn’t have any data for 2017, because it was not available at the time the dataset was created

When comparing countries over time it is therefore best to either exclude 2017 from your analysis because it will only be relevant for a subset of the countries, while all countries can be included from 2013-2016, or to compare only those countries that have data for every single year.

Otherwise what can happen is something like the below line chart where a sharp drop for 2017 makes it look like exports dropped dramatically, but all that happened was that only some of the countries have data, reducing the total sum of export values for that year compared to the years where every country contributed a value.




We can achieve creative and impactful designs through the use of color and many of us use it as a way to draw attention to our vizzes. Each week’s topic can serve as an inspiration for certain color choices and some that quickly come to mind for this week’s focus on drug exports include green, purple, and red.

What we need to keep in mind, regardless of which color(s) we choose, is that our choice should ideally carry meaning and therefore support our overall story.

For example, we can use color to call out a specific country in the dataset, in which case it makes sense to use the same color for that country wherever we refer to it, e.g. in different charts and/or titles or subtitles.

When it comes to color, less is more and using color sparingly but very deliberately, can be a powerful support for your overall data story.

Staticum produced a great example this week where a single color is used to highlight South American countries.




Author: Tom Pilgrem
Link: Tableau Public

Author: Jack Horton
Link: Tableau Public