Launch, grow, and unlock your career in data

April 3, 2017

Makeover Monday: Which UK workers are most at risk of being replaced by robots?

Most week's when we're starting the Makeover Monday process, we have a pretty crappy viz to work with. Eva has mixed things up on us this week by picking a viz that is already quite good. This makes the process of making it over even tougher.

This week, she chose this viz from The Guardian about the jobs that are most likely to be replaced by automation:

What works well?

  • The overall design is super simple to understand.
  • The gaps are evident by using distinct colors for the two categories.
  • Including references to the data source
  • Clean fonts
  • Grey background bars help make the blue lines pop
  • Axis labels only include the % sign on zero
  • Sorting is easy to understand even though it isn't stated
  • Long scrolling view suits this design

What could be improved?

  • Grey background bars don't need to extend the entire width. They would help accentuate the difference if they only extended between the two blue bars.
  • Why are some of the jobs in bold?
  • The title is way too long.
  • Is there a better way to make the gap easier to interpret?

Since I really like the design of this visualisation, I decided to try to recreate it while also including the things that I would like to see improved. I'm going to post this as a future Workout Wednesday, so if you participate in that project, hold off on downloading this workbook to see how I made it.

Another fun week. I learned a bunch creating this viz and it was easy to keep all the work under an hour.


  1. I'm honestly struggling to clearly interpret this plot as it is:
    does it show that, say, 46.4% of all jobs in manufacturing are at high risk of being replaced, but in total people employed in manufacturing are only 7.6% of all employed everywhere?

    wouldn't scatter plot work better here -- with percent of total employed population and percent of jobs under high risk as two axis?

    1. That's the way I interpret the meaning Eugene. As for scatterplots, check out the weekly Makeover Monday recap. You'll see a couple good scatterplots there.

    2. Thanks Andy, I had a look and quite a few excellent ideas there.
      and thanks for running all those blogs,btw - always interesting and useful.