So, there I was, creating a dashboard for my application to The Data School, thinking about which charts would go where, what colours to use, and how to present. I would tell the story of migration in Japan, who moved where and what jobs they did. It would tell the story of population movements and potential skills shortages.

After a largely successful interview, with some constructive feedback given, it dawned on me just how much of a dashboard’s impact is conveyed through its storytelling – i.e. 99.475% of the impact to be precise.

I think my storytelling hit somewhere in the 78.902% ballpark initially. And while I did tell a story, after just 1 week at The Data School, I knew the story could be better, more targeted, more impactful.

On the Friday of the first week, we spent some time updating out interview dashboards. A perfect opportunity to do a ‘storytelling pass’ to finesse the presentation.

And here’s what I decided to do.

1 – The map shows the number of migrants coming in to each of Japan’s 47 prefectures. But it was hard to tell which prefectures were losing more people than they gained, so I added a colour code to make this immediately apparent to the dashboard user.

2 – The Labour Status bar chart shows a breakdown of migrants by their occupation. Interesting, but it doesn’t really uncover much. By adding a parameter to toggle between actuals and % of total migrants, then adding a second table to compare one prefecture to another, I was able to quickly discover that a similar number of people migrate to Tokyo and Hokkaido for government positions – BUT a much higher % of all migrants go to Hokkaido for this reason.

3 – I had tables for Top and Bottom origins of migrants. Now, who wants to read a table of numbers?? Instead, this was changed to an ordered bar chart with a dynamic benchmark. Now we quickly and easily see which prefectures had more than X people coming into the selected prefecture. This is a better way to isolate and identify prefectures that may be losing too many people.

4 – Finally, in the bottom right corner I had a chart using shapes to separate Male and Female. Functional but not extremely legible so I changed it to a stacked bar chart. Now we can easily see how Male and Female compare.

Key Takeaway

Moral of the story? It’s worth allocating some time to truly figuring out what story you’re going to tell – how a dashboard progresses, how each chart contributes to the next, and how each chart reads.

Here’s how my storytelling changed:

  • OLD: “look at this map, lot’s of people migrated here”
  • NEW: “Hmm, these 2 prefectures have the same migrants in, but one is also losing people, let’s hone in on that one”
  • OLD: “here’s how many people migrated for government jobs”
  • NEW: “Hmmm, equal numbers seem to be migrating for government jobs, but a much higher percentage of migrants do so for these jobs in Hokkaido – is there a skills shortage there? Are government officials from other prefectures better trained?”
The Data School
Author: The Data School