Data Storytelling Case Study: The London Underground Map

Storytelling isn’t “Once Upon a Time”, it’s how you frame data for the audience.

When we produce analytics for our Podcast Clients at Pikkal & Co I encourage the engineers to think about “the user story we are trying to tell.” Engineers are trained to think about solutions to problems, not necessarily how people will use those solutions.

Our hypothesis at Pikkal is that Podcasts are the Ultimate Big Data source. There is a ton of data that we can potentially mine. That’s why when we first started building our Ai Powered Voice Dynamics graphs to help clients analyse their audio content, the default engineering position was to make those graphs accurate.

Accuracy means nothing unless the user wants accuracy. Most people want a solution. How does this solve my problem?

Consider the London Metro “Tube” Map, one of the first metro maps published in the world. When Harry Beck took the map in 1931, he radically transformed it. Technically, it’s not accurate. It’s not even near. London transport authorities were sceptical because it was over simplistic and opted to publish the map as a pamphlet only. But, the popularity of Beck’s map with the public won over, and this version eventually became the map we are familiar with today.

Let me explain a little about maps first. You might think that maps are scientific representations of the world, but you’d be wrong. Maps are extremely subjective. At the most basic levels, someone had to decide how to translate a 3d sphere into 2d paper. At the outset, maps are representations rather than facts. You don’t fully appreciate how subjective maps are until you see alternative narratives.

The Mercator map, as seen on most classroom walls, looks very different compared to alternatives (this Reddit post has a good visualisation). In the old fashioned Mercator, Greenland is bigger than South America (which it isn’t). Africa and India are tiny compared to Europe.

The Tube’s iconic map has changed little in style over time as the Tube network has grown. One reason why the map has been resilient is its effectiveness. For those who know London beyond a tourist eye level view will know that the drop pins associated with train stations are actually geographically very inaccurate. But that’s the point. The London Metro map is actually a masterclass in data storytelling. The inventor didn’t think, “how can I show better data?” but started asking the question, “how is the user using this data?”

London Metro commuters don’t use the map to walk between stations, in which case it would need to be geographically accurate. What they use it for is to know which stop is next, or are they on the right line? Data accuracy, therefore, is less important than its clarity.

Key messages:

- Only data scientists care about data, most people care only about solutions

- To create effective data you need to understand how people are going to use it

- There are real world tradeoffs to be made: accuracy and purity of data is less important than its usefulness.