Early last week I posted a tweet that discussed my hatred for ‘long-form’ style visualisations. By that I mean, one that needs use of the scroll bar to be fully consumed by a user.
It was meant as a thought provoking tweet. I wanted it to spark a debate and discussion.
Over the last few months ‘long-form’ visualisations seem to have been popularised, what caused this I cannot quite pin point, but every time I have tuned into my social network recently they have been everywhere. It genuinely pains me.
I think part of the reason is because we are so used to consuming information by the internet. The internet tends to have vertical scroll bars, why can’t visualisations too?
Well I will tell you why, because they go against a number of key data visualisation best practices.
Best Practice 1. Scroll-bars
One of the first things I learnt when beginning to study data visualisation is the importance of having all the information and context your users need to be in the view.
By having scroll bars in your dashboard this immediately becomes difficult.
This visualisation by Tableau Zen Master Andy Kriebel is a good example. He’s made the decision to include the results of all 50 states to allow users to understand the voting predictions at a state by state level.
However by visualising in a long-form style it is impossible for users to compare all states in one go.
Best Practice 2. Not concise
Most visualisations that tend to be in long-form style are in long-form style because they are simply not concise; situations where the designer/analyst has failed to pick out they key points that they wish users to consume and instead give all the detail possible.
Lets think more about what information is necessary, and what are the key takeaways we want for our users. Andy’s post above could have been made more concise by focusing on the different questions his users are likely to pose.
What are Clintons most dominant states?
What are Trumps most dominant states?
What states look likely to be the ‘swing states’?
Perhaps then we could present the top 5 that fitted in each of these categories, leaving us with just 15, rather than 50 states.
Yes doing so would remove the ‘state by state’ element to the visualisation, but instead it emphasises the key points, and it does so in one view.
Sometimes the content that has caused the dashboard to be made in long-form is actually completely irrelevant to the information at the top, in cases like this perhaps it would be better to break the visualisation into different dashboards or story points.
Best Practice 3. FUgly
By that I mean, they are f*****g ugly.
Seeing half a chart at the bottom or top is not visually pleasant and can be hugely distracting to what you want your users to focus on in that specific part of the visualisation.
Take the bottom right part of this visualisation for instance.
Best Practice 4. Practice what you preach
Okay, I get that we use Tableau Public as this experimental tool, I’ve done it myself on a number of occasions with my #ReViz project, looking too see what’s possible and what works.
My worry is that we take long-form visualisations into our professional practices, and that would make things a whole lot worse.
Lets remember how most of our professional stakeholders want to view their content, quickly, and at a glance. Long-form visualisations definitely do not impact positively to either of these criteria.
They are negatively impacted because they have to spend time scrolling through the content, and also they may not be able to see all of the context required.
I’m going to sign off with a pretty bold statement, but long-form visualisations infuriate me far more than pie charts.
Disclaimer: I fully endorse the use of long-form visualisations designed for mobile