Ralph Spandl developed web and UX projects for 15 years, and then he decided to switch to data visualization. A natural move, considering UX is no stranger to data.
Ralph and his team at r42 communication are sharing their work with the Data Studio enthusiasts through three beautiful community visualizations. They are also the creators of the first commercial Google Data Studio chart library. Go and check it out!
“Are we able to effectively transform data into digestible and actionable information?” is the Why of the r42 team’s purpose. So we wanted to hear Ralph’s ideas on how to effectively make data digestible in an era of information overload. Let’s hear him out!
Ralph, after working for years as an Information Architect, you are now working mostly as a Data Visualization Designer.
Is there anything both occupations have in common?
Information Architecture has a lot to do with defining a project: Listen to the client, understand what they want to achieve, talk to all stakeholders to gather all functional and content requirements, and translate all this information into rough wireframes and flowcharts in a way so that everybody working on the project (clients as well as UX / UI designers or developers) get a solid understanding of how the final product should look.
As a data visualization designer things are not that different. Listening to the data, understanding the context, asking all sorts of questions is still key in the process.
Then I still do what my designer background leads me to do – quick sketches to test out some scenarios before starting to code and make sure my ideas really work with different datasets.
In what way does the quality of a data designer’s work affect the user experience?
In an ideal world I would probably say, make the chart as easy as possible to read.
However, these days data is often very complex, and oversimplifying does not always help the matter. In this sense, we must sometimes force the user to spend more time with a chart.
If however, a reader can really understand the data after having invested some effort learning how the chart works, he will walk away with a good experience.
So it’s not just selecting the right chart for the job, but also distilling the information, organizing a series of charts to tell the full story, and guide the user through the different layers of information. As I said, you have to know who your user(s) is. Understand their needs and abilities to use the information.
In your day-to-day work, what are the principles that guide you to make data digestible?
Finding the ultimate question, one question at a time.
With data, you hold all the answers in your hands. It is however more important to find the right questions. The questions that matter and that are worth being asked. This isn’t easy, because there are so many angles you can look at data. The challenge is to select all these questions and find a clear thread to tell the story that is hiding in the dataset.
With data, you hold all the answers in your hands. It is however more important to find the right questions.
What’s needed to go from digestible to actionable data?
A clear set of business rules.
With data visualization, you can design shapes and patterns. Once we learned how to read these shapes and patterns, they can act as triggers. The actions that need to be taken when a trigger goes off must then be established by the experts.
Simple example: I recently designed a multi-target gauge for Google Data Studio that lets the user define multiple ranges/goals and attach colours to each range. The user can configure the chart so that all gauges are drawn in a neutral colour until a certain limit is passed and the gauge turns red. The colour would act as a trigger to take a previously established action.
But we can also design more subtle triggers using patterns. This requires more knowledge on how to read a chart. In the past year, we all learned that exponential growth in a line chart should result in fewer social contacts.
And the same way we could learn to read patterns of a point cloud, a sunburst diagram, or a radar chart. The specific distribution of volumes and positions, agglomeration of points, or repetitive patterns are just some indicators that can act as a trigger and require appropriate action.
You’re saying that data visualization can help us identify patterns.
How can one get better at spotting and analyzing patterns?
First, you will need a complete understanding of how the data is drawn and what each pattern really means. You learn by reading charts, just as you learn to read the alphabet: you need to look at the same charts with different data over and over again.
The process is called pattern recognition and involves matching the information received with the information already stored in the brain. This requires training of your long-term memory which can be achieved with repetition of experience.
The ‘Fullstack D3 and Data Visualization’ book from Amelia Wattenberger contains a fun section that teaches you to design radial weather charts. One year of weather data in a single circular diagram.
I extended this little exercise into a one-page website that allows you to compare how the weather is in three cities: Montreal, Munich and Tokyo. These cities actually have quite similar patterns, but if you look carefully you can for sure spot the differences. Imagine for a moment how the chart would look like from a city in the Caribbean (same temperature all year round) or in the southern hemisphere…
I believe that in order to be a data visualization expert, you also need to be good at storytelling. What’s your take on this?
Absolutely, charts are like tiny dots. A good story can connect these dots and help us see the bigger picture.
I am no expert in cognitive science, but the same way we help our brain understand data when we transform it from a spreadsheet to a visual chart, we can further help the understanding by telling a story using multiple charts and look at the same data from different angles.
By connecting the dots with other patterns that everyone understands, relatable, memorable, humanizing information if you will, we can make the pattern clearer to see.
Do you consider there is such a thing as good statistics or bad statistics? What extra steps can help specialists make sure they’re collecting the right data?
I am usually not responsible for collecting the data, so I can’t really give too much advice here.
However, as data visualization designers we can do our share to avoid bad charts. First, we must have a complete understanding of the data we received.
In which context was the data collected? What does each metric and dimension precisely mean? Are there any causalities between the numbers or are obvious patterns just coincidences?
We can’t just assume the answers and must ask the questions. And if the data is not coherent we should ask for clarification until everything makes complete sense.
Ralph, how should a data-driven business culture look like?
This question may be outside of my expertise. However, I am sure you can’t implement a data-driven business culture by throwing numbers on a dashboard and giving everybody access to it.
I think the key lies in encouraging everybody to discover data on their own. Therefore you need easy access to raw data and visualization software with a gentle learning curve.
And then you will need to increase data literacy within your company. As I mentioned before, reading data is something that needs to be learned, like reading the alphabet or a new language. And the same iterative approach learning how to read data can be used to apply it. Both take time and effort.
We’re both heavy users of Google Data Studio. I’m curious - what do you feel this tool is missing at this point, feature-wise?
To me, Data Studio is a visualization or dashboarding tool, not a data preparation tool.
In this sense, Data Studio has a very limited vocabulary and is missing many helpful charts. This is why I started to design and develop some free community visualizations and maintain a commercial chart library.
I love the way Data Studio makes data visualization very accessible. It is super easy – even for data novices to visualize their data within minutes.
And I try to keep this spirit with my charts: They should work with the first click and configuration should be comprehensible, easy to configure, and adhere to data visualization best practices.
Thank you for your time, Ralph! Our last question - is there a data visualization that you grew fond of? I’m not talking about a particular type, but about the story that it tells.
I am a big fan of circular charts. Especially when applied to time-based data, they make a lot of sense. One day or one year of data in a beautiful and attractive circular chart… we are getting close to heaven.