Data Science and the Value of Design

Forecast.io's Forecast Lines app.

Forecast.io’s Forecast Lines app.

Data visualization is far more than simply translating information into charts. Effective visualization reveals insights and tells stories through rigorous analysis and effective design. As a result, operationalizing data visualization can prove challenging.

Complex datasets and counterintuitive insights aren’t necessarily communicated effectively with the venerable pie charts and graphs that dominate many a dashboard. Finding the correct design language for a particular use case often requires a counterintuitive approach of its own. Conversely, there’s a risk of creating visually impressive visualizations that fail on the most basic level by confusing rather than clarifying.

Given these demands and constraints, some of the most impressive examples of the craft can be found on mobile platforms, which favor economical design that communicates information to users quickly and clearly.

Recent exemplars include forecast.io, a mobile web app that displays a wealth of local weather data in a simple animated interface optimized for playfulness and clarity. Sitegeist, a mobile app by the nonprofit Sunlight Foundation, performs a similar feat for neighborhood-based civic data, displaying demographic information such as age distribution, political contributions, and home values, as well as listings of local venues and entertainment.

In a post at Fortune, designer Olof Schybergson shares his insights from a decade of consulting on interface design for corporations, illustrating the value of “a design-led approach to data”:

Designers have it ingrained to focus on simplicity, and bring a singular focus to delighting the end user – regardless of whether they are a business user or consumer. Designers know how to take complex or disparate information and make it tangible, understandable, and importantly, more human. Design can play a key role to make digital data adapt to our messy lives and the real world. Designers can bring stories and humanity back into the digital services we increasingly rely on for all aspects of our lives.

Among Schybergson’s examples of this principle in action is the online billing system for Swedish mobile provider 3, which takes “information that can be classified as amongst the dullest on the planet, and potentially one of the most aggravating for customers–and turns it on its head.” The interface, which helps customers easily determine relevant account information, such as whether they are close to incurring overages, has paid off in increased consumer satisfaction ratings. He also points to the SMART initiative, a project of Harvard Medical School and Boston Children’s Hospital, a pediatric growth-charting app that presents customized views to parents and pediatricians that communicate user-relevant data.

Harvard Medical School and Boston Children's Hospital's SMART app.

Harvard Medical School and Boston Children’s Hospital’s SMART app.

Within risk-averse organizations, such principles may seem unlikely, but there are success stories of traditionally button-up organizations reaping the benefits of placing an emphasis on design-centered thinking and rapid, data-driven development. One such example was the Obama 2012 reelection campaign’s Dashboard app, a voter outreach and organization tool that provided volunteers with a useful tool to submit self-reported data, while providing sophisticated analytics capabilities to those working within the campaign.

In his Fortune post, Schybergson emphasizes that effective visualizations come from organizations willing to take risks. The end result should prefer elegance and simplicity, but achieving those goals requires the embrace of a degree of messiness and even intuition. Data can do much, but communicating its insights still takes a human touch.

Permalink

Tags: , ,

Leave a Reply

Your email address will not be published. Required fields are marked *

*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>