The objects between big data and UX

As a proponent of user-centered UX design, I am always on the lookout for new takes on the design process that may play a role in allowing the systems that we design to be more user friendly. Sophia Voychehovski’s article on Object-Oriented UX struck me as an interesting method toward approaching projects where knowing the content structure beforehand can allow a modular approach toward the application’s layout hierarchy. By designing around “objects” rather than pages or sections accessible through navigation, we can break down barriers that we needlessly put in front of our users in their search for information. The article got me thinking how it would be possible to put this method into practice for applications that we typically design, which include data visualizations powered by big-data.

The strong selling point of our applications are that we design powerful tools that can provide the user with as much information as they’ll ever need through comprehensive data and give insights through predictive analysis based on that data. While this goal is a huge selling feature, it presents a problem from the angle of the user-driven design process. Mountains of data in the hands of someone skilled in the realm of parsing spreadsheets and intricate line charts is powerful in and of itself. But what about users who do not have backgrounds in data science? Providing a firehose of data with no method to break things down into easy to manage portions can inundate and eventually frustrate a user looking for a specific story within the library of data.

While the Object-Oriented UX (OOUX) method seems to simplify the design of a complicated system where we know what the user is looking for, it becomes more complicated for open-ended systems in which the user dictates their own story each time they use the application. After approaching the design of our application applying fundamental best-practices, how do we continue to adapt our methods to make sure we are not placing barriers in front of our user.

For our applications, it starts by looking at the datasets themselves as objects. In the OOUX method, after breaking down the user’s needs into objects, the designer can then identify opportunities for nesting these objects within each other to increase contextual navigation. The same logic can be applied to data itself to build out simplified data visualizations that take multiple streams of complex data and compile it to an easy-to-read story for the user.

Take, for example, the large amounts of data provided by the city of Chicago. A user could take a look at neighborhood tax statistics to get a sense for the spread of income throughout the city. But if we take this data object and nest within it data for crime by location and locations of businesses, namely grocery stores, we now have a real story showing the impact of urban food deserts on lower income communities. A social enterprise may see value in any one of these data objects, but the combined story provided by all three give them an actionable goal with a tactical approach. It is not hard to see how this can be applied to most any industry that can benefit from data-driven solutions.

By constructing our application to be accommodating for modular data sets, we can make sure we are prepared to handle the user’s needs as they shift. Standardization of data with consistent and easy to understand visuals can help ease the transition of the user between datasets and help them rely on contextual navigation rather than segmented sections driven by top-level, navigation. Data analytics can provide the razor’s edge companies need to make well-informed decisions and drive customer engagement. By making our data modular we can provide a tool that is flexible enough to not only show our users exactly what they want to see, but what they may not have thought to look for.

Written by Tony Knaff

Leave a Reply

*