With data aggregation becoming easier and faster, professionals have to be able to make decisions rapidly. Problems arise in the analysis and display of data, where some insights can be missed. When visualizing your data, three things should be considered:
Less is more: Often the instinct is that once we’ve gathered all the data, it's important to show ALL of it. But understanding the context of what you’re seeing, coupled with the context of the needs of the the person viewing the data, is vital to design a visualization that gives powerful insights. A great example is our work with an energy company. The instruction from the operators who were viewing this data on a wallboard (or digital wall) was: "it needs to be red, green and yellow," which is a common statement when users are asked how they want to see their data. The problem is that a wallboard full of red indicators has an ergonomic component – its impossible to tell how to prioritize what’s bad and what’s really bad unless you have months of experience working with the data. Instead, working with the operators, we learned what was the key reason that motivates them to examine a piece of data in detail: deviation from average. So, we decided as a group that the more something deviates from its average, regardless of whether its good or bad (which is subjective anyway), the brighter we made the text. In this way, operators could immediately discern which pieces of information they needed to pay attention to and examine in more detail on their own desktops. In this way, we saved the operators time to focus on the items that matter and become more effective.
What is important to the user: Often, data is simply accessed, but not necessarily leveraged in a way that is meaningful to executives. Your data platform is something that needs to be designed by your own specific needs and what you hope to gain as a business. For example, with a financial client we worked with, we could have taken the typical route of aggregated dashboards for executives – make the dashboard look great and show the big picture. Instead, we created a visualization that shows the executive both the big picture and the details at the same time by using a ‘watermelon’ metaphor – something that could be green (good) on the outside, but red (bad) on the inside. Users simply click on what KPI they want to see the ‘insides’ of. This creates an actionable tool for the executive, not simply a passive viewing experience that gives a distorted view of health that lacks real transparency.
Connect the dots: A piece of data on its own does not give you the full story; how that data flows through time can tell you just as much if you connect the dots. For a news media company in the UK, we created a timeline of a piece of content. As the user swipes left and right through time, they can see how that content was shared or commented on and on which social platforms via which devices. This creates a picture of the newspaper's demographics – age range, gender, location, preferred device, and preferred related content. This not only helps create a more 360 editorial coverage, but by fully understanding the consumer through time, the paper now has a powerful advertising platform by being able to place targeted advertisements based on their needs of demographic, content, and device.
Our examples show the myriad ways data can be shown. By understanding the user, the context and the story you want to tell, effective data analysis, and decision making becomes a consequence of good data visualization.
Note: James Horgan will be speaking in person at upcoming event in October in NYC, covering Data Visualization in the media industry. Details coming soon.
No comments:
Post a Comment