IBM Content Analytics helps companies access, aggregate, analyze, and visually explore large volumes of unstructured content to unlock new business insights and identify areas that deserve deeper investigation. The analysis of documents, e-mail, database records, and other enterprise content can help solve decision-making challenges across a wide range of industries, such as:
The user interface is highly visual in nature with much less text than other interfaces, and users can select different visualizations to explore content in different contexts and from different perspectives. Unlike keyword searches, where users try to find the needle in the haystack, Content Analytics reveals patterns, trends, deviations, and unexpected correlations that can help users discover what the haystack is all about.
To be successful, the visualizations needed to be both informative and easy to understand. The first challenge for our team was to design visualizations when we didn't know what users would be looking for – to design for something that doesn't have a concrete answer and whose purpose is to facilitate discovery. We used an iterative design and evaluation process with users to try to understand what they were looking for and how they thought they'd use it.
Because it was new, the users really didn't know what they wanted until they saw it, so we tried a lot of different approaches. Every time we talked to a user from a different industry, we got yet another insight into how the product could potentially be used, and thus how to best design for that. We knew we were on the right path with our Connections visualization, for example, when we heard things like "Wow…that could really help me understand my drug trials and the interactions...now, could you just add..."
The second challenge we faced was how to help users get started when confronted with an interface that includes multiple paths. The order of the visualizations in the interface is not linear. Users can explore content through a single view, such as exploring trends in customer satisfaction records over several years, or they can explore content from multiple views and iteratively drill down through a vast amount of data to discover critical business insights.
Feedback from the many usability sessions that we conducted convinced us that users needed help, but that they were unwilling to click "Help" to get it. We learned, however, that the users were willing to learn and learned quickly by being shown how to manipulate the visualizations. While they were resistant to reading, they gladly invested a few minutes to watch a video.
Based on this feedback, we embellished the user assistance by:
Because we embellished the user assistance after the first round of usability tests, we were able to achieve validation of the improvements in follow-up sessions. Test survey results showed that participants who chose to watch a tutorial before attempting a task returned much higher ease-of-use scores than those who attempted to jump directly into the product and tasks.