If you've been following any recent releases in all product lines,
you'll know how making our products visually engaging has been an
integral part of the drive towards simplicity of the user experience. Our design jam next week on Visualizing Information is a result of that drive. I'm pretty excited about this event since it's bringing together so many experts on user experience and data visualization, as well as product subject matter experts at Oracle into one place for two days of design, and in an area that I'm particularly passionate about as a UX professional.
For any design jam or hackathon, we do our best as organizers to ensure that we prep participants with enough training and information, or give them some good guidelines with what they should do in advance to ensure that their limited time together is as productive as possible. For this design jam, we just held a training session on some ideas of how to approach the data visualization you're trying to design, and I thought might be good to share here.
1. Start with the user, not the data
In UX, this is our mantra, so nothing new here. But, when it comes to visualizing data, the approach can often start with a goal of making extremely large data sets understandable. Of course, that's a critical goal, but we usually step back and start asking questions about who the user is, his motivations, goals, how he spends his day to understand what he may want to learn from the data. There is a wealth of information in understanding the user through personas or user profiles, business process models to understand tasks and dependencies, or even job postings, so I won't belabor it here.
2. What are the users Big Questions; in other words, what problem are you trying to solve? In our example, our user is a Sales VP.
She has a lot of big questions, but for the sake of this story, we're going to focus on whether she is going to make quota.
3. Distill the question(s) into something you can design around.
Is there data available that can help answer her questions? This is where the magic mix of art and technology come to play. Assuming you do have access to the data..
4. Start by figuring out which visualization technique might work.
Time for me to give kudos to my colleague, Katia Obradovic-Sarkic, and owner of the Data Visualization Blog. In our recent training, she went through many use cases and examples of effective visualization techniques that work well for given problems you're trying to solve. There are so many others with greater expertise than me, in particular Katia, so let me just refer you to Katia's blog to learn from the expert and continue my story on approach. In our event kickoff, our UX experts, John Cartan and Julia Blyumen are going to spend a few hours training, so I'll be sure to report back some of their pearls in this space, as well.
5. Design your visualization.
Easy, right? Well, of course not. I'll paraphrase a favorite quote about design, "everything is in a nightmarish state of total failure until the moment it is not". Wish I knew who to attribute that to, since it pretty much describes most creative processes.
In my experience, there is a whole lot of iteration between my steps 4 and 5, including a lot of testing with your data to see if it all falls apart. But, if we wanted something easy, we'd be brain surgeons and not designers ;-)
Looking forward to what the next week will bring. Until then, will be spending time trying to get my left brain to talk to the right side. Ciao!