UX Design 101: Context in BI and Analytics Design
What’s the Context, Dude?
The first question I ask the stakeholder when starting any new UX research and design project is:
"What’s the context?"
If that question can’t be answered, I know there’s more up-front work to be done before anything else happens. When you ask this question, be prepared for a range of responses that swing from blank stares to, “I’ll send you the document”.
Without context, that expectation I hinted at in the last blog – Think UX Isn't Crucial...– means nothing. Context gives the perspective for starting, and is a checkpoint for conclusion. Here is a blueprint you can use to do this research in your shop, starting with your stakeholder.
We have a group of analysts who each search through 100 GB of data a day looking for credit card fraud. They have to generate detailed analytical documents with alerts for our regional managers that are delivered every Wednesday afternoon.
- Credit card fraud analysts
- Big Data
- Weekly analysis reports
- Regional managers
- Alerting capability
Our technicians are having a very difficult time figuring out if the best course of action is repair or replace. We need them to diagnose a problem in two minutes, and then get the correct advice so they can make the decision that best solves the problem while reducing our overhead.
- Repair shop technicians
- Parts inventory
- Part wear and tear history
- Predictive and prescriptive analytics
- High degree of accuracy
- Loss prevention
Our head of research travels 250 days a year. We need a dashboard application that he can use on his smartphone to keep up with very specific results from lab work from our six international facilities.
- Senior VP
- Clinical trials
- Web-based, mobile ready
Each of these context examples would now allow you to get started with a focus towards the best solution. There’s more research to do, but now you know what research needs to be completed.
"Can I see your data dictionary?" is another up-front question that must be asked, especially when the “context” question falls flat. And it’s no one fault. Some shops are just farther along the curve than others, but it is inevitable if your UX project is going to succeed because you need to have a full understanding of what specific data points are called and what they represent.
You’ll find more often than not, that one division refers to, let’s say Data Widget 45X, as “Data Widget 45X”, and another division calls it “Data Thingy 45”. I know it sounds like silly semantics, but these are real data problems that have to be ironed out before proceeding with any kind of design effort.