One of the critical factors to a successful data intelligence program is the ability to define key objectives in how to use the information. “Much like any other project within an organization, the absence of a clear purpose and goals can result in issues such as ‘scope creep’ and unrealistic expectations,” according to the AFP Data Intelligence Survey. Fully 28 percent of financial professionals surveyed indicate that not having clear key objectives is a major challenge facing their companies’ data intelligence programs.
According to Graham Wills, RAVE chief architect and statistician with IBM SPSS, any successful data visualization project starts with clearly identifying your objectives, which starts by asking the right questions: Is the goal is to assist in managing global risk and governance, to increase sales, or something else? “Assuming it’s about risk, you’d probably want to know which areas are most risky, or where do things go unreported? Visualization will answer those questions,” said Wills.
A successful image asks and answers multiple questions. “Visualization almost always asks better questions,” said Wills. For example:
- Where do problems occur that are not handled rapidly
- How long does it take for them to be handled?
- What is the geographical location?
“You have to be able to answer the broader questions so people can discover more about the problem,” he said. “It’s a sort of partnership with the user.”
“It’s also important to try to tie in predictive analytics,” continued Wills. “If something happens, what is the total amount of money you’re predicted to lose, and should you be concerned about the average amount or the big amounts that are the outliers? That’s the data you choose to highlight and the context for the analytics. The measure of value or success of a visualization project is whether it helped answer the questions.”