Remember the old Reese’s peanut butter cup advertisement? Two people are walking down the street – one eating peanut butter, the other chocolate – they run into each other and the chocolate and the peanut butter get mixed up. They sample the new concoction and marvel at the wonderful mixture of flavors.
“Two great tastes that taste great together” was the tagline.
Just like SAS and JMP! Two great analytic tools that work great together. To illustrate the point, team members from SAS and JMP (a SAS division) held a free one-day seminar at the Hartford, Connecticut regional SAS office entitled “JMP for the SAS User.” We had excellent representation from a number of industry verticals, as well as a broad range of SAS analysts and statisticians who were all interested in understanding where JMP fits within the SAS family. Since this is a SAS blog, I’m going to assume that our readers may not be familiar with JMP. So here’s the scoop:
- JMP is a powerful desktop data visualization and statistical analysis application. The JMP division was founded by SAS co-founder John Sall, who is also an executive vice president of SAS. John developed JMP (aka “John’s Macintosh Program”) more than 20 years ago, and the product has a loyal following among a diverse group of industries. In addition to its robust data visualization capabilities, JMP is widely used for design of experiments and visual Six Sigma.
- What differentiates JMP from other toolsets in the business visualization category is the advanced statistical analysis capabilities. In fact, in my past role at a major insurance company, we selected JMP as our preferred visualization tool because we recognized the importance of increasing the analytic competency of business analysts in our organization, not just the ability to create more attractive graphs and charts.
In a presentation at the Connecticut event on creating and maximizing analytic bandwidth, we began to put this into context. With our audience, we put together a definition on “analytics” and “analysts.” We discussed the skills analysts needed and the immense variation in skills that businesses are asking analysts to have. We talked about “the decision-maker lens” – communicating and presenting information in a variety of ways to influence decision-making both internally and externally, and finally how JMP plays a role in helping analysts visually interact with data and communicate results to multiple classes of decision-makers.
I noted that the “80/20” rule of percentage of time spent on data preparation versus analysis is also out-dated because we haven’t carved out any time to focus on the effective communication of the analysis! A question I posed to our audience was: Do we do a good job at getting the right information into the right format for our decision-makers? And if not, is it an organizational priority? And finally, do we even know what information will be relevant?
What’s the typical output of a standard business analysis process? We prepare lots of data (most likely in SAS) and put the output into large spreadsheets of information that don’t tell us anything. We send these spreadsheets out to large groups of information consumers – but we don’t know what’s relevant, we don’t know how to make the information actionable and we often lose sight of (or don’t even know!) how our end-consumers need to use the information (or if they’re even using it at all!). Analysts should be empowered to find trends, correlations and causality in conjunction with their business partners; and we need to find better ways to present and communicate that analysis.
We’ve got so much data at our fingertips that there’s no way that we can make sense of it through merely aggregating and transforming it. While the core SAS technologies provide an excellent platform for preparing data and performing statistical modeling, it’s not so great at helping business analysts visualize business problems: JMP can be a great tool to help “make the invisible visible.”