The SAS office in New York City has one of the best views in the city. Sitting on the 47th floor of a skyscraper on 7th Avenue, the Executive Briefing Room overlooks Central Park and it has a highly distracting view – especially on such a nice sunny day. What a great setting for the next stop on the JMP for SAS Users tour! What is that? Well, I’m glad you asked…!
JMP is SAS’ visual statistical discovery tool. It’s a really neat (and powerful) desktop application that provides anyone from a business analyst to a statistician a way to visually explore data and uncover hidden relationships. Its agile approach to exploring data through visual tools and graphs provides an excellent complement to SAS’ “heavy lifting” capabilities (and helps you avoid SAS/GRAPH code, if that's not your thing).
So, our JMP for SAS Users day is geared towards educating SAS programmers, analysts and statisticians on how JMP can help perform and communicate analysis more easily. The roster for the day included presenters and practioners from JMP and SAS as well as guest presenter, Wayne Levin, from the analytic consulting firm Predictum. The enthusiastic (and overflowing) room of attendees provided for an engaging day of learning and sharing. I’ll cover some of the other sessions in subsequent posts, but there is so much good stuff, it’s better to start at the beginning.
My good friends at JMP invited me as a guest speaker to kick off the day, and in many ways, lay the foundation or context for the session (the “what” and “why”) – the changing role of the analyst, the need for better tools and methods to communicate and present information - to complement the “how” you make that happen.
In my roles over the years, I typically have helped organizations figure out (1) how they can use analytics to help transform their business and (2) what to do with the analytic after it’s been created (I rely on all the really smart statisticians to do the actual analysis itself). Unless you can put all of those pieces together - the purpose of the analytic, the creation of the analytic and the deployment or communication of the analytic output – you won’t be successful with your analytic initiative. How do we close the gaps and link all of these critical elements together?
First, let’s look at the 6Ws. I recently read Dan Roam’s book, The Back of the Napkin, which provides a great framework for problem solving. He reminds us that we should approach any initiative by answering the 6Ws: who, what, why, where, when, how. By answering the 6Ws, you can begin to bridge those gaps in the analytic lifecycle. Understanding what the purpose of the analysis is, who the audience is, why it’s important to the business are essential to the qualitative aspects of the analysis – how the insight is presented, communicated and shared. Then the other half of the 6Ws begin to address the quantitative aspects of the data needed to validate problem statement – where is the trend occurring, how much or how many times do we see it, when is it happening. Yet how many analysts in your organization forget to answer the first 3Ws: who, what, why?
Okay, let’s not lay all of the blame on the poor analyst. Most organizations don’t do a good job at helping frame the problem statement (i.e. what they’re trying to solve for), and the analysts are often so abstracted from the problem that they can’t provide that missing context on their own. BUT, if we can get people (analysts, managers and executives) to start demanding (yes, demanding) the answer to those questions, everyone will be better positioned to understand what’s being worked on, how it impacts the business and who the intended audience is (reminder: never show statistics to sales people).
What JMP does is help provide analysts with new capabilities for simply communicating complex information to different levels of stakeholders. I’ve long been frustrated with the old adage around data and analysis: We spend 80 percent of our time on data preparation and 20 percent of our time on analysis. Guess what we didn’t leave any time for? Communication and presentation of results. How many good analytic projects have failed because the results couldn’t be communicated in a way that was easily explainable? A great example is in the insurance industry – you can create all of the fancy statistical models you like to optimize pricing and underwriting, but if the people that sell your products can’t explain it to their customers (your future or current policyholders) on why their insurance rates went up, you’re sunk!
So let’s start thinking as organizations on how to increase our focus on what, who, why; help train analysts to more effectively communicate their results visually in an easy-to-understand format, and give them some good tools to do it in!