Analytics Spoken Here? Finding a Common Language

The term "analytics" has become so overused that I'm not sure anyone really knows what it means anymore. Or at the very least, it means a lot of different things to different people. Sometimes analytics gets blended in with "business analytics," which gets blended in with business intelligence. What do all of these things mean? I realize just how meaningless these terms are when I try to explain to my parents what I do for a living. They gently remind me that I don't make any sense from their point of view (everyone should have parents who tell them these things!). As analytics professionals, it's our job to communicate and educate on what analytics is, how it adds value, how it is used (with real examples) and how to implement it.

For example, one insurer brought its different lines of business together to discuss analytics. Each line of business was at a very different maturity level from an operational perspective (especially for newer lines where the company was not yet making significant investments in the business). For one line, analytics meant basic reporting, because it didn't have any mechanism to tell it anything about the business. For another line, analytics meant continuing to make strategic investments in statistical resources so it could continue to grow its predictive modeling capabilities. And for yet another – further along the maturity model – it was how to embed predictive capabilities in operational systems. This represents three very distinct "analytics" goals. Now if this company hadn't begun to have this conversation, how would it know where it was supposed to make its enterprise analytics investments? Once there was a common understanding, or language, the organization could identify needs, capabilities and investments necessary to ensure it was marching down the same path.

[Previously from Alt-Simmons: Why Asking "What if?" Can Spur Insurance Sales]

But perhaps more importantly, the work that the analytics professional performs needs to be accessible and understandable by users. In effect, analytics need a soul. Some time ago, I had a conversation with an analytics leader who eloquently reminded me of the importance of context in decision making in different scenarios, and told me:

Rachel Alt-Simmons, SAS

When lives are on the line the 'soul' or considered thought of analysis is much more important . There is something that defines us as humans that tells us that the analysis doesn't tell the whole story. Analysts, if they are honest with themselves, understand that all analysis contains uncertainty, either by design, the limits of the data used, or our inability to articulate and account for all the variables. Some senior decision makers have a firm grip on the analysis and some don't. When the issues involve something less than a human life, many senior decision makers will 'go with the numbers.'

As soon as a life or even the quality of someone's life is at stake, the numbers and the analysis alone are not enough. He continued, "One of the mistakes my junior (or new) analysts would make would be to crunch the numbers and present the analysis as the final product. They missed what I called the 'so what' of the analysis. So what? Is that good or bad, better or worse? Do we have a trend developing? If so, what is it we are doing? Is there a correlation between our efforts and this number?"

They didn't have the experience yet to think about the implication of the analysis. They put no soul into it. In spite of everything analysts think about the work they do, the numbers rarely speak for themselves. If you can't articulate what they mean in the context of the organizational mission, what good are they? Thinking through the implications, impact, or "so what" of the analysis requires a little analytical soul.

Another analytic leader within an insurance company told me: "We need a 'soul' to make analytics successful. We recognize that a culture change is needed to move decision making from an art to a science, but we need the right balance of art and science. Our business will not be run by a computer – we need people with a feel for intuitive decision making who understand the business, but we need to give them tools to better inform their decision-making process. How do we evolve this decision-making process?"

An obvious mechanism is to ensure that your business and analysis community has strong analytic skills and a solid understanding of the work that your analytic team is performing on their behalf. This includes helping them synthesize both qualitative and quantitative information in their day-to-day decision-making processes. Like any skill, the ability to think analytically can be taught and nurtured – but in the words of Richards Heuer, the father of the psychology of intelligence analysis, analysis techniques "are not learned by sitting in a classroom ... analysts learn by doing."

This learning and education process is bidirectional. The analytics group needs to understand the business context, and how the insight (whether a predictive model, a report, a self-service analysis tool, etc.) will be used by a businessperson to make decisions; and the businessperson needs to understand how the insight fits into the business process. So if you're rolling out new insight, make sure your analytics professionals spend some time with the business community, and leave plenty of time for training and supporting the business. Putting in place a formal change management strategy will help you communicate the vision for analytics throughout the project so there are no surprises! Collecting feedback along the way (let them kick the tires – find out what's working, what's not working) will help ensure that the insight is relevant and being used appropriately. Give your analytics a soul!

About the author: Rachel Alt-Simmons is the Senior Industry Consultant for Insurance at Cary, NC,-based SAS. Simmons has driven business intelligence initiatives at Travelers and Hartford Life and has been Research Director for Life & Annuity at research firm TowerGroup.