As a business manager, it's useful to understand the different kinds of analytics. Descriptive analytics are the most familiar kind: they describe what has happened. So, for instance, I have dashboards that show our monthly financial results—revenues, expenses and profit—and sales results. I can compare these results against our targets to ensure that we're meeting our goals and to flag performance issues that need further investigation.
That same descriptive data becomes even more powerful when viewed historically. I track our finances and sales monthly, getting trends of our performance over the past month, quarter, year and beyond. I can compare our results against prior periods, to understand rates of change and to recognize the impact of seasonality.
Sometimes, however, the results vary significantly from month-to-month, making it hard to pick out trends. For instance, we have quarterly cycles in our revenues, which obscure the longer-term trends in our performance. So, I use trailing 12-month and 3-month averages (T12M and T3M). These rolling averages help smooth out short-and-long-term variability, so it's easier to spot the actual trends.
With trailing averages, our analytics start moving from descriptive to predictive. Rather than showing us past performance, we're able to start forecasting the future. Individually, the trailing averages stabilize the trends. By overlaying long (12-month) and short (3-month) trailing trends, you get forecasting magic. The 12- and 3-month trends, respectively, are less and more sensitive to recent changes. When your 3-month revenues are above the 12-month revenues, you're moving up. But if the 3-month drops below the 12-month, it's a signal that your performance is moving down.
Predictive Analytics and Beyond
True predictive analytics move beyond forecasting. Whereas a forecast predicts the overall direction, predictive analytics identify which individual elements are most likely to cause that forecast to come true.
Forecasting will tell you what overall performance to expect from your sales team; predictive analytics will tell them which individual prospects are most likely to close. With predictive analytics, we are starting to move into the world of artificial intelligence and machine learning. In that realm, you also find prescriptive analytics, which tell you what to do to achieve your result. That level of sophistication is beyond what I've achieved so far, but it's coming soon to all our futures.
Commercial Lines Analytics Overview
Policy Works introduces the first out-of-the-box suite of commercial lines analytics that incorporates policy and workflow data.