Last week Grant and Tadd (co-owners at CAN) presented their insight on the concept of being “Less Wrong” on day two of the Infotec Conference and left the audience intrigued with the possibilities of the use of predictive analytics in their line of work.
Last week Grant and Tadd (co-owners at CAN) presented their insight on the concept of being “Less Wrong” on day two of the Infotec Conference and left the audience intrigued with the possibilities of the use of predictive analytics in their line of work. CAN has used the “Less Wrong” approach in predictive analytics to predict when roofs will fail, track down graffiti artists, date, predict the S&P 500, and make software go viral.
The basic idea of Less Wrong is that in business, and almost anything in life, you can never be perfectly right, but you can be less wrong and by striving to continually become less wrong you get closer and closer to being right. By using predictive analytics you can analyze data from multiple sources to capture information and determine what’s happening, what will happen, and what is the right thing to do.
A good example of Less Wrong is the CAN Roof Failure model. CAN worked with the facilities department of a Fortune 500 company to help improve the scheduling and budgeting process of their facility maintenance division. Each facility manager was responsible for 500+ server huts and they had no way to predict when the roof on a hut was going to fail. However a single leak could quickly destroy the sensitive electronics below leading to blackouts for their customers. CAN was able to use information about what the roofing material, weather patterns, age of the roof, previous maintenance, location, and other variables to make a reasonable assumption on when each roof was likely to fail. While it was impossible to know exactly when a roof would fail, we were able to provide a reasonable estimate as to when failure might occur thus allowing for maintenance schedules and budgets to be prioritized.