RECAP: Mark's Thoughts from the Altamont Predictive Intelligence Summit


Landing on the Sun by Mark Donnelly, Manager of Business Intelligence, Red Door Interactive I recently attended the Altamont Predictive Intelligence Summit held in San Diego at the Andaz Hotel.  There were several well prepared presentations on competitive and market intelligence, financial forecasting, sales and demand forecasting, and leveraging technology. The general theme of the event and many of the presentations reflects all too human responses to data flooding every aspect of our lives.  How the data astute to ignorant deal with this information data can be summed up in a cold war joke.  This is a statement by General Secretary Leonid Brezhnev on the USSR’s strategy to outdo the U.S. moon landing that goes something like this:

“Comrades,” he said, “I have a plan to overtake the U.S. in the space race — you will land on the sun!”

”But Comrade Brezhnev,” the scientists protested, “we’ll burn up!”

”Don’t take me for a fool,” he said, “you’ll land at night!”

This statement may be more humorous in Russian, but whether this statement was made in jest or he was serious, it provides an excellent metaphor for the fallacy of what happens when our understanding of strategy and data becomes decoupled from reality.  I have experienced my share of Brezhnev-esk like moments in business intelligence and analytics.  One experience was when I was working on a call center staffing model that was heavily influenced by multiple overlapping direct mail campaigns.  After several unsuccessful attempts at creating a staffing model that would meet the required service levels, while minimizing the overtime and idle time, a senior manager said that the solution was simple; take the average staffing requirement by day and use the daily average to create a quarterly average, which would result in a single number to manage.  By the time I overcame my shock (curse of knowledge bias), the proposed model could have almost been considered a sabotage, as it all but ignored services levels, and would maximize overtime and idle time.  In essence, he was asking for an average of an average of an average, and saw no problem in doing so.  Smoothing things over might work with peanut butter and jelly sandwiches, but in the world of forecasting, one needs to know the right questions for the highs and lows.  After long deliberations about this problem, the solution was in changing the marketing strategy by spacing out the direct mail campaigns in order to smooth the actual volume of responses. The tsunami of data that is quickly overwhelming modern businesses does not show bias to anyone particularly.  It is drowning the analysts and executives with same ferocity.  Individuals and organizations that can develop the skills to filter out the signal from the noise, and respond with speed, precision, and accuracy will succeed.

comments powered by Disqus