In 2002 I spent two days in Oslo: June 21 and December 16 -- the shortest night and (almost) the shortest day of the year in a place where the contrast is dramatic. In the summer the sunset never turns to night, and just before Christmas, black starts at 3:30 p.m. and doesn't turn back to twilight again until 9:00 the next morning. I was there both times to help European companies explore the benefits and methods of building share of customer and customer equity.
Today is the longest day of the year in 2011. Where am I this year? In Bentonville, Arkansas, at the Wal-Mart Analytics Day sponsored by SAS Institute and Deloitte, with 500 attendees from across Wal-Mart companies. (We're meeting in the auditorium at world headquarters for Sam's Club.) Wal-Mart has been a controversial company, but I can tell you that up close, the folks here believe the customer is number one. It's even part of the "cheer" that the hundreds of people here did. Loudly. The sign on the wall here quotes the mission for Sam's Club: "Simplifying Members' lives by helping them make smart choices." (Disclosure: I've been a member for at least 20 years.)
In a fascinating keynote, Colleen McCue, senior director, social science and quantitative methods, for GeoEye, talks about using analytics to figure out where the risk areas are for crime in Richmond, Virginia, on behalf of the police department there. She's been able to help police put beat cops in the right place at the right time, which significantly reduced the number of homicides and robberies and increased the number of assault weapons seized.
Her suggestion to the rest of us? Start by understanding rare and independent events. Think about it: The vast majority of commercial and private vessels sailing through the oil trade routes in the Middle East get through fine, but the ones we need to understand and predict are the one third of 1 percent that are successfully "shipnapped" by the pirates. As she points out, she can't ask why nearly every citizen in Richmond made it home safely last night without getting robbed or killed, but she can use analytics to predict where the problems will arise. She has been asked to help predict where the next natural disasters will hit and how the likely human response will affect the long-term outcome, and I honestly would like to know why the Pentagon hasn't asked her where the next suicide bomber or 911 attack will be.
One of the keynote speakers is one of my favorite contributors to better thinking about how to make better decisions through improved analytics: Tom Davenport, Ph.D., whose new book is Analytics at Work. Davenport describes "descriptive analytics" as the what, and "predictive" and "prescriptive analytics" as the so what. He emphasizes the need to embed analytics in business processes.
Davenport makes the case that analytics occurs at all levels of decision making and summarizes those levels like this (from lowest to highest):
- What happened? (Standard reports)
- How many? How often? Where?
- What exactly is the problem?
- What actions are needed?
- Why is this happening?
- What happens if we try this?
- What will happen next?
- What's the best that can happen? (Optimization)
According to Davenport, and Harrah's Entertainment CEO Gary Loveman, Harrah's measures everything. As a result, the company found a high correlation between "smile frequency" and guest satisfaction (so employees try to increase the number of times per visit that guests smile). The only thing with a higher satisfaction correlation was "winning," which, for the record, they decided not to optimize. Loveman is a big believer in careful analytics and research methodology. According to Davenport, there are three ways to get fired at Harrah's: 1) cheating or stealing; 2) harassing women (apparently, it's OK to harass men); 3) not using a control group.
Davenport emphasizes the importance of developing an analytical culture, embedding analytics into key processes, and using analytics to make sure the company makes better decisions systematically. Sam's Club has started this process with the work being done by Linda Vytlacil, Ph.D., vice president of member insights and innovation, our colleague from our days at Carlson Marketing, and Dan Thorpe, senior director, advanced analytics in insights and innovation, formerly senior vice president of statistics and modeling at Wachovia, where he was named a 1to1 Customer Champion. Davenport emphasizes how even minute decisions improved by an assist from analytics can make a difference and cites Stanley Tools, whose executives say that looking at all the inputs for pricing made it possible for them to acquire Black & Decker, not the other way around.
The panel on "Analytics at Work in Wal-Mart" featured Cindy Davis, executive vice president of global customer insights at Wal-Mart; Rick Webb, senior vice president of global business processes at Wal-Mart; Karenann Terrell, executive vice president and assistant CIO for ISD at Wal-Mart; and Tom Waldron, senior vice president of Sam's Club People. The moderator was Gene Gsell, vice president and general manager, SAS Retail. The discussion was chock full of "aha" moments driven by surprising findings that came from the analytics these guys do. Davis has only been in her job for a month, and her challenge is to understand each customer across Wal-Mart, not just in the silos of each of the different businesses. She challenged the audience to help her understand: What is it going to take to understand the data we have, and how do we infuse those insights into the decision making? She calls it insight logistics. The panel members are authorities on analytics and insight, and they all emphasize the importance of making the reporting simple and usable. (Amen.)
Davis concluded by saying, "I have a dream. I dream that every executive who gets up at 4:30 in the morning to check the product sales numbers [will] get up and first check the customer metrics. The company that does that will be unstoppable."
My own keynote address was about how product-centric companies become customer-centric companies, how Return on Customer must accompany ROI when resources are being allocated and success is being anticipated, and how social media and the resulting transparency mean every company has to be a lot more trustworthy than companies could get away with 10 years ago. These are subjects for another day (and for the new book Don Peppers and I are writing now, called Extreme Trust).
Meanwhile, it'll be a short night. Get some sleep.