Arthur C. Nielsen Jr died last month in Winnetka, Illinois, at the age of 92. He joined his father’s small Chicago coupon counting firm in 1945 and grew it into a billion dollar market research conglomerate. “AC” as he was called by his employees, believed that there would always be a thriving business in counting and measuring things for other companies. He didn’t have a strong preference for what was being counted, as long as someone would pay to count it. Grocery store coupons, radio listeners, TV programs, magazine sales, oil well production, it didn’t matter.
AC had an interesting 5-step concept about the flow of information.
- The first principle was that counting/measuring something early, was more important than counting it accurately. If a Texas oil well had a gauge right at the well, that reported the number of barrels pumped every hour, with an accuracy of 95%, it was far more valuable than something that measured what was in the big tank at the end of the pipeline with 99% accuracy, but a week after it left the well.
- The 2nd principle was the accuracy of your counting was important, but the amount you could spend taking that measurement was not unlimited. If it cost $100 to count 10,000 coupons with 99% accuracy, it was never profitable to spend $200 to get to 99.5%.
- The 3rd principle was that consistency was important. If only 22,000 families were being used to represent 100 million TVs, it was important that the families stayed a consistent 22,000.
- The 4th principal was the delivery speed of the recorded information. If a paying client didn’t get their promised information in the hour, day or week promised, it wasn’t worth delivering. Often, the client got a full refund.
- The last principal was archive your results longer than you guessed you needed them, because someone would always pay for an analysis of something you kept.
My first 9 years in IT were spent working at AC Nielsen. I only met him a couple of times, when he was on one of his cross-country tours of the Nielsen divisions. As the company that bought the first IBM mainframe for a private company in 1954 and the first electronic scanner in 1967, it was an interesting environment.