The industry made what seemed like a very logical assumption: the shift was obviously caused by the incredible crossover appeal of Country superstar Garth Brooks, who had recently taken the music industry by storm. They also took very predictable actions in response: several promising rock bands were dumped while resources were shifted to other country acts.
In the short run, these actions seemed to reinforce the trend, with even more country music sales. But then something very strange happened: the sales numbers slowly drifted back to the exact pre-Garth Brooks percentages, with rock being preferred by the same percentage it had for decades. Industry analysts were left scratching their head. What just happened?
What they found after some analysis was surprising. In March 1991, the industry began counting record sales using the Nielsen SoundScan system. Before that, sales were counted by calling stores across the U.S. to collect sales data – an incredibly ineffective collection method. Unfortunately, not all record stores were able to implement the SoundScan system immediately and continued using the old method for months or even years. On the other hand, one behemoth was online immediately: Wal-Mart. In the early days of SoundScan, every single time a Wal-Mart sales associate scanned a CD, it was counted by SoundScan and reordered, whereas record store sales (and reorders) were hit-and-miss.
Here’s the thing that nobody had thought about before: in 1991, country music fans primarily bought their music from Wal-Mart and rock music fans primarily bought their music from record stores. Once all of the record stores were online, it became clear that the appearance of a shift in preference was nothing more than a measurement data collection problem.
The lesson to this story is that it is critical to resist the urge for a knee-jerk reaction to data such as dumping promising rock bands! There is a process discipline to performance analysis and improvement and the steps are simple. First, a Smart Chart should be used to make sure you are correctly interpreting the data. Then, a root-cause analysis is in order to understand why you are getting the results you are getting. This root cause analysis would have likely revealed the issue with the data in SoundScan being dominated by Wal-Mart sales. Finally, an improvement action plan is implemented and the results are monitored over time.
To learn more about how to interpret, report and react to your performance data, see the KPI Professional Certification Workshop, or see The Institute Way: Simplify Strategic Planning and Management Using the Balanced Scorecard.