Author: Michael Stus
Some time ago I read an article by Inc. columnist, Norm Brodsky, discussing unique indicators or as he dubbed them, magic numbers. Magic numbers are described as particular statistics discovered by business operators to have special correlations that can be used to predict future business performance. Examples in the article include a restaurateur who predicts his nightly receipts by the length of wait for a table at 8:30 p.m. and Mr. Brodsky himself who uses the current number of new boxes to forecast a period's sales within his records storage business. What makes these magic numbers extremely valuable beyond their correlation with traditional business performance measures is that they are easier to obtain and/or available well before their lagging counterparts. Once identified and monitored, the numbers can allow earlier more accurate business operations decisions. In the referenced article, Mr. Brodsky describes how he, using the afore mentioned indicator, slowed his company's hiring rate well ahead of when he had the quarterly sales totals which he previously used to drive this decision; this allowed him to not over hire rather than being forced to lay off personnel -nice!
This concept is not unique to business. Doctors use indicators like blood pressure, cholesterol test, blood cell counts that have shown to correlate with certain conditions. Using these statistics they predict and take steps to prevent health issues before other symptoms may present. Car owners and mechanics may watch miles per gallon or RPM values in the same manner. How does any of this relate to EIM / SPM systems?
First, EIM / SPM systems, because of their early access to order and customer account data, are a gold mine for finding magic numbers that correlate with sales or other key business statistics. You may find that the commissions system's transaction count on a particular day of a pay period correlates with that period's eventual sales commission payout total or perhaps an increase in the percentage of sales of a particular type of product is a leading indicator of an upswing in your company's market.
Additionally, there are likely magic numbers that can help assess the health of your incentive compensation plans or the EIM / SPM system. These can help IT, administrators, and compensation analysts make EIM / SPM related decisions and adjustments earlier than they might otherwise. The number of sales representative disputes per week might correlate to sales plan health or the number of credits awarded per sale may help administrators detect system inaccuracies before your sales reps or auditors find them.
So how do you go about finding the appropriate magic numbers? Well obviously they should be easier to obtain and available earlier than what you hope they will predict (otherwise what would be the point?). Beyond that, you will need to become very familiar with your system and / or business and watch candidate numbers over time, sometimes a lot of time; you shouldn't hope to find valuable indicators until you have adequate history to test correlations. Brodsky describes a process of understanding "the relationships between the numbers" which sounds to me like relying on your gut. The statisticians among you might suggest ANOVA or other tests of correlation as magic number discovery techniques; unfortunately (or is it fortunately?) it's been far too long since my last stats class to address these. Whether it's via gut feel or a more scientific approach, hope you find some magic!