Michael LoBue writes: Earlier this year the 13th Edition of the Operating Ratio Report (ORR) was released by ASAE. This edition is a solid improvement over earlier editions. The ORR is an important, no make that an essential, benchmark report for AMCs of any size.
Included in the improvements in the 13th Edition is the reporting of medians, in addition to the usual averages. For those who took “Stat 101” too long ago to remember, median is that value in a range of numbers that is exactly in the middle of the distribution — half of the values are above that median and half are below. So, if the median is greater than the average (aka: mean) it means that the range of numbers contains “values below the median that are pulling down the average”. The opposite is true if the median value is less than the average. Thus, including medians gives us more information about the total distribution of values than the average alone provides. It also includes key performance ratios by organization types — these ratios are:
- Productivity & Efficiency
- Revenue & Expense Management
All this is good! But, we’re still lacking important data in the ORR that will give association managers a much better sense of how our organizations compare to like organizations. For example, it is misleading to compare a single organization to the ORR by revenue class and draw any useful conclusions, or even inferences, about how well our organization may be performing relative to the industry.
Chances are, any organization we compare the ORR metrics will be above or below the average and median. If our subject organization is below, does this mean it is leaving revenue on the table, or under-investing on the expense side? Presumably, deeper comparisons to the Key Performance Ratio will give us a better handle in comparing our organization. But, there is a missing metric that must exist in the data sets that would answer a key question. That question is: “How does my organization compare to the surveyed organizations that occupy the same point on the distribution?” To answer this question we need the “missing metric” and associated data from the organizations corresponding to the metric:
- incremental data points – either standard deviation values or quartile values; and
- demographic characteristics of the organizations at each incremental marker.
For example, if your subject organization is “some-number” in percentage points above the mean (and median), wouldn’t it be more valid to compare your subject organization to the organizations in the ORR that are at or near 8 percentage points above the mean?
Take the following model distribution curve… all the response form the “bell curve” distribution, where 68% of the responses (value) are 1 standard deviation above and below the mean value and 95% of all values in the distribution are between 2 standard deviations above and below the mean value. Assume your subject organization is somewhere between +1 and +2 standard deviations (see red X inside red oval below — wouldn’t it be more useful to compare your subject organizations to organizations in the distribution that were near the position of your subject organization (within red oval) than the general mean for the entire distribution? After all, isn’t the mean the best of the worst and worst of the best — that’s no target to shoot for!
The following list of comparisons should be available from the existing ORR dataset:
- number of members
- age of organization
- scope (e.g., local, state, national or international)
- line of business or profession
- revenue source ratios
In conclusion, this should not be a difficult threshold to reach for the professional management of associations. After all, it’s being taught these days in high school!
This has been cross-posted at: Association Voices.