Posted by Milos Sugovic
2009 is long gone and the results of your Q4 PR efforts are finally in. Let’s say you had 100 media placements during this time period and you want to know how you did. Before you even read the executive summary of the PR measurement report, what can you expect? The good, the bad, or the mediocre?
The correct answer is: all of the above.
But many of us fall in this trap where our expectations are shaped by extreme events: “We had a great hit in the NYT this quarter. That means we did great!” or “The WSJ article ripped us apart, we are dreading the report findings.” We tend to focus too much on one or two good or bad eggs out of a whole basket, and wrongfully so. Now don’t get me wrong, it’s important to identify and be mindful of the outliers, but if you’re measuring the overall success of a PR initiative, it’s the central tendency of the effort that really matters.
To illustrate, let’s go back to my initial question: If you had 100 media mentions in Q4 of 2009, what should your expectation be? I can bet you that, on average, you’ll see the following: 16 poor hits, 68 mediocre hits, and 16 exceptional hits. How do I know this? It’s not magic; it’s called the normal distribution.
In statistics, the normal distribution is a bell-shaped curve that describes, or at least approximates, any variable that tends to cluster around the mean (or average). So, if PR measurements are designed to do just that – measure PR outputs – one would naturally expect that some media placements “underperform” vis-à-vis expectations while others “outperform” expectations, with the remainder clustering around the average.
If you see anything other than this pattern it should automatically raise a red flag. Measurements that do not look like a normal distribution may be inaccurate and that’s enough to be a cause of concern, especially if they’re used to make important strategic and tactical business decisions.
So the next time your PR measurement team says, “You did great this quarter!” ask them to see a frequency distribution of your media placements. If it doesn’t resemble a normal distribution at all, their yardstick is seriously crooked.
But if it passes the test, you’re in for some insightful findings. You’ll be able to not only see the central tendency (average) of your and your competitors’ media placements, but its spread (standard deviation) as well. This will tell you how well, or poorly, you’re doing – on average – and to what extent there’s under- and over-performance. Take that a step further and track it over time, and you now have a few interesting trends to examine and/or explain.
At the end of the day, PR measurements are not immune to the “laws” of statistics. And if they seem to be, then there’s a pretty high probability that they – to put it nicely – belong on the lower end of the bell-curve.