This one has been a pet peeve of mine for a while, so you’ll have to endure my ranting.  As we head deeper into 2009 and more and more data points are released we, as users of the data, need to be careful of relying too much on year over year comparisons.  I can’t tell you how many articles I read where some statistic is 2008 numbers are down XX% from 2007 and YY% off 2006 numbers.  OK, it makes for an attention grabbing headline, but what does it really tell us?  Not much.

I think we can all agree 2008 numbers probably aren’t going to be very good (in any industry).  But we have to keep in mind that 2006 and 2006 were very good years for most businesses.  Any comparison between 2008 and 2006 & 2007 is going to be negatively skewed, even more so that the data would suggest.  I could be down 15% from 2007, but still be exceeding an 8% compounded annual growth rate over the past 5 years.  Does that mean my business is suffering badly?  No, it means I’m doing quite well. We need to take a longer term view and put 2008’s numbers into a complete historical context to fully understand their implications.  It may not sell as many newspapers or generate as many clicks or diggs, but it will allow for a more meaningful interpretation of the numbers.

Any statistician will tell you that you need to throw out the outliers before you begin your analysis.  For many industries and many comparisons, 2006 and 2007 were outliers and not indicative of long term sustainable growth assumptions.