Is the data set a sample large enough to be statistically significant?
While traffic may have increased, did shopping increase?
Was the decrease in sales a dollar measure or a conversion rate measure?
Were the data for the three months in question compared to same calendar periods of past years? If not, what seasonal factors might be in play?
Did the product line and prices remain the same throughout the said three months as they were previous to such?
Is there any increased competition, be it new competitors or existing competitors changing their product lines and/or prices?
Have you considered doing a test by reverting to the previous system to see if sales revert to their previous level(s)?
Last edited by deepsand; 12-07-2011 at 03:57 AM.
Last edited by deepsand; 12-07-2011 at 08:49 PM. Reason: merged traffic flow
Aside from the fact that requirements differ according to the merchant's classification level, and that it is the acquirer who sets them for Level 4, the PCI guidelines themselves are open to multiple reasonable interpretations.
Even in the absence of web based analytics, you still have a wealth of information from the server logs available for analysis.
Would not be surprised to learn that the client's perceptions are less than accurate.