The disadvantage to customer surveys

I recently posted this comment (slightly modified) in response to a RetailWire discussion on the use of satisfaction surveys:

While the instinct to learn from customers and the market is a good one (otherwise we're just sitting around making things up), there are some definite disadvantages to customer surveys. Not only the irritation factor, but also surveys have sample bias (those choosing to take them are either super involved and therefore biased or pissed off about something and therefore biased) and suffer from the distortion that always accompanies self-reported information. That's one reason that more and more retailers are going out of their way to gather and learn from the huge quantity of performance data that stores inevitably generate. These data are not just sales at the register but also information about traffic to the store and behavior within the store. By correlating these data back to such factors as marketing and promotions, staffing, cyclicality, and even the weather, retailers have actionable intelligence they can use to improve performance without irritating their customers or suffering from the other disadvantages listed here.