Tyler Cowen lists reasons for longer lines in some stores. I think there is another reason too:
8a) Some stores do a better job of dicing and slicing customer behavior data that lets them develop more sophisticated plans for the number of registers that need to remain open in a given store/location/ season/time/day. One company headquartered in Bentonville has a bit of a reputation about that! Not everyone is equally good. (In one famous example i read recently, Bank Of America researchers found that if they provide cable Television on top of their teller counters, the perceived wait time would go down by as much as 30% (I think). The cost of HDTVs per counter in that geography was much less that the cost of implementing an EAI project that would have reduced the actual wait time. Increase in revenue per branch because of the TVs were projected to to be higher than the cost for the TV purchases. The TVs were in. The EAI project was thrown out)
b) The point about impulse purchases is related to this. The success of ‘impule buy’ merchandising is not only a function of the instinct of the merchandisers / store buyers, but also (I would say more so) of the sophistication of the business analytics applications used by the retailer.
c) Some stores have buggy software or messed up user interface. We go to ‘Stop and Shop’ and ‘Shoprite’ for our groceries. Shoprite’s software has trouble reading discounts correctly and often brings lines to halt. But Shoprite managment deduced (correctly) that its price sensitive target audience in unlikely to walk away. Stop and Shop is at a slightly higer price point and (probably) made the necessary investment in more state of the art POS software
I think retail is one of the last frontiers. Regular Margin erosions and huge cost of implementations have made the majority of retailers risk averse.