I have a theory that the two best audiences for online retail stores are the urban core and rural areas.

Urban core areas do well because people living in high density areas tend to get everything delivered. Not just pizza, but groceries, beer, and hardware. Parking, commuting, or even owning a car can be difficult, so it's easier to hop on the web, whip out the plastic, and watch stuff show up at your door.

Rural people benefit from online retail due to a lack of choice. There are few interesting stores within easy driving distance. When faced with a choice of making a list of things to pick up on a shopping excursion, or simply buying the same stuff online, the choice of moving online becomes pretty obvious. There are some exceptions for people who's bandwidth is so slow they'd still rather drive 20 minutes to shop at Target.

I've attempted to illustrate this with a color coded map of the greater Chicago area:

The red zone is what I call the online retail dead zone. That's obviously going a bit too far, but the point here is that suburban families have relatively easy access to a wide array of shopping options and easy parking, so are less reliant on retail delivery than those further in or further out.

If you're an online retailer with a finite amount of money to spend on advertising, it's possible than you may see higher conversions from outside the red donut. This assumes that you ignore all other variables, such as which market is the best fit for the products you're trying to sell.

Actually targeting this type of an audience online is something I haven't fully worked out yet. Any suggestions?