I echo Jeans point here, based on experience that I have watched with Banks misguided attempts in this approach.  It usually stems from some internal research that proves the usual stuff about 80% of profits emanating from 20% of customers.  In fact its often more like 95/5.

Recently UnionBanCal announced that they were moving away from pursuing a broad consumer base in order to go after more high-end, profitable clients and small businesses (American Banker, January 9, 2007). However, I have my doubts that shrinking the retail base will actually help UnionBanCal grow its deposits.

Source: Javelin Strategy and Research » Focusing on high-end consumers may not always pay

The problem is that its not as simple as that for several reasons:

• customer loyalty is not necessarily associated with the accounts within the 20%
• accounts in the 80% have unknown/ unseen connections to the 20%, and this complex relationship is not understood.  We need better data. 
• finally, Bank costs and distribution are scaled to reach the mass audience (otherwise this discussion would not be happening).  To take that mass distribution, and refocus it on the dramatically smaller account base constrained in the 20% means loss of revenue from the 80%, that was supporting the distribution, by at least covering costs.

Taking an analytical view of account by account does not properly represent the dynamic of account holders, and customers with the Bank.

Don’t get me wrong – focus on a segment is a good thing … if you build your business model that way from inception.  To make the shift midstream requires a significant organisational adjustment.