Do you really know your small business customers: emerging big-data analysis can uncover commercial client needs and generate new revenue opportunities.

• Author: Krishna, Bc
• Position: Company overview

D0 BANKS REALLY UNDERSTAND THEIR SMALL-BUSINESS CUSTOMERS? Nearly every bank (except, perhaps, those that specialize in a vertical market, such as Silicon Valley Bank and its technology clients) includes a large swathe of clients that populate the poorly surveyed land between retail consumers and corporate businesses (that is, large businesses with annual revenue greater than $250 million).

Who are these nonretail, noncorporate clients? This enormous population--typically about .15 percent to 30 percent of a bank's clients--are their long-suffering small--and medium-sized business (SMB) customers.

Most banks tend to segment their commercial customers by the value of their annual sales or revenue, and virtually every institution carves up its business clients by some revenue range--for example, less. than $5 million, $5 million to $20 million, etc. However, these segmentation ranges are quite arbitrary and often based on criteria that were rational at one time, but later lost their usefulness. Additionally, the "ownership" of .these individual segments at the institution is always a bit arbitrary and buffeted by the winds of change. Simply put, revenue is a terrible way to segment business clients--and it's especially true for SMB clients. Except in the rare cases when revenue estimates are actually reliable and indicative of their needs, the knowledge gleaned from this single magic number does not help banks truly know their customers. In most cases, revenue-based segmentation fails on both counts--reliability and as a proxy for client need.

Revenue estimates are extremely unreliable

Typically, SMB clients are not public companies. These clients have no obligation to report revenue, are...