There are several types of data issues that these tools might identify, and these issues must be resolved, or the suspect records must be eliminated from the “usable” list:
• Invalid address, based on either CASS or NCOA.
• Duplicate record. For example, consider a company with IBM as a primary customer. Its newly merged database might hold 40 different IBM sites. However, for a number of those sites the only difference could be the way the name is represented (i.e., for some sites, the corporate name is IBM, while for others, it’s Inter-national Business Machines, Int’l. Business Machines or I.B.M.). These inconsistencies must be resolved. In this example, while you thought you were starting with 40 unique sites, you could end up with only 15 unique sites.
• Invalid company, based on inability to match to a valid DUNS number.
In going through this process, a typical Massini Group client might start with 150,000 records spread across 10 to 15 data sources and end the process with 50,000 usable records. While this result could be considered disappointing, it is better to concentrate future investment in 50,000 known valid records than to spread it over 150,000 records, wasting two-thirds of the investment.
The result of this initial set of steps is that an unstructured set of data is transformed into a usable database with a single format, in which all relationships are connected and each usable record is verified to offer a high probability of sales success.
Identify the Customer Universe
The data are now well structured and consistent. It is safe to presume, however, the database remains incomplete. Based on Massini Group’s experience working with Fortune 1000 B-to-B technology companies, the sites in a B-to-B enterprise’s marketing database represent only 25 to 40 percent of the potential sites in its target universe.