Medici Turns Data Into Single PackageAs database marketing systems become more complicated, a shrinking number of firms can afford full-blown implementations. The obvious question is what systems will be used by everyone else.
Nowhere is the issue more clear than in the granddaddy of all marketing database segments, the mass of small banks, credit unions and savings and loan associations that spawned the original desktop marketing customer information file (MCIF) systems. Although no more than 5 percent of these institutions ever purchased an MCIF, their numbers have always made them by far the largest group of users.
The hope of somehow attracting the rest of the market, combined with the more certain need of existing users to replace their now-obsolete original systems, makes this segment irresistible to system developers. Many have pursued this market in recent years with offers ranging from bare-bones service bureau solutions to whiz-bang technologies at popular prices. No one has achieved mass success.
Medici (Medici Technology Inc., 800/882-9394) represents another attempt at a modern system for low-end customers. Medici's approach is to integrate every possible type of data -- customer, prospect, competitor and geographic -- into a single package accessible to a thoroughly nontechnical user. It is targeting institutions so small they may not have a full-time marketing director. The idea is a system that the bank president can use personally, with no training, and for a variety of applications beyond traditional direct marketing.
The key to this flexibility is Medici's data content. While traditional MCIF systems focus primarily on customer data, each Medici database holds prospects (from the Polk TotaList national household database), competitive information (from Sheshunoff Information Service's compilation of branch statistics) and geographic information (from Geographic Data Technology's street-level geographic database). This is all appended during an update at Medici.
The update also links existing customer accounts into households, applies geocodes and Polk lifestyle codes, calculates profitability and groups customers into four profitability segments. The profitability calculations are fairly sophisticated, using actual transaction costs and fees rather than industry averages or cost of funds calculations. But householding is based only on shared physical location, which means the system won't bring together even accounts sharing the same Social Security number. The vendor plans to enhance its householding logic in future releases.
The final database is shipped to the customer on a CD-ROM. Users see a book-like interface with "chapters" for overview statistics, customers, prospects, competitors, market areas and background information.
The customer section lets users look up information about an individual account and see all other accounts in the same household. It presents a variety of online reports showing service usage, cross sell patterns, market areas, profitability and demographic information.
Each report simultaneously displays data in tables and bar charts, providing an unusually productive analytical tool. The bar chart shows data from whatever row is highlighted and the user can change the sort sequence of a table by selecting different columns. Users can easily print a table, graph or both and can move the data into Excel for additional manipulation.
Reports can show the entire customer base or a user-defined segment. Segments are created by checking location, relationship, product, branch, lifestyle and demographic attributes. This gives considerable power, although more complicated selections, such as those based on user-defined calculation, are not available. Users can extract a customer list showing name, address, telephone number, profitability, deposit value and loan value, but no other elements. They can define the sort order and limit the number of records extracted.
Although the extract file can be stored outside of the system, Medici doesn't retain an internal selection history. Nor does it allow posting of responses between updates, or tracking of changes from one update to the next. These limits make serious campaign management virtually impossible. The vendor is evaluating methods to remove these limits in future releases.
Medici also can display selected customers on a map and show the accounts associated with a specific location, but users draw an area on the map and select customers within it.
Prospect functions resemble customer functions. Current customers are automatically excluded from the prospect list, and a penetration analysis can compare customer to prospect counts by market area. The system will rank prospects on expected profitability or likelihood to use a specific service, using the Polk demographic "Niche" codes that showed the highest profitability or service use among the customer base. Users can analyze prospect data for free, but must pay to select prospect names for a mailing list. The list is generated from the Polk data already stored on the user's computer, which is decrypted after Medici issues an authorization code based on the number of records purchased. Users pay about $70 per thousand, significantly less than the $120-$130 Polk would charge for similar selections if purchased directly.
Competitive analysis shows summary information about other banks and deposit information at the branch level. The market area analysis lets users compare two existing markets, defined by drive time calculations from branch locations. Users also can create up to 20 proposed markets by drawing boundaries on a map. They then can analyze and select customers and prospects within these areas.
Medici was released in the summer and is installed at seven clients. The system is offered directly to banks in the eastern United States and will be sold through third parties elsewhere. Pricing is based on the number of accounts. A bank with under 10,000 accounts would pay $5,000 for setup plus $16,000 for quarterly updates. A bank with 150,000 accounts would pay $19,000 for setup plus $30,000 for four updates.
This system offers an all-in-one marketing tool for companies that cannot afford several separate, more capable ones. But in choosing where to compromise, the vendor has chosen to skimp on campaign management -- even though cost savings and revenue enhancements from targeted marketing have traditionally provided much of the justification for investing in MCIFs. Can a system without real campaign management offer bankers enough value to justify even a relatively low purchase price? Medici may
soon find out.
David M. Raab is principal of Raab Associates, a consultancy based near
Philadelphia that specializes in marketing database systems.