David Raab Article Archive

Many vendors now offer software to identify marketing opportunities by examining customer transactions. These systems all work roughly the same way: users define patterns of customer transactions that indicate marketing opportunities; the system scans transactions for those patterns; it then sends the resulting opportunities to sales departments as leads. But the products differ in important details. These include:

– automated or manual methods to identify patterns. Automated systems are easier to run and may find unexpected patterns; manual systems are easier for developers to build and for users to understand.

– analysis of summarized historical data or individual past transactions. Individual transactions allow greater precision and flexibility; summarized data takes less processing.

– complexity of patterns the system can track. Complex patterns may involve time series analysis, identifying expected events that did not occur, and relative value calculations that identify what is normal for a given customer. They are more precise than simpler patterns, but harder to define and more reliant on detailed information.

– use of standard relational database or specialized file formats. Standard databases are more familiar and their contents are more accessible; specialized formats can be more efficient for specialized processing.

– hosted or in-house operation. Hosted operation is easier to deploy and often requires less initial investment; in-house operation keeps customer data inside the company and may cost less in the long run.

– end-user ability to change inputs, patterns, and outputs. End-user control allows quick changes but sometimes limits what can be done.

– opportunity ranking based on static priority codes or customer-specific value estimates. Static codes are simple to define; customer-specific values are theoretically more precise but hard to calculate accurately.

– linking of leads to specific marketing approaches such as telephone scripts. Explicit instructions help agents know what to do with a lead, but may deflect them from exploring a customer’s actual situation.

– provision of a lead management tool for sales agents. Many firms lack such a tool; others already have one or would rather purchase a more sophisticated product separately.

– option to simulate the impact of rule changes using past transactions. Such simulation is helpful but adds complexity to the product.

No single combination of features has yet emerged as standard. Perhaps it never will: users with different needs will purchase different configurations. Still, the variety of options makes it particularly important for buyers to examine each product to ensure it matches their needs..

Customer Opportunity Advisor (ASA Corp., 412-220-9300, www.asacorp.com) provides an opportunity extraction product configured for simplicity. The system is built on ASA’s DecisionBuilder rules engine, which allows users to define patterns in terms of hierarchical rule trees. These are built with a graphical interface that is accessible to non-technical users.

DecisionBuilder can generate complex rules with calculations, table look-ups, calls to external programs such as scoring routines, and references to other rules. Although it does not easily support certain types of complex pattern detection, ASA reports most opportunity identification users do not feel these are important. Data mapping and transformation tools can connect with nearly any data source and rule sets can be called as services through a broad range of technologies.

In setting up Customer Opportunity Advisor, users define DecisionBuilder rules to identify marketing opportunities. Each opportunity is classified by type, such as retention or cross sell, and can be linked to text that describes the opportunity or provides instructions to sales agents. The system can also assign a priority rating, using either a fixed value or customer-specific calculations. Rules can limit the number of opportunities per customer over a specified time period–for example, no more than one opportunity every two weeks.

Customer Opportunity Advisor can also distribute leads to sales agents and to help agents manage those leads. Set-up is done by non-technical system administrators, who choose from configuration options, define bank branches, and create accounts for individual agents. Administrators can specify the number of leads sent each day to each branch and to individual agents, with different limits for each opportunity type. Leads can be distributed to agents automatically, by the branch manager, or based on assignments of specific customers to specific agents. The system can also import leads from external sources such as campaign management systems and from referrals made by other agents within the system. Agents can also enter their own leads directly.

The agents interface is straightforward, listing leads and letting users drill down to details. These include the opportunity description, other opportunities for the same customer, contact information, a form to enter call results and sales, and logs of previous interactions. System security limits agents to their own leads and lets branch managers access all leads within their branch. Standard reports show processing volumes and basic sales management information including leads sent, lead status, outcomes, and sales. These can viewed by date range at agent and summary levels.

DecisionBuilder runs on Windows, Linux or Solaris servers. Administrators and sales agents access the Customer Opportunity Advisor interface through any Web browser. All current clients operate the system in-house, although ASA offers a hosted option as well.

Pricing of Customer Opportunity Advisor is based on the number of branches using the system, unusual for this type of software. It is remarkably inexpensive, ranging from $2,000 to $500 per branch per year depending on the total number of branches. Implementation fees are additional, but are fairly small since set-up is limited largely to configuration of existing software options. ASA helps clients establish their initial rules, starting with a dozen standard opportunity definitions. These are very simple, such as “customers with a new mortgage and no other accounts”, and are supplemented by policy rules such as constraints on contact frequency. Clients typically refine the rules and add new ones by themselves soon after deployment.

Customer Opportunity Advisor was introduced in early 2004 and has about a half-dozen clients in production. The product is sold mostly to institutions with 15 to 20 branches, although the largest client has more than 200.

* * *

David M. Raab is a Principal at Raab Associates Inc., a consultancy specializing in marketing technology and analytics. He can be reached at draab@raabassociates.com.