Data warehousing in the telecommunications industry
The telecommunications industry offers a wealth of opportunity to those who take on the challenge of providing it with data warehousing capabilities, but the data storage and analytical requirements can push the limits of current technology. Few markets are moving more rapidly than telecommunications. Some of the hottest products--and in fact, the hottest companies--did not even exist a year ago. The basic identity of this market has undergone profound changes.
Three forces have converged to propel telecommunications into one of the most competitive business arenas of today's market: technology development, user demand and deregulation. Ten years ago, the heart of the telecommunication industry was long-distance service. Now an array of new services ranging from paging to cellular phones and wireless Internet are technically feasible. Users have responded by signing on and demanding ever-increasing quality (not to mention bandwidth). Deregulation has allowed competition to open up even in such traditional areas as local phone and cable service. It has also prompted a series of mergers and acquisitions among both small firms and the industry giants.
All that leads to one bottom line: a data management nightmare. The large, long-established firms have legacy data on mainframes that may need to be integrated with data stored somewhere else on a series of new products. Worse yet, data from a small company bought to provide a new product may need to be combined with that legacy data. Newer, smaller firms have a simpler integration task, but face limitations in resources. Yet the mandate to "know the customer" means that no matter how challenging, the task must be achieved or the company will not survive.
"Putting the customer at the center of the business is one of the key trends in the industry," said Stewart Meyer, telecommunications industry analyst for MicroStrategy . "The best way to do this is through a data warehouse." He also noted that, as in other industries, there is a need to integrate customer input from numerous channels, including call centers and the growing number of online transactions. Each service representative should be able to see a full history of the customer's interactions with the firm. New transactions need to be fed back into the database. Although those issues are similar to the ones faced by many companies implementing data warehouses, the telecommunications industry does have some unique requirements. The most outstanding one is the sheer volume of data.
"Some switches are producing millions of records an hour," said Rick Plaster, a technical consultant for Perspective Technologies. , a MicroStrategy business parter, is an experienced implementer of data warehouses for the telecommunications industry. Plaster drew a comparison between large companies, which were doing sophisticated data analyses before the newer software tools were available, and smaller companies that are focusing on the simpler task of data consolidation.The call volume translates into major storage requirements, among other things.
"Two years of call detail records (CDR) can require 100 terabytes," said Rich Peters, manager of telecommunications industry solutions at FileTek . implements call detail warehouses for large telecommunications firms as well as financial, retail and government organizations. The volume of data also can make it inaccessible to the decision-making process if it is not properly managed.
"The firm is trying to get a certain amount of 'stickiness' with the customer in each interaction," said Peters, "and to know ahead of time which of the many available services that customer is likely to buy."
The well-established goals of customer retention, cross-selling and upselling can be effectively supported by viable data warehouse and analytical tools such as online analytical processing (OLAP). Those are among the primary customer-related applications for data warehousing and OLAP. For example, telecommunications firms can examine the CDRs for calling patterns that might indicate that an additional service would be of interest. Customer satisfaction is at the heart of those efforts.
Clearnet Communications , a digital wireless communications company, has a strong incentive to keep its clients across Canada happy. Clearnet clients are not locked into binding long-term contracts--unusual for the Canadian wireless industry --and therefore can change service providers any time they choose.
Clearnet integrated its information into a data warehouse and uses Cognos' PowerPlay and Impromptu and other tools to analyze its data.
"We need to keep our client satisfaction high," said Clearnet's decision support systems manager Victor Antunes. "We are now bringing in data at the site level to analyze performance and, by next year, will do so at the client level."
Antunes pointed out the importance of designing a flexible system that can adapt to the fast-paced communications world, and of allocating sufficient resources to the maintenance of the data warehouse. Clearnet is using its system successfully to keep its "churn" rate (the level at which clients leave their service provider) at low levels.
Combining customer sales data with external data can lead to some sophisticated marketing strategies. For example, one firm used social security numbers (which indicate the location in which an individual was born) in combination with the individual's present address. Those who were living away from their birthplace were targeted as candidates for a long-distance plan. Companies can also target their high-profit customers for retention efforts, maximizing the use of marketing dollars. Customer data can be used to develop a profile for identifying high-probability new customers, leading to more effective marketing strategies. Well managed data reduces the cost of acquiring new customers, particularly if it is quickly made available to all the departments that need it, regardless of the channel through which it was acquired.
Warehouse data allows a company to anticipate possible problems and respond proactively in a variety of ways. For example, dropped calls represent not only an obstacle to customer satisfaction, but also a possible indicator of the need to expand the network. Analysis tools let a company determine whether there is a correlation between those two factors. Knowing the location of customers relative to network resources can help determine where capital investments should be made so that they have the greatest benefit.
Some of the strategies for managing the large volume of information apply to any industry where the quantity of data is an issue. If only a portion of the available data is needed for key analyses, one strategy is to populate the warehouse only with the detailed data needed for those analyses. In certain cases, summary data is sufficient for the warehouse. Another option is to have all the data in a warehouse but set up data marts for focused applications. Each of those approaches has trade-offs in terms of speed and availability, and each requires careful planning to ensure the system accomplishes the desired goals.
In another move toward enhancing customer satisfaction, some telecommunications firms are beginning to "push" analyses out of their systems to the customer. That allows the customers to understand their own business better and provides a value-added service. As data warehousing and OLAP become more widespread, we can expect other similar extensions of their use. In today's competitive market, telecommunications providers want any edge they can get.