BACKGROUND: Telef-nica Chile (formerly Compañia de Telefonos de Chile CTC), with 8,200 employees, is the leading provider of telecommunication services in Chile, servicing more than 90 percent (2.6 million) of Chile's phone lines and more than 35 percent of the country's long- distance market. We offer a broad range of telecommunication services including cable TV, Internet access, cellular telephone services and data transmission.
PLATFORMS: Telef-nica Chile uses a combination of Sun Solaris (to host the data mart), Windows NT and Windows clients. The Sybase IQ data marts run on a four-processor Sun Ultra Enterprise with 2GB RAM and 100GB of storage, running Solaris 2.6. The WhiteLight Analytic Application Server runs on a Windows NT Server (4.0 running Service Pack 3 or above) with 1GB of RAM and two CPUs. We use Windows 95 and Windows NT clients.
PROBLEM SOLVED: Telef-nica Chile recently reorganized by customer type (corporate, small business and residence) to better serve our customer base. To maintain our market share, we needed ways to analyze customer records. We wanted to segment the customer base according to behavior in order to better create services and cross-market existing services for the different segments. We also wanted to improve customer retention by identifying those customers likely to switch providers. The challenge was sorting and analyzing the massive amounts of data involved. Our analysts wanted to look at customer information on a daily, weekly and monthly basis for more than 2 million customers. Our transactional systems generate approximately six gigabytes of customer data each month, and we needed a solution that could handle these large volumes of data easily and quickly. We selected a data mart/analytic application solution comprised of Sybase IQ and the WhiteLight Analytic Application Server (Sybase PowerDimensions). We built two data marts based on Sybase IQ: one for customer segmentation and one for customer retention. The WhiteLight Server drives decisions based on this information with applications that support interactive modeling and analysis.
PRODUCT FUNCTIONALITY: Customer information is extracted from transactional systems and loaded into the Sybase IQ data marts. The WhiteLight Analytic Application Server integrates this information into a complete business model so users can test scenarios, view customer behavior by segment or drill down to specific records and predict which customers are likely to switch carriers.
DELIVERABLES: Together with the Sybase consultants, we created an application that business users can access easily. In addition, the WhiteLight Server hosts a number of analytic models specifically based on different aspects of our business. Business users can use these models in a graphical environment to test scenarios and analyze information on an ad hoc basis.
STRENGTHS: The WhiteLight Server's advanced analytical capabilities enable complex analysis on large data sets from multiple data sources, helping us to better understand customer dispersion. The WhiteLight Server and the Sybase IQ data marts provide interactive performance for complex analysis on very large data sets.
WEAKNESSES: Ours was the first telecommunications implementation for the WhiteLight Server. Creating the right model and formulas required some trial and error.
SELECTION CRITERIA: We needed a solution that could handle massive amounts of data and was easy for our business users to use. The WhiteLight Server together with the Sybase IQ data mart solved the problem. By visiting other Sybase/WhiteLight installation sites, we determined that this product architecture provided both the flexibility and performance we needed.
VENDOR SUPPORT: As a reseller of the WhiteLight Analytic Application Server (PowerDimensions), Sybase provided both pre- and post- implementation support.
DOCUMENTATION: The documentation from WhiteLight and Sybase IQ was very thorough. In conjunction with Sybase Chile, Telef-nica Chile developed all the documentation needed for pre- and post-project administration.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access