Data Mining and Warehousing:
Many utilities have no ability to turn raw customer information into significant insights about their business.
In the utility industry, any discussion of improvements in either customer or revenue management requires consideration of the critical importance of the customer-information system (CIS). Pivotal to effective customer and revenue management, the CIS typically contains the vast majority of information needed to achieve the best possible levels of customer satisfaction. It also houses much data critical to a utility's operational efficiency. The CIS is the system of record for both customer and meter data, and in combination with asset databases, these information repositories comprise the lifeblood of every utility.
Although CIS modules typically are integrated for billing, account management, meter reading, credit and collection, work orders, rate management, and customer contact, they usually are not tuned for analysis or for easy reporting, and while almost all of the utility's customer interaction, revenue, and work order functions are handled through the CIS, a single view of the customer and the business processes surrounding the customer does not exist in most of the systems in use today.
Many utilities, as a result, are data rich but information poor, with repositories consisting of gigabytes or terabytes of customer data but with no ability to turn this raw data into significant insights on management, marketing, and investment strategies. This is particularly true when the CIS is running on a legacy platform, of which there are many still in use. The convoluted nature of the legacy software and the monolithic architectures upon which old solutions were built are not conducive to flexibility of data access and manipulation. Fortunately, the return on investment of replacing legacy software with new technology based on open, Web-based architectures is increasingly attractive, and as more legacy data is migrated to new CIS software platforms, it becomes increasingly easier to access and manipulate this data.
Enter the Data Warehouse
Utilities seeking to improve both strategic and tactical decision-making while achieving greater efficiencies are benefiting through efforts to create "one-stop" repositories for the retrieval of information, with an emphasis on more fully leveraging the information that's already been created and stored in existing data archives. Here, the focus is not so much on what's taking place but rather on understanding why specific events happen.
The first step in combining archived customer and operational data with emerging analytic functionality is to consolidate into a data warehouse all the data stored in disparate databases. A data warehouse stores large quantities of data by specific categories so it can be more easily retrieved, interpreted, and sorted by users, thereby enabling executives and managers to anticipate and respond faster to market situations and make better-informed business decisions.
Some business forecasts predict that every business will have a data warehouse within 10 years. But merely storing data in a data warehouse does a company little good. That's where data mining and customer analytics come in.
Data-mining tools enable user-defined analysis of data based on a robust data warehousing framework that provides a single view of the customer by combining customer information, account management data, and information on installed services, meter reading, billing, payment, and collection.
What Is Data Mining?
Data mining provides tools and techniques that add intelligence to the data warehouse. It derives its name from the similarities between digging through and extracting meaning from information in a large database and mining a mountain for a vein of valuable ore. Both processes require either sifting through mountains of material, or intelligently probing it to find its hidden value.
Data-mining tools use pattern recognition technologies, along with statistical and mathematical techniques, to sift through warehoused information and unearth significant facts, hidden patterns, relationships, trends, exceptions, anomalies, and predictive information that otherwise might go unnoticed. These new analytical tools can answer important business questions that, until the advent of these tools, were too complex and time-consuming to consider, much less resolve.
Currently, data mining and the associated use of customer analytics and business intelligence software are more prevalent within the European utility industry than in North America because of the higher level of competition under way (and thus the more pressing need to fully understand customer buying and usage patterns). For example, to survive a price war, German power company Hamburgische Electricitats-Werke AG (HEW) implemented a data warehousing project followed by a data-mining project to analyze customer acquisitions and better position itself to compete.
At Electricité de France, the French national electric power company is using data mining to better understand and predict electric power load curves of individual customers, and to characterize records that fail consistency checking of the data warehouse. This makes it easier to produce accurate reports even when data is missing, through the use of statistical adjustment.
Patterns That Lead to Better Decisions
Using data-mining tools, utilities also can analyze any correlations between customer profiles and payment histories. Analysis capabilities can enable users to query CIS data in an manner to identify patterns and analyze root causes for key issues like delinquency and slow payments, thereby increasing the potential for collection. Similarly, data-mining techniques have been used to develop systems that can detect fraudulent credit card transactions in near-real time.
Predictive models can take the uncertainty out of projecting timing and magnitude of utility consumption and thus facilitate and lower the risks of load forecasting, so that more accurate load information can be used to negotiate better rates. Many utilities that collect and maintain meter data reflecting actual energy usage for individual consumers to support accurate customer billing are installing advanced metering technology to enable time-of-use and real-time pricing rate programs. This data can be mined as a means to improve geographic load forecasting and subsequent targeting of energy efficiency and demand reduction programs, or it can be mined by utilities and state regulators to target specific regions or major customers (by class or individually). Data analysis could identify promising tradeoff opportunities to mitigate price and supply volatility in commodity markets, and to provide unprecedented capabilities to more precisely target public energy management programs. It could also help competitive energy service providers in emerging retail markets support efforts to open retail markets more quickly.
Furthermore, data-building management and metering systems can be aggregated and analyzed to spot trends, issues, or problem areas and to identify energy-saving opportunities, which can be posted on secure Web sites for 24/7 access for everyone from executives to maintenance personnel.
Data mining also enables utilities to better meet operational and regulatory reporting needs. Utilities are increasingly required constantly to monitor their information and practices and benchmark their performance against competition. They are expected to analyze existing customer information and apply measurement criteria to improve the scope of the services provided to the customer, improve the cash-collection process, streamline workflow practices, carry out management change processes, and improve customer communication.
Moreover, using data mining, those responsible for customer relationship management quickly can access information about customer consumption, revenue, cost, and service disruptions. As a result, they can gain greater insights into what customer segments are most profitable, the demand profile for each segment, which services and products are being bought by various segments, and the associations among these services/products and consumption patterns.
Articles found on this page are available to subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.