Every business strives to bring more revenue through the door. Business leaders understand the importance of satisfied customers, but they also realize that catering to some customers can cost more money than is brought in. How can organizations predict a client’s behavior and purchasing power? Through advanced analytics.

Traditional business intelligence (BI) enables companies to understand the “here” and “now,” and even some of the “why,” of a given business situation. Advanced analytics goes deeper into the “why.” Employing a business-focused approach, it comprises techniques that help build models and simulations to create scenarios, understand realities, improve decision making and determine forecasts. These techniques include—but are not limited to—data mining, predictive analytics, applied analytics, data visualization and statistics.

This method of analyzing data provides intelligence in the form of predictions, descriptions, scores and profiles that help businesses better understand customer behavior and business trends. Its capabilities can drive a wide range of applications, from operational functions such as fraud detection to strategic analysis like customer segmentation to build loyalty and stem attrition (see company example #1).

Company example #1 A large wireless phone service provider was concerned with the number of customers it was losing—millions of dollars in revenue was lost each month. Using advanced analytics, the company developed an attrition model to predict which customers were most likely to terminate their contract. In doing so, the company developed a model to cross-sell, helping retain customers by providing products, services and other incentives targeted to their profiles. This program increased contract renewal rates by 60% and decreased attrition from 2.4% to less than 1.5%.

Company example #2 A consumer goods manufacturer ran analytics as part of its customer loyalty program; however, as the company collected data from its retailers, the analytics began to severely impact the system, taking more than 300 hours to run. The group partnered with Teradata and pushed the analytics into the data warehouse, reducing runtime to 12 hours.

Issues in deploying advanced analytics
When implementing advanced analytics projects, it is not uncommon to encounter challenges along the way. Businesses today are storing more data than ever before to support product development, marketing and inventory and point-of-sale (POS) transactions. The more data that can be processed and analyzed, the better the advanced analytics techniques can be at finding useful patterns and predicting customer behavior. As data complexity and volumes grow, so does the cost of building analytic models. Clearly, performance and cost can significantly impact an enterprise’s motivation to conduct these analyses.

Time-to-market constraint is another potential pitfall of an advanced analytics project. About 60% to 80% of human effort is involved in the up-front work of gathering, cleansing and manipulating the data, all of which are essential to a successful project. This, of course, can affect data latency, which in turn can impact the analyst’s ability to deliver analytics in a timely manner. The challenge is reducing the time to prepare the data for analytics.

Furthermore, market forces, customer requirements, governmental regulations and technology changes collectively demand that data that is not up-to-date is not acceptable. As a result, today’s leading organizations are constantly working to improve access to and analysis of current data. Businesses no longer have the option to take weeks and months to analyze data, but they must deliver tangible results to make smart business decisions within hours and days.

Having addressed the benefits of advanced analysis, the next step is to examine the solutions Teradata offers, including Teradata Warehouse Miner (for in-database analytics to build or optimize existing analytic environments), best practices (to create an Enterprise Analytic Data Set (ADS)) and Teradata advanced analytic partners (to provide near real-time data visualization capabilities).

Optimizing analytics
Teradata Warehouse Miner is a set of advanced analytic features that automate and optimize your enterprise’s analytic process to:

	Control and reduce the cost of analytic development by pushing data preparation tasks such as extract, transform and load (ETL) and data management tasks into the warehouse. Teradata Warehouse Miner also consolidates analytic data marts to eliminate unnecessary data movement, redundancy and latency.
	Speed time-to-results by streamlining key, repetitive tasks through automation and deploying analytics in-database for dramatic performance improvement (see company example #2).
	Improve analytic precision with comprehensive and optimal analytic data, and improve data quality through automated summarized and record-level analyses.
	Improve quality of advanced analytics by creating embedded analytics and scores to support real-time analytics.

Teradata Warehouse Miner helps reduce the cost of analytics by pushing the data-intensive tasks directly in the Teradata Database, which results in significantly improved performance. It provides scalable enterprise analytic modeling technology that reduces data mining cycle times for faster delivery of information. The solution also supports an interface that enables other tools to run models directly in the Teradata Database and facilitates the creation of the analytic data set required for all other advanced analytic modeling tools.

Remember the speed-to-market issue? Instead of requiring days or weeks for data extraction, joins, subset development, merging, aggregation and transformation, all of these functions can be done directly in the Teradata Database against large data volumes. Teradata Warehouse Miner facilitates the setup so structured query language (SQL) novices can quickly create an analytic data set optimized for Teradata.

Company example #3 One customer built a 10-million row analytic data set containing customer records with more than 500 variables. The analysis took more than six hours; once moved into the Teradata Database, the process took about 15 minutes and the customer was able to reduce the analytic development cycle from weeks to a few days.

Company example #4 A global entertainment company was running an SAS model to forecast the sales of new movie titles. Each forecast required more than seven minutes per title, a translation of 36 hours for 300 titles. By splitting the task and pushing 90% into its Teradata Database, the company dramatically improved performance by a factor of 28.

Leveraging best practices
Teradata provides best practices that condense the data preparation tasks which account for 60% to 80% of the analytic development process. When models supporting analytic projects are run infrequently, or if there are only a few models to run, data preparation tasks can be done as part of each specific project. However, when an organization incorporates dozens or hundreds of models into its business environment on an ongoing basis, the repeated manipulation of large amounts of data becomes inefficient.

As many models are built over time, certain standard metrics and manipulations become apparent. An Enterprise ADS takes the standard data rollups that are used in a variety of analytic tasks and centralizes their generation. Any required cleansing or recoding of the detail data that is used to facilitate rollups will be constant once the right analytic procedure is established. The standard metrics are created in an automated fashion and on a regular schedule, and then made available to all analysts and processes. Any entity that will be the focus of a wide range of analytics is a candidate for an Enterprise ADS (see company example #3).

Basically, the Enterprise ADS offers a simplified view of the data warehouse by providing a condensed, manageable number of analytic tables that represent hundreds of tables of detailed data. This process improves consistency in the methodology used by various analysts to generate their analytical data sets. The chance of an error due to the omission or altering of appropriate logic is eliminated. Overall, system processing cycles are greatly reduced since variables requiring a lot of heavy processing are computed once, stored and shared, rather than run time and again.

Deploying the best practices in advanced analytics by leveraging Teradata Warehouse Miner can improve time-to-results and analytic performance of other analytic environments. For example, by using external tools such as SAS, SPSS and KXEN, Teradata customers can rapidly build models from the analytic data set created directly in the Teradata Warehouse. Instead of downloading detailed data and aggregating it, the offline tool can simply access the Enterprise ADS, minimizing the impact of data transfer and analysis translation (see company example #4).

The bottom line
With advanced analytics, businesses can more accurately analyze large volumes of data. From this output, CEOs can interpret and forecast their customers’ needs, providing them the opportunity to deliver individualized customer service. This predictive ability, coupled with the cost savings from applying best practices within their enterprise, puts organizations at an advantage over their competitors. They are able to increase their profitability by making more intelligent and effective decisions faster. T

Craig S. Mullins is a data management consultant, author and speaker at conferences.

Real-time data visualization with Compudigm enlarge Compudigm’s seePOWER gaming portfolio creates data visualization that provides an accurate, near real-time view of the operation. Teradata supports real-time analytics projects with advanced visualization techniques. Compudigm, a Teradata partner, provides the capability to transform massive volumes of transaction and customer data into critical, real-time visual insights to make smarter, faster and more profitable business decisions. The combination of the Teradata Warehouse and Compudigm’s seePOWER software delivers intuitive visualizations across multiple industries that can be viewed over space and time. By combining thousands of data points into a visual representation, business users can more rapidly glean trends and patterns, making information more readily available, digestible and actionable. For instance, you can use seePOWER in a Teradata Warehouse to analyze customer history, then, based on that analysis, manage real-time activity. To do this, customer rankings (scoring) need to be determined to assure the proper level of service. These rankings are generally done in batches via an advanced analytical tool such as Teradata Warehouse Miner and used in real time. Compudigm leverages this information to present the real-time picture of the business. In the gaming business, for example, seePOWER offers a particularly useful implementation of visualization and real-time analytics (see figure). The focus of real-time analytics capitalizes upon events and opportunities as they unfold in order to deliver optimal customer service. The focus on customers is based on their analyzed profile: Top VIP customers are given preferential and individualized attention, hot players are monitored to proactively acquire new loyalty card members and other carefully selected customers are offered an up-sell while they are on the floor. The result: better-serviced customers and a more profitable gaming business. —C.M.

Teradata Magazine-December 2006