[Skip Global Navigation]

 Downloads

Customers

Customers Home

Globo.com

Situation

Globo.com is the Internet service provider/portal operator of Globo, Brazil's largest media conglomerate. The company's business units include television, cable television, newspapers, magazines, and entertainment parks. In 2003, Globo.com began providing both standard and paid Internet services. Currently, the company has more than 200 employees and 200,000 subscribers. Globo.com's portal receives between four to five million visitors per day.

Challenge

Internet service providers such as Globo.com know that in addition to strong advertising revenue, success requires healthy subscriber revenue. Globo.com, one of Brazil’s largest ISPs, adopted a strategy to improve its customer service with the goal of converting portal visitors into new customers and reducing its subscriber churn rate. As part of this strategy, customer service representatives needed better insight into what individual customers wanted in an Internet service. By understanding users’ individual online preferences, customer service representatives could tailor new products and services to user interests, thus maintaining their loyalty as well as improving the bottom line.

Solution

Previously, Globo.com generated market reports on customer behavior from legacy system queries. The process was complicated, inefficient, and did not give marketing teams access to critical customer information, such as search and download history, number of e-mail accounts, and content preference.

With PASW Statistics, Globo.com can now collect and act on this data to customize offerings on its paid site. By improving these offerings, Globo.com acquired 200,000 new customers in two years.

Globo.com — In The News

Brazilian ISP Globo.com Gains Customer Insight With SPSS
by Melissa Campanelli, DM News, June 21, 2005

Globo.com is using SPSS predictive analytics to improve customer service
by staff, CRMAdvocate, July 07, 2005

Interested in Globo.com? Download the PDF

Results

Globo.com is part of the Brazilian-based Globo media conglomerate, which includes the Globo television network, the O Globo newspaper, Globosat satellite services, Globo Radio Network, and Editora Magazine. Founded in 1965, Globo television is the nation's number one network, renowned for its cutting-edge programming, professional reporting, and state-of-the-art technology.

In 2003, Globo launched Globo.com, now one of the largest ISPs in Brazil. Globo.com hosts a robust search portal homepage, as well as several free and pay Internet offerings, such as streaming audio and video news.

Recently, Globo.com determined that to meet profit goals, the company needed to upgrade its customer service capabilities. Most importantly, it wanted to provide customer service teams with specific information pertaining to individual subscriber Internet usage, as well as general usage trends. This strategy would help convert new portal visitors into customers and current service subscribers into loyal customers.

To achieve this goal, Globo.com sought to better understand its customers' interests, and to deliver products that reflected those desires. Quick access to this sort of information was elusive, however, as legacy query and reporting software proved unable to produce such data in a timely manner without heavily taxing the company's network infrastructure. By using PASW Statistics, Globo.com was able to overcome these obstacles, drawing robust insight from its database and personalizing its services on an individual basis.

Using the PASW Custom Tables module in PASW Statistics, Globo.com detailed a set of criteria that allowed for a more finely tuned Web experience specific to each subscriber's preference.

With PASW Statistics, working on streaming, Internet access, and e-mail logs, we were able to develop a customer usage model, which can segment the customer behavior and identify different utilization levels. With these segments in hand, we made an effort to understand the motives of churn, giving us a chance to implement an effective retention strategy. Today we can track a set of indicators, per customer, which allows us to know what kinds of offerings better suit their taste. With this insight we can help to ensure customer loyalty.

Jeremias Coral Klausner
Data Mining Coordinator
Globo.com

Developed indicators to better understand customer behavior

Globo.com understood that to retain current customers and increase its market share, it needed to determine what type of Internet experience people preferred. Using PASW Statistics, Globo.com was able to develop specific indicators that allowed the company to score the customers as low, medium, or high users. These indicators measured such variables as number of e-mail accounts, average megabytes of content downloaded, and number of downloaded streaming videos.

This information was then analyzed against reports from customer service that assessed churn motives, such as competitor benefits, problems with e-mail access, and quality of content. Globo.com was then able to segment customers and create loyalty strategies based on discounts, packaged plans, and specified content for each user.

"With PASW Statistics, working on streaming, Internet access, and e-mail logs, we were able to develop a customer usage model, which can segment the customer behavior and identify different utilization levels. With these segments in hand, we made an effort to understand the motives of churn, giving us a chance to implement an effective retention strategy," said Jeremias Coral Klausner, data mining coordinator. "Today we can track a set of indicators, per customer, which allows us to know what kinds of offerings better suit their taste. With this insight we can help to ensure customer loyalty."

Reduced churn rate by 30 percent

Globo.com was able to transfer insight gained from the PASW Statistics reports into the hands of customer service representatives who deal with customers thinking about moving to other Internet services. By recognizing customer preferences, such as online video or audio usage, customer service representatives were able to offer these potential churners the kind of Internet options that fit their individual wants and needs. In addition, customer service representatives were given competitors' retention strategies and patterns, information tracked by PASW Statistics. As a result, Globo.com was able to reduce its churn rate by 30 percent in about a year's time.

"The capacity to collect and process not only identifying information such as name, age, sex, etc., but preferences and habits of use to identify trends and patterns, and ultimately make the correct offer, is most important," said Klausner. "With this information, our customer service representatives are able to develop a set of discounts and benefits per customer to ensure retention."

Stimulated subscriber demand for add-ons and upgrades

Knowing customer preferences also let Globo.com representatives know which premium offerings to up-sell to current clients and new prospects. Globo.com has also been able to extend user segmentation by topics of interest. For instance, the customer service representatives might package support services with additional premium sports services for customers that use the Internet to follow their favorite team.

Another way that Globo.com has increased premium service purchases is by using customer-specific e-mail. By utilizing topic-based segmentation, Globo.com increased premium video viewing using e-mail campaigns to communicate the availability of new videos and special reports.

"The customer usage insight we have gained using PASW Statistics has enabled us to sell premium services and additional add-ons more efficiently," said Klausner. "Using e-mail and voice, we have been able to communicate to customers offerings that we feel will increase the value of their Internet experience."

After spending a considerable amount of time looking into which data mining solution to use, Klausner is glad he chose PASW Statistics.

"We tested PASW Statistics against other data mining solutions. In addition to its affordable price, PASW Statistics’ report capabilities are better and easier to use than its competitors’." Klausner continued, "With SPSS, we are able to focus on our business needs, and more importantly, our customer’s Internet needs. PASW Statistics has been an integral part of our success in Internet services in Brazil."