Published on 23 Dec 2009
Validated on 08 Jan 2013
"By incorporating IBM’s composite data analysis technology into our blog service delivery, we’re finding new ways to engage our readers, provide the seeds for new communities, and energize overall activity on our blogs." - Shinsuke Usami, managing director, CyberAgent
Media & Entertainment
Empowering People, Enterprise Content Management, IBM Research, Smarter Planet, Business-to-Consumer, Leveraging Information
The goal of service provider CyberAgent—a blog site host—was to put in place an intelligence capability that would enable a more comprehensive understanding of behavior as it pertained to blog usage and preferences.
CyberAgent, Japan’s largest blog service provider, sought to create a new level of engagement and loyalty among its blog readers.
CyberAgent and IBM Research developed a first-of-a-kind analytics framework that enables an unprecedented level of insight into blogging preferences and behavior.
Expected increase in loyalty among blog readers and blog authors; expected increase in page-view volume by virtue of smarter recommendations and a more engaging blogging experience; expected increase in advertising revenue.
What makes it Smarter
CyberAgent can translate user profile data into highly targeted blog recommendations and loyalty programs, increasing loyalty and stimulating higher volumes of blogging activities.
- Expected increase in loyalty among blog readers and blog authors
- Expected increase in page-view volume by virtue of smarter recommendations and a more engaging blogging experience
- Expected increase in advertising revenue due to a higher volume of page views and the ability to charge advertisers more because of highly targeted nature of the readership base.
- Text Network Analysis (TENA)
- IBM Content Analyzer
- IBM Engineering and Technology Services – Component Design Services
- IBM Engineering and Technology Services – Technology Consulting Services
- IBM Research – Tokyo
Smarter Media—Building reader loyalty by increasing engagement
- Instrumented: CyberAgent’s user activity databases capture a rich array of information on site visits, blog reading and comment writing.
- Interconnected: TENA framework mines user activity data by simultaneously running advanced text and network analysis to paint a comprehensive profile.
- Intelligent: A deep understanding of blog readers enables CyberAgent to deliver highly targeted blog recommendations and loyalty-building programs.
Along with social networking, the explosive growth of social media—and of blogging in particular—is perhaps the most visible sign of how Internet usage is becoming more and more a participatory experience, built around communities. With more than 100 million blogs thought to be in existence today, the sheer size and diversity of the “blogosphere” makes it difficult to broadly characterize. For instance, while blogs as a rule focus on specific subjects or figures, they offer a great deal of latitude as to how they meet their goals, including the kind of media (text, video, photos) they employ, their length (from long, in-depth posts to “microblogs”) and their frequency, to name just a few factors. Ultimately, however, the acid test for a successful blog is its ability to attract a readership base and engage it with a rich, interactive experience that keeps readers coming back for more.
Achieving this naturally requires an understanding of the unique dynamics of the blog experience and—just as important—of the needs that blog readers expect them to fill. By and large, blog readers are motivated by deep interests in specific subjects. What blogs can deliver that most content-specific Web sites cannot is a high degree of spontaneity, currency and perhaps above all, interactivity, in terms of what is discussed. This last point reflects the central importance of reader comments in animating—and often guiding—the blog’s activities. Indeed, by promoting active participation in the blog, reader comments are often integral to their vitality and represent the most engaging and interactive aspect of the blog experience.
Giving bloggers what they want
To the blog owner, establishing and maintaining a loyal following—a base of readers that considers the blog the “go-to” source on the subject—is the touchstone of success. But here’s the irony of blogs and loyalty. The very appetite for information that attracted them to the blog in the first place tends to whet their appetite for even more. That’s why blog sites that can provide readers with inspired recommendations on finding related content tend to make themselves stickier with readers and cement themselves as an indispensable destination. To service providers like Japan’s CyberAgent, which host blog sites, helping their blog site clients provide the best possible experience—and thereby serve and grow their audience—is a top-tier strategic priority. It was with this goal in mind that CyberAgent, the largest blog hosting provider in the world’s most actively blogging nation, collaborated with the IBM Tokyo Research Lab to leverage its knowledge of reader activities to create a more compelling blogging experience.
Knowing blog readers through analytics
A key aspect of CyberAgent’s vision was to provide blog readers with highly personalized recommendations on related content and resources. In contrast to traditional approaches, which employ generic algorithms that deduce preferences based on site visits, IBM aimed for an unprecedented level of insight by factoring in both behavioral inputs (such as blog viewing and posting frequency, online purchases and music listening) and text data (drawn from blog comments and postings), and incorporating all this data into a single, unified analysis. What made it possible was a first-of-a-kind analytics framework—known as Text Network Analysis (TENA)—developed by IBM Research – Tokyo.
The Inside Story: Getting There
The Challenge … CyberAgent’s goal was to put in place an intelligence capability that would enable a more comprehensive understanding of behavior as it pertained to blog usage and preferences. To achieve the leap in analytical depth this required, the CyberAgent solution melded the two dimensions of behavior—what people do and what people write—that are arguably the most accurate indicators of a blogger’s interests.
Taking a New Angle … CyberAgent’s solution leverages the core functionality of IBM Content Analyzer and TENA’s powerful analytics to enhance the capabilities it can offer its blog customers. Intelligent blog recommendation is a prime example. Using activity records stored in CyberAgent’s customer database, TENA takes a two-pronged path to creating a customer preference profile. The first is a comprehensive textual analysis of blog comment postings based on keyword frequency over a period of time, hence whether it is increasing or declining. The second—or “network”—angle of analysis factors in patterns of blog usage over time. By interweaving both analytical approaches, CyberAgent—via the TENA framework—is able to provide blog readers with truly insightful recommendations on blogs that are a close match with the reader’s demonstrated interests.
Rewarding Loyalty … One way for a blog to strengthen loyalty among its readers is to let them know that you recognize that loyalty. With that principle in mind, IBM developed an application that explicitly ranks blog readers based on a comprehensive set of activity inputs, and then uses that information to identify especially loyal supporters of blogs. To create this measure, the loyalty application analyzes not only the quantitative factors (such as the frequency of reading and commenting) but also the qualitative patterns that lie within unstructured data, factors such as the writing style and word usage in reader comments that can indicate a reader’s powerful affinity for a blog site.
A catalyst for new communities
This ability to detect and measure loyalty at a granular level gives CyberAgent and its blog customers a far more nuanced view into the readership base on which they both depend. One implication of this deeper understanding is an enhanced ability for individual blogs to meet the needs of their readers. But as CyberAgent realizes, the dynamism and vitality of the blogosphere comes from the constant search for new connections to other, like-minded readers. In the same way readers are always looking for new content, the communities they form are in constant evolution. By bringing together people with common affinities and giving them a means to interact, CyberAgent sees its loyalty application as a powerful catalyst to the formation of new communities, as well as a way to energize its existing blog community as a whole.
“On the strength of IBM’s analytics research and expertise, CyberAgent has made a major step in redefining the blog experience.” —Shinsuke Usami
With more than nine billion page views—double the level of just a year ago—CyberAgent’s celebrity-laden Ameba blog is among the world’s largest. Having recently introduced its new TENA-based functionality to Ameba, CyberAgent expects it to further boost the growth of blog readership by introducing blog users to new and highly relevant sources of content. Since CyberAgent’s business model is built on advertising revenue, the continued growth of its blog readership—as well as the general increase in engagement level—is expected to translate into substantial revenue increases for both CyberAgent and the blogs it hosts and shares revenue with. In addition to growth, CyberAgent also sees its ability to provide a more personalized and engaging blog experience as an important leg up in building loyalty among bloggers—thereby strengthening its competitive differentiation over the long term.
Toward a more engaging experience
Managing director Shinsuke Usami sees CyberAgent’s work with IBM as part of an ongoing effort to bring the quality of the blog user experience to a new level. “By incorporating IBM’s composite data analysis technology into our blog service delivery, we’re finding new ways to engage our readers, provide the seeds for new communities, and energize overall activity on our blogs,” explains Usami. “On the strength of IBM’s analytics research and expertise, CyberAgent has made a major step in redefining the blog experience.”
For more information
To learn more about how IBM can help you transform your business, contact your IBM sales representative or IBM Business Partner.
Visit us at:
IBM products and services that were used in this case study.
© Copyright IBM Corporation 2009
1 New Orchard Road
Armonk, NY 10504
Produced in the United States of America
All Rights Reserved.
IBM, the IBM logo, ibm.com, Smarter Planet, and the planet icon are trademarks of International Business Machines Corporation, registered in many jurisdictions worldwide. Other product and service names might be trademarks of IBM or other companies. A current list of IBM trademarks is available on the Web at ibm.com/legal/copytrade.shtml.
This case study illustrates how one IBM customer uses IBM products. There is no guarantee of comparable results.
References in this publication to IBM products or services do not imply that IBM intends to make them available in all countries in which IBM operates.