Q&A: Harnessing Big Data for Business Advantage
Big data introduces uncertainty and changing business requirements, but having the right processes in place to guide engagements can help smooth the way to success.
- By Linda L. Briggs
- September 11, 2012
In addressing big data concerns, recent research from HP shows that 50 percent of organizations lack an effective information strategy, and more than 60 percent claim that business intelligence solutions fail to meet current business needs. Those two statistics are linked, says HP's Brian Ng. "It's critical that organizations step back and address big data challenges from a business perspective," he maintains, with initiatives that are rooted in business value.
Ng is a seasoned information management executive with more than 20 years of strategic consulting and hands-on implementation experience for global Fortune 500 companies. At HP, Ing develops and drives the HP information and analytics strategy, service portfolio, methodologies, and delivery competencies.
BI This Week: What do you see as critical success factors in managing big data successfully today?
Brian Ng: Critical success factors for managing big data are largely the same as for any type of information management initiative. We have to take a fresh approach to some of these things to be successful, however. I would define success factors in terms of three categories: business alignment, process and organizational alignment, and technology alignment.
It's still critical for organizations to take a strategic view of big data first, then follow an iterative, incremental implementation plan. HP recently published research that shows 50 percent of organizations lack an effective information strategy, and more than 60 percent of respondents claim that business intelligence solutions fail to meet current business needs. I don't think that's a coincidence. It is critical that organizations step back and address big data challenges from a business perspective. Decisions about the initiatives we want to invest in and the key capabilities we want to implement relative to big data must be rooted in business value.
Big data introduces a high degree of uncertainty and changing business requirements. As a result, there is still a lot of exploration and learning about how we leverage it for business. Having the right process to guide big data engagements is critical. The use of business use cases, early discovery, assessment, and proofs-of-concept -- all coupled with detailed robust implementation methodologies -- will enhance the probability of success.
Finally, organizations need to balance emerging big data technologies with foundational and core capabilities. It is absolutely necessary to explore emerging technologies that are focused on big data capabilities, but not to the detriment of traditional capabilities such as enterprise data architecture, master data, data integration, data quality, and metadata. These are all very much a part of the foundation that is required to support the emerging capabilities. Without the corresponding foundational pieces, we run the risk of our big data solution being siloed and thus unable to be leveraged for enterprisewide value.
What common mistakes do you see clients making with big data?
The first mistake I see is organizations creating technology-led solutions instead of business-value-led solutions. Product and technology-led solutions typically focus on addressing issues the technology can solve, which is often a subset of the broader challenge and opportunity of big data. Many times, this results in the next common mistake, which is addressing big data in a siloed manner -- just as organizations have been doing for years, even before today's big data.
What are some of the promises that analytics, used on big data, brings to the business?
Big data can be leveraged by enterprises to enhance customer experience, increase revenue, drive efficiency, and manage risks. Here are some sample use cases in the world of big data:
Consumer packaged goods: A consumer goods company can leverage analytics to understand how a product is perceived by different market segments in social media. Marketing teams can leverage analytics to track the impact of a particular marketing campaign and adjust it accordingly. Incremental value can help results significantly -- in a large company, even a one percent improvement can have a huge impact on the sales results. In the world of big data, we do all these things with extreme high volumes of data and in real time.
Telcos: A telecom products-and-services provider can leverage analytics to provide information on mobile number portability to accurately show data on customer migration and churn. This enables them to analyze and identify new customer decision paths, thus helping gain a greater market share. They can also identify social and relational communities to improve their campaign management target models to coordinate marketing activities more efficiently.
Web 2.0 and gaming: Today's online and social gaming operators leverage massive amounts of player engagement and interaction data to perform complex, real-time analytics. That can help enhance the gaming experience, direct advertising, drive customer retention, and cross-sell or up-sell activities.
Financial services: With the advancement of big data technologies, financial services companies now have added capabilities to leverage previously challenging or unusable human information, such as voice, videos, images, and e-mail to manage and control risks and compliance in a much more comprehensive manner. The ability to detect, report, and mitigate risks based on real-time analysis of human information can make a difference toward their bottom lines.
Can analytics be used on both structured and unstructured data? What about differences in formats, semantics, and organization?
Yes, analytics can be performed on both structured and unstructured data. In fact, the ability to reduce semi- and unstructured data into a structured form, then subsequently integrate those types of data with traditional enterprise and external structured data for analytics, is the best approach to gain complete and holistic insights for a given analysis.
In working with customers, how do you achieve business alignment specific to big data and information management?
We have developed some core methodologies and tools that we use to help drive the business alignment that is required for success with big data. HP's big data Business Use Case Framework helps guide the conversation with our clients, starting from a business point of view. These use cases help clients understand what is possible with big data and relate that to their own businesses.
HP's Master Plan methodology, which guides our strategy engagements with clients, includes a whole set of activities that focuses on business needs and outcomes. Through this series of activities, we work with executives to understand the potential value of information in the context of the strategic goals and objectives of the client's major business units. This information creates value statements and business justification supporting the recommended approach.
Through this process, we begin to identify major key performance indicators (KPIs) that may be most valuable to each of the key functional groups and their relative priority within each group. Ultimately, this information helps determine the overall cost vs. benefit ratings for the data or business subject areas identified. Throughout the process, we constantly validate the observations and understanding with business users to be sure we are accurately addressing their needs.
The underlying point here is that the business side of an organization has to participate fully in this process. Business alignment cannot happen without the direct involvement of the business.
What does HP bring to the table in terms of managing big data to benefit the business?
HP brings strong experience and expertise in managing big data, both from the technology and services perspectives. We have a portfolio of consulting services that help organizations capture, integrate, protect, share, and analyze their data. These services include everything from information strategy and governance to data integration and master data management. They also include business solutions such as our Social Intelligence offering, which helps organizations bring social media data together with traditional data to drive better customer experiences. Our consulting staff has a strong heritage in dealing with the most complex big data challenges organizations can bring our way.
HP also provides a comprehensive offering (combining software, hardware, and services) that allows organizations to manage, understand and act on 100 percent of their data. HP's next-generation information platform, IDOL 10, combines Autonomy's information processing layer with HP Vertica's high-performance analytics engine to enable organizations to analyze unstructured, structured, and machine data, all from a single platform.