Big data presents opportunities for communications service providers (CSPs) to capture new revenue streams by optimizing large amounts of structured and unstructured customer data. To take advantage, CSPs must develop a strategic plan and roadmap to transform how they use customer data, identifying specific business values. Success stories show how CSPs have improved operational efficiency, provided targeted marketing offers, and created new business models through partnerships. The document recommends CSPs formulate a big data strategy and business case with measurable outcomes to guide strategic transformation and monetization of big data opportunities.
2. Viewpoint paper | Monetize big data
Table of contents
1 Take advantage of big data
1 Understand big data
3 Review big data’s evolution
4 Gain from big data opportunities
5 Understand—data is the new global currency of business
5 Review big data success stories
6 Review our recommendations
8 Transform strategically
9 About the author
3. Viewpoint paper | Monetize big data
Big data is a phenomenon brought about by rapid data growth,
complex, new, and changing data types, and parallel
technology advancements; it brings huge possibilities. By
optimizing these enormous amounts of structured and
unstructured data, CSPs are in a unique position to capture
these opportunities and create new revenue streams.
Daily, from our mobile devices, we: Download news
from home Wi-Fi; connect with friends on Facebook
Messenger, Google Hangouts, WhatsApp, WeChat,
or Viber; watch movies while waiting for the train;
settle bills using online banking; browse the web
to find the best air ticket price; and more. Each of
us is now a walking data generator, providing two
critical things—location and identify.
Take advantage of big data
Big data is a paradigm shift. At a tactical level, the analytic paradigm is shifting from analyzing
data using well-known schemas to finding hidden relationship patterns. At a strategic and
management level, it’s an opportunity for business transformation and a decision-making
evolution—making data-driven decisions that are speedier and better.
To take full advantage of big data opportunities, communications service providers (CSPs)
must not just remain on the tactical level by implementing tools only. We recommend you start
by engaging in transformational workshops to identify business value that can be achieved
through optimal use of new and old information sources. The key is formulating a strategy with
a road map to transform and monetize these opportunities.
Understand big data
McKinsey Global Institute, in the June 2011 report,
estimates that location-based data alone will
generate $100 billion of value to CSPs in the next
10 years. Asia is the leading region for personal
location data generation simply because of its
population density and high volume of mobile
devices used.
1
Big data is a class of data challenges—due to increasing volume, variety, velocity, and
complexity—that are beyond the capabilities of traditional software, architecture, and
processes to effectively manage and use.
To illustrate the impact of big data, in April 20131, three economists published an eye-opening paper.
It said Google Trends data was useful in predicting daily price moves in the Dow Jones industrial
average, reversing their earlier research published in 2010. What’s the key difference? The
data—faster data (real time) is better, and bigger amounts of data matter when it comes
to predictions.
F
orbes, “Big data gets bigger, now Google trend
can predicts the market,” April 2013
1
4. Viewpoint paper | Monetize big data
Figure 1. What is big data to CSPs?
Structured information
it used to run the business
• Devices
• Session
• Context
• CDRs, XDRs
• Subscriber data
• Usage
• Deep packet data
• Network QoS
• Web mobile behavior,
transactions
Billions of interactions
Millions of
transactions
Unstructured information
it used to gain insight on
business drivers
• Online sales
• Downloads
• Call notes
• SMS
• Web chat
• Blogs
• Social networks
• Mobile apps
• Sensors
• Survey response
• Emails
• Office documents
Big data sources are generally one of two types:
• High-variety unstructured or semistructured data, which contains human language or rich media
• High-volume, high-velocity machine-generated data, which can be messages or sensor readings
Today, 80% to 85% or more of new data is human-generated in a unstructured format.
Structured data is the millions of transactions
and information that operators have used to run
their businesses for many years.
Unstructured data is the billions of transactions
that are as varied as data from office documents,
sensors, survey responses, or call center operations.
2
The structured data that operators collect amounts to millions of transactions and is the
information they have used to run their businesses for many years. This area is a familiar
territory—they just haven’t applied this data to areas outside their traditional operations
to generate value, either because they didn’t see the value or were concerned about
privacy governance.
Unstructured data is a far more complex and new terrain for operators. There are billions of
transactions to take into account and sources can be as varied as data from office documents,
sensors, survey responses, or call center operations. Unstructured information can be
used to gain insight into business drivers, and bringing together data from structured and
unstructured sources is at the heart of operators’ plans to derive value and generate revenue
and differentiation from big data.
5. Viewpoint paper | Monetize big data
Table 1. Big data implications to CSPs
Big data characteristics
Volume and variety—Data collected by CSPs are
growing in terms of volume (from GB in 2000 to TB
in 2010) and variety (few different voice services in
2000, thousands of Internet services in 2010).
• elocity—This is the speed and frequency of data
V
generation and its delivery. It is also the velocity or
capability to analyze data in real time or streaming.
This is fundamental when you have to connect
analytics and action-ability in real time.
• Complexity—This is the inherent characteristics
needed to do analytics on a wide array of data—
structured and unstructured, under a global
environment, and in real time. As a result, the
difference in quality, cost-effective forms of
processing, security, regulatory, and compliance
and access requirements, at different stages of
the data life cycle form the complexity of big
data. And the combination of any two “V”
forces—volume, variety, and velocity—
increases analysis complexity.
PCRF and future software-defined network (SDN)
must be driven by sophisticated real-time analytic
systems, able to identify in real time, issues at
the subscriber level. Sophisticated analytics are
needed to compare events with historical trends to
keep the right decision.
Value—CSPs need to monetize data to support
increasing investments in network to support
growing traffic. They need to create new revenue
streams from their data.
• ariety— Big data comes from a greater variety of
V
sources than ever before. These sources range
from structured to unstructured data—text, audio,
video, and human language—and semistructured
data including XML and RSS feeds. Plus,
multidimensional data can be drawn from a data
warehouse to add historic context to big data.
So, with big data, variety is just as big as volume
and variety, and volume tends to fuel itself.
It is not just question of efficiently (cost,
operability, manageability, and security) storing TB
and petabyte (PB) of traffic, but it is fundamental
that the capability easily add new protocols, format,
and manage semi- and fully unstructured traffic—
for example, logs, social media, multimedia.
Velocity—Policy and charging rules function
(PCRF) are just the beginning of new era
where telecommunication networks will
adapt themselves in real time to subscriber’s
needs, profile, and quality of experience (QoE)
management. The network SLA is replacing
subscriber QoE.
• Volume—Data volume is the primary attribute
of big data. It’s not possible to quantify big data
volume. But in general terms, when you start to
worry about data, that’s your big data volume.
Implications
CSPs must extract more value from the
information they collect. They have to switch from
service analysis to preference analysis to activate
new ecosystems partners. The models are Google,
Amazon, Facebook, and LinkedIn.
Review big data’s evolution
Gartner analysts predict that data will grow 800% over the next five years and that 80% of that
new data will be unstructured.2
Further, Gartner predicts that nearly 70% of all business intelligence vendors will incorporate
natural-language capabilities into their applications by 2016.3 The result—users will be able
to perform searches and analyze data using natural language or even voice commands rather
than traditional SQL queries.
Big data requires technology enablers such as Hadoop, which is a widely adopted and proven
distributed file system for big data. However, it does not understand the meaning of concepts
or meaning contained in information. To do that requires technologies such as intelligent data
operating layer (IDOL) of Autonomy that can “understand” all forms of content, not just text, in
any language.
Integrating unstructured data with structured data will be a challenge, but it will provide
business value. For example, performing analysis on content from social media—a list of what
people’s interests are based on what they “say” on their accounts and their friends—will enable
mobile and ad experience personalization.
Technology solutions that embed natural language understanding and provide telco-specific
analytics solutions, such as deep packet inspections, are the preferred choice.
F
orbes, Big Data—Big Money Says It Is A
Paradigm Buster, June 2012
3
G
artner, Predicts 2013: Business Intelligence and
Analytics Need to Scale Up to Support Explosive
Growth in Data Sources, December 2012
2
3
6. Viewpoint paper | Monetize big data
Gain from big data opportunities
Customer loyalty is diminishing. Customers are easily attracted to other CSPs by a more attractive
offer, plan, or new device. Traditional revenue streams, such as short message service (SMS)
and international direct dialing (IDD) calls, are being eroded by over-the-top (OTT) players such
as Facebook, Google, WhatsApp, Skype, WeChat, and Viber. Meanwhile the explosive use of
mobile devices creates complexity and big demand on network capacity. While there is a need
to generate more revenue, keeping an eye on expenses is the other end of the equation.
The key business drivers for CSP today are:
• Customer-centric and customer experience management objectives—to remain in the game
and enhance revenue
• New business models—to create new revenue streams
• Risk/financial management—to prevent revenue leaks
• Operational optimization—to optimize network and operating costs
Figure 2. Can CSPs avoid becoming a data pipe?
Currently located two
miles from sports store
GPS
Video
conferencing
Data
transmission
Transfer
data
4
Understand customer experience
across network, services, and
social conversation.
Network optimization
Digital
music
Streaming
video
Saves
department
store coupons
Six slow
data sessions
yesterday
Digital
coupons
Experience
providers
Social
media
Web
browsing
Understand customer use,
behavior, and interests.
Targeted products and
marketing offers
Monthly data
usage mostly
used on social
network and email
Browsing baby websites
last two months
Connect with OTT players,
advertisers, and verticals.
New business models
Data is the new currency of business
7. Viewpoint paper | Monetize big data
Understand—data is the new global currency of business
Information is at the heart of transforming to communication experience providers in addition
to core services. While the notion of average revenue per user (ARPU), understanding the
customer and personalizing the experience, and connecting and monetizing with ecosystem are
not new, traditional information management approaches are no longer sufficient to address
the emerging demands of agility. To understand opportunities brought about by big data, the
following table illustrates example of big data levers.
Table 2. Big data lever grouped by functions
Function
Marketing
“HP is the only vendor who
could understand SBM’s true
issue in usage data
management. HP Dragon
streamlined our CDR
operation and can now
leverage our view of
customer usage to provide a
better customer experience.”
Akihiro Muranaka, manager of Rating System
Development Department, Information System
Div., SoftBank Mobile
Big data lever
• Targeted products and marketing offers (personalization)
• Cross-selling
• Location-based marketing
• Customer microsegmentation
• Sentiment analysis
• Enhancing the multichannel consumer experience
• Customer/product life cycle analysis
• Churn analysis (predictive analysis)
New business model
• Geo-targeted advertising and couponing
• Revenue sharing with digital retailers
• Monetization of (obfuscated) subscribers’ data
• In-store behavior analysis
• Insurance pricing (for example, go-as-you-drive insurance)
• Health care monitoring (for example, senior citizens or family members)
• Safety tracking of family members
Network engineering
• Network optimization (for example, wi-fi offload)
• Quality of experience
• Proactive resolution (QoS policy)
Customer services and
operation
• Customer experience management
• RT FAQ broadcasting
Finance
• Fraud management and prevention
• Revenue assurance
IT
• Operational efficiency
Review big data success stories
Gain operational efficiency and customer experience management
With billions of usage records per day, SoftBank Mobile (SBM) realized the business benefits to
be gained by improved management and better access to their usage records repository. HP
Dragon RED—rather than additional storage—was deployed to compress huge customer data
record (CDR) archives. The unlimited scalability architecture, high compression ratio, highload performance, and ready-made processes of data analytics enabled SBM to achieve the
following business benefits: Scaled systems to keep pace with double-digit growth in billing
data volumes over six months; reduced archiving area by 90% without additional storage; and
improved operational efficiency, including faster daily packet CDR query performance.
5
8. Viewpoint paper | Monetize big data
Provide targeted offers and new business model
Two major North American carriers and a tier one CSP in Asia Pacific deployed an HP Smart
Profile Server (SPS) so they could partner with an advertising company and collaborate with
OTT players to create new revenue streams.
With input from HP SPS Customer Sentimental Analysis—a result from social network analysis
and mining, and call record analysis—the solution correlates a live session with network-based
data. It then provides in real time (milliseconds), a meaningful masked profile—based on
customers opt-in approval. The profile can then be used by ad networks, publishers/destinations,
and others through real-time bidding marketplaces. This improves advertising effectiveness
and offers customers a more personalized digital experience. It also enables generating
targeted offers/plans for customers. By providing a personalized customer service based on
customer sentiment coupled with Service Experience Analytics, the result is happier customers,
better stickiness, and stronger customer lifetime value, while creating new revenue streams.
Use new business model—machine-to-machine
Some CSPs offer pay-as-you-go car insurance. Telemetric devices are plugged into the car’s
OBD port. Information collected from these devices provides data to insurers. Rates might be
lowered from 1% to 45% if the device finds the driver drives less over time, detects no high
speeds or erratic braking, and finds little late-night or rush-hour driving.
Gain operational efficiency and customer care
By combining information from networks, call centers, and social networks, you can transform
from a reactive to proactive approach in customer care. Following are the key features and benefits:
• Anticipate questions from customers thanks to a better understanding of what is happening in
real time
• Significantly reduce calls to customer care by broadcasting frequently asked questions to
every touch point—in real time
• Improve first call resolution and reduce average handling time
Review our recommendations
To optimize big data’s benefit, we provide a broad set of recommendations to proceed down the
journey of big data, eventually evolving it into a core competence.
Formulate strategy and road map
Big data-related initiatives will bring about significant changes for business, organization,
technology, process, and potentially the industry. It sounds quite obvious that a strategic plan
is required. However, most companies start at a tactical level by buying tools, with no strategic
plan in place. This most often ends up with costly and ineffective solutions or inaction.
The power of the plan is to create a shared vision among senior executives, technology
professionals, and data scientists of the potential business values. As big data initiatives evolve,
investments typically involve competing strategic priorities. Senior executives should oversee
this to ensure efforts are aligned with the company’s strategic intent. Big data strategies provide
input into road map development. The road map should include priorities and associated
business values, actions required for implementation, and how to mitigate potential risks. It
should also include a data strategy covering data governance, data privacy, and from a
customer experience perspective, the need to incorporate opt-in/opt-out strategies.
In a recent survey by European Communications, a group of industry executives found
agreement across the board—91% believe that big data strategies should be a priority for
every service provider.4
4
6
B
ig Data Survey: New Revenue Stream
top CEM as biggest Opportunity, European
Communications, May 2012
9. Viewpoint paper | Monetize big data
Create a business case with measurable outcomes
CSPs should always commence on big data initiatives with measurable outcomes and senior
management commitment. Before developing any business case, we would also recommend that
you commence your big data journey with a proof-of-concept scale project with a small investment.
Further big data initiatives should be not viewed as an IT project unless the initiatives are purely
IT-related. It is recommended that a big data business case always have a business owner. The
key success factor is for business functions to take the lead, “roll up their sleeves,” and refrain
from being a back-seat driver giving IT instructions to produce analytical results. On the other
hand, IT should refrain from the model of “build it and they will come.” Business and IT must
partner for success.
Gartner analysts predict that data will grow 800%
over the next five years, and that 80% of that data
will also be unstructured.5
In fact, there is a trend for sales and marketing or operation/customer care functions to
increase ownership. Gartner predicts that by 2017 the chief marketing officer (CMO) will spend
more on IT than the chief information officer (CIO).6
Develop a proof of concept
To get started, we recommend picking a business function to be the testing ground. The
business function should be analytics-friendly backed with data scientist skills. Based on
initiatives from the road map and committed initial efforts to customer-centric or customer
experience management outcomes, these areas should provide value for the business to
remain competitive and potentially generate additional revenue. Start small with simple
analytics, measure the outcomes, and replicate the model across the enterprise quickly,
increasing the analytic complexity gradually based on the big data strategy.
Realign processes
The challenge often lies not in the analytics but in their effective integrations into organizational
processes. Organizations must re-align the end-to-end process from what and how data is
collected, to who analyzes them and the final decision-making process. If the process is not
aligned, there is a risk for analytics redundancy or data discrepancies, which undermines big
data’s value.
Manage talent
Some of the most crucial big data success factors are the availability of data scientists and
other professionals skilled at working with large quantities of information. The best data
scientists have business domain expertise and are comfortable speaking the language of
business and helping leaders formulate their challenges in ways that big data can tackle.7
They should have coding or technical skills to use the tools and be interested in finding hidden
patterns in large data sets—internal and external. Data scientists understand how to fish out
answers to important business questions from today’s tsunami of unstructured information.
According to Gartner, by 2015 big data demand will reach 4.4 million jobs globally, but only onethird of those jobs will be filled.8 As companies rush to capitalize on the potential of big data, the
largest constraint many will face is the scarcity of this special talent.
Consolidate and remake architecture
Many CSPs have overlap or redundancy in warehouses or data marts. This may be due to
organizational differences or the result of mergers and acquisitions. Regardless, it increases
costs and can contribute to management confusion if it results in “multiple versions of the
truth” being reported. With big data deployment, IT organizations should leverage the big data
strategy to consolidate data stores, clearing up confusion and remaking big data architecture
and applications to ensure success.
F
orbes, Big Data—Big Money Says It Is A
Paradigm Buster, June 2012
6
G
artner, Forecast Analysis: Enterprise IT
Spending by Vertical Industry Market,
Worldwide, 2Q13 Update
7
“
Data Scientists: The Sexiest Job of the 21st
Century,” Thomas H. Davenport and D.J. Patil,
HBR Oct 2012
8
G
artner, Predicts 2013: Big Data and
Information Infrastructure, November 2012
5
7
10. Viewpoint paper | Monetize big data
Use a third-party cloud platform
A recent report from AnalysysMason claims that more than a quarter of CSPs have no
strategy for big data analytics, and the vast amount of information that CSPs hold about their
subscribers has been largely untapped.9
To address this shortcoming, you must define your big data strategy. For executing the strategy,
some CSPs are turning to third-party cloud computing services to monitor their networks and
software infrastructure and analyze customer behavior. This partnership achieves two key
objectives: It identifies new patterns and opportunities to build customer loyalty and creates
new revenue streams. And it enables you to move quickly by avoiding large investments and
going through the learning process internally, leaving time to focus on the business strategic.
Transform strategically
Use data in decision making
One of the most critical aspects of big data is its impact on how decisions are made and who
gets to make them.10 Although most companies today rely largely on data for their decisions,
often, the most important decisions rely on HiPPO—the highest-paid person’s opinion. Most of
time, the decision is made and then data is collected to support the decision.
The first question that a data-driven organization asks itself is not “What do we think?” but
“What do we know?” “What does the data say?” and “Where does the data come from?” When
it comes to what problem to tackle and what questions to ask, of course, domain expertise
remains critical.
To nurture an analytical culture, organizations should also consider deploying simple and
usable analytics tools to their front-end employees, for example, sales team, with proper
training, where applicable.
Big data does not undermine the need for vision and human insight. The company who can
combine vision and domain expertise together with data science will be ahead of their rivals.
Seize this window of opportunity
Big data is no longer hype. The payoffs from joining the big data analytics management
revolution are no longer in doubt. There are numerous success cases re-enforcing that
companies who inject big data and analytics into their operations show productivity rates and
profitability that are 5% to 6% higher than those of their peers.11
Consumer attitude is changing towards data privacy. OTT players such as Google, Facebook,
and Twitter have changed the attitude towards sharing personal information. Having said that,
to capture the big data opportunities, CSPs should have a data strategy addressing data privacy
concerns, such as opt-in/opt-out measures.
Data has become the new global currency. Ultimately, the insights that are derived from big
data strategies will increase your knowledge of your operations and customer base. This
knowledge enables you to better direct network investment, improve the efficiency of operational
processes, deepen knowledge of customers’ behaviors and needs, enhance customer experience,
strengthen customer relationships over time, and create new business models.
Seize this crucial window of opportunity to become a competitive new service provider. Use big
data analytics to transform, monetize, and eventually become your most valuable core competency.
While CSPs would have similar data sets and thoughts, the real differentiators will be the “first
mover” advantage, using holistic strategies from the beginning and superior execution, which
needs to be supported by relevant skills and solutions.
B
ig data analytics: How to generate revenue and
customer loyalty using real-time network data,
January 2013
10
“
Big Data: The Management Revolution,” Andrew
McAfee and Erik Brynjolfsson, HBR October 2012
11
Ibid
9
8
11. Viewpoint paper | Monetize big data
Learn more at
hp.com/go/scs
About the author
HP Industry Advisory Program
The HP Industry Advisory Program is a unique
HP Solution Consulting Services program that
delivers innovative thought leadership to address
our clients’ key business issues. The program
is built on the global knowledge, expertise, and
experience of our industry business consultants. It
incorporates proven HP methodologies, industry
frameworks, and intellectual capital to deliver
true business value through a collaborative, social
media-based environment.
Helen Chu
Helen Chu is a principal consultant from Solution Consulting Services in the Asia Pacific region.
With an extensive background in business consultancy and strong industry experience, Helen
harnesses her capabilities in comprehending business challenges faced by individual clients and
her deep technology knowledge to develop and execute effective strategies and transformation
programs, while innovating and leveraging technology solutions that deliver business value to
HP clients.
9