Page 23 - Mobile World Daily - Day One
P. 23
AMDOCS | BIG DATA
Matt Roberts, Marketing Director, Amdocs Big Data and Strategic Innovations Business Unit
The Three Layers of THE MONEY
Big Data Implementation:
The top layer contains the actionable
Hard Work, Toys and Money analytical applications and discovery
capabilities. It is just a simple fact that
Colin Powell once said, “A dream Big data has transformed from a buzzword into a reality for service providers machines can recognize patterns and trends
doesn’t become reality through magic. across the world. However, to date the main hype and focus has been on the that humans may miss. This does not mean
It takes sweat, determination and hard data lakes and infrastructure while the real heavy lifting needs to also come that the machine is set to replace the human.
work.” There are a lot of complex, labor- from the tough tasks of (1) data extraction and cleansing, as well as (2) putting Machine learning and big data technologies
intensive activities that still need to be data science to work in the operational systems of the service provider. are designed to give timely advice and prompt
considered to bring big data benefits to life. users and systems to make smarter decisions.
Discovery and The Money
Today, statistics support the fact that Actionable Analytics The Toys Analytical benefits span across three main
businesses which invest in cross-company domains for service providers. Of course
data aggregation and then leverage modern Data Management there are additional uses, but the main areas
machine learning to derive competitive Infrastructure that provide material business and bottom
insights are capable of making better business line performance improvements are:
decisions and fine-tuning their business to Data Sources The Hard Work
meet key performance metrics. Many are 1) Dynamic marketing and segmentation to
forecasting that companies which do not make Three Layers of Big Data Implementation allow CMOs to better promote the right
the transition to a data-driven business have a services to the right people at the right time.
high probability of losing their competitive
advantage and will struggle to survive. 2) Care and business performance for the
optimization of the self-care portal, commerce
As Gartner has predicted, an average e-shop, and call centers to lower call handling
service provider could potentially generate time and optimize call center operations.
$300 million a year in additional margins
through successful analytics. Companies that 3) Network optimization for better
are more data driven are 5 percent more connectivity experience while optimizing
productive and 6 percent more profitable infrastructure investment.
according to McKinsey.
Across all of these categories, there are a
Becoming a true data-driven organization wide range of customer experience use cases
isn’t just about installing and integrating new which can analyze thousands of actual data
infrastructure. The problem is much more points or KPIs to materially improve Net
complex. Promoter Score.
The route to success is beset with three This top layer brings new requirements that
distinct challenges. To articulate these touch all service provider organizations, but
challenges, we have simplified and segmented has more profound impact on the CMO
big data analytics (BDA) into three layers. behavior and processes. CMOs in the past
would be charged with championing the
THE HARD WORK THE TOYS to as many critical data sources you have and company brand and driving interest in the
making sure these sources are coordinated products and services that it produced. This
The bottom layer is where CIOs and CTOs face The majority of big data projects to date have from release to release; (3) scrubbing and would typically be executed through large,
the big challenges of extracting data from a focused on the sexy, new technologies which hosting the data in a clean manner which is expensive and creative campaigns. Marketing
myriad of disparate systems. This problem is store and process this data. Although we call analytics ready; and (4) integrating the now is depending on IT to supply contextual
complex, time consuming and fraught with data this layer ‘the toys’, this is not to denote that infrastructure into the northbound information about its customers so that the
inconsistencies. Without the help of the service these systems are trivial or simple – far from applications and visualization capabilities. CMO organization can improve targeting and
providers’ vendor community, the probability of it. There is huge industry investment and customer interaction effectiveness. This
getting the data out of these systems and some of the brightest brains working on new One of the drivers towards new means that the CMO and the CIO will need to
hydrating their lake with clean, actionable data is storage and streaming technologies. The infrastructure is the fact that Hadoop can spend a lot more quality time together.
very low. Often service providers depend on result is that this open source software is both provide a lower-cost storage option than
third-party integrators to perform this function, very vibrant and very confusing – a virtual traditional data warehouses. Most In 2012 Gartner made a prediction that at
but the integrators are not intimately familiar circus of options. If not managed carefully it organizations analyze only about 12 percent the time raised many an eyebrow. They
with the data as it is obfuscated in various can be a big distraction to the end goal of of the data they hold, so the race is on to get forecasted that by 2017 the CMO would spend
operational and network systems. We are seeing adding value. faster, interactive SQL working well on more on technology than the CIO. Although a
the dawn of a new trend where service Hadoop. This doesn’t mean that the transition to this degree is unlikely, the
providers are reverting to the source vendors to The trick with this segment of the BDA traditional data warehouse will cease to marketing department will clearly play a
extract and clean this data in a reliable and real- layer is to choose a pragmatic vendor who exist. As more and more data needs to be greater part in such technology decisions.
time fashion in order to better service the rest of can consolidate the key elements of the data retained and analyzed in a cost-effective
the organization. It has been proven time and management infrastructure from this open manner, the CIO’s challenge is to cap the THE RESULTS
time again that relying on your core systems source market and bring it together in a spend in the expensive data warehouse and
vendors for this task is a much safer path to secure and cost-effective way. The most relational database technologies. Hadoop In conclusion, there are three layers of
hydrate a CIO's data management infrastructure important aspects of this are (1) choosing technologies are five to fifteen times less execution that need to be addressed to realize
with clean, ‘analytics ready’ data. and bundling the right mix of open source expensive for this function. the full potential of big data. To date the
technologies; (2) productizing data collectors programs have been fragmented and too
much focus has been on the infrastructure and
platforms. It is clear that given the specific
challenges in telecoms, more emphasis needs
to be given to the quick, efficient extraction
and ingestion of data and then to the value-
added applications that need to be applied to
drive the real value across all three of the
vertical analytics application segments
mentioned above. Only then will we even
begin to start talking about applying these
technologies to real financial benefits.
MOBILE WORLD CONGRESS DAILY 2015 | www.mobileworldcongress.com Monday 2nd March PAGE 23
