LS_WhitePaper_NextGenAnalyticsMay2016

Seeing the big picture about
big data
Next-generation analytics for
life sciences organizations
May 2016

Seeing the big picture about big data 1
Contents
2 Transformative data governance enabled by next-generation analytics for
life sciences organizations
2 Big data as the big disruptor
3 The “democratization” of data governance
4 Similar, but expanded
5 Look ahead of the curve
5 Conclusion
6 Contacts

2
Transformative data governance enabled by next-
generation analytics for life sciences organizations
In a rapidly evolving marketplace fueled by regulatory
changes and global competition, data is possibly the most
valuable asset that life sciences organizations have. While
most organizations see the potential of leveraging data
streams from new technologies such as wearables and
connected devices, they also see a growing number of data
challenges before them. How can they handle the massive
volume of data being produced today? How can they keep
data safe and compliant? How can they ensure that the
right audiences are using data in the right context? The
answers often lie in their approach to data governance and
the information management and business intelligence
platform they use to enable it.
While almost all life sciences organizations have data
governance structures, new technologies, and the “big
data” they acquire, produce and consume for decision-
making purposes, are straining these legacy environments.
In many instances, standard models and processes are
no longer sufficient. The journey to becoming an insight-
driven organization is a three-step process, moving
from information to insight to impact. Since good
data governance is a prerequisite for effective analytics,
organizations won’t be able to arrive at the insights and
impacts they desire without a change in focus.
Big data as the big disruptor
Many disruptive forces are converging upon life sciences
organizations today, ranging from collaborative business
models to healthcare reforms. However, “big data” could
be the most challenging, and transformative, of them all.
While the definition of “big data” differs across industries,
for purposes of this discussion it comprises five main
data streams:
1. Web and social media—Through social media sites and
a number of other channels, customers can share their
experiences with a medication or device as well as their
sentiments regarding a company or a product.
2. Machine-to-machine—From sensors deployed in a
hospital environment to wearable devices consumers use
at home, communication between devices through the
Internet can produce terabytes of information per day.
3. High-volume, high-velocity transactions—This
encompasses major administrative areas such
as patient claims and interactions with benefit
management organizations.
4. Biometrics—While still evolving, biometrics can produce
huge volumes of highly sensitive, personal data such as
fingerprints, voice scans and face maps.
5. Personally generated—From emails to blogs, personally
generated data represents massive streams of mostly
unstructured content.
This deluge of data represents an opportunity if well
managed, but a liability if not. Big data, however, is
voluminous and complex in terms of its breadth of sources,
varied formats and unstructured nature. The rate at which
it is being produced is also much faster than anyone could
have anticipated even a few years ago, and employees
often become overwhelmed in trying to process it—not to
mention in trying to extract valuable insights from it.
The velocity, volume and variety of big data additionally
compound many of the existing data-related risks and
challenges inherent to the life sciences sector. These include
difficulties in accessing and managing the complexity of
clinical and R&D data, which is typically strewn across many
companies and third parties. Indeed, a heavy reliance on
third-party data in general poses unique risks to life sciences
companies, raising questions not only about which data
sets to purchase but also how to use them effectively.
Furthermore, the availability of a wider variety of data sets,
both from internal sources and from third parties, raises
new concerns regarding how to secure sensitive Personally
Identifiable Information (PII) or Protected Health Information
(PHI). It also elevates the risk of regulatory non-compliance,
since patient data is increasingly a component of these data
sets. Meanwhile, the ownership of the data is becoming less
clear, all while companies are being called upon to share
data with collaborators in a safe and compliant way.
In order to address these challenges, while simultaneously
unlocking the insights that big data has to offer,
many life sciences organizations will need to reassess
their data strategies as well as evolve their data
governance organizations.

The “democratization” of data governance
With huge volumes of data coming from many different
sources both from within and outside the enterprise, a
new “democratized” approach to data governance may
be required. In order to be effective, data governance can
no longer be driven solely by a centralized governance
organization. Instead, various functions of business and
IT should be involved, with their roles, responsibilities,
handoffs and interactions being guided by a collaborative,
responsive, and agile framework. The good news is that
taking this new “democratized” approach rarely means
starting over. The traditional dimensions of governance
still apply to big data, but with a shift in focus—mainly
realigning objectives and activities across people, processes
and technology. These data-governance dimensions, and
proposed shifts, are discussed in greater detail below:
• Organization—As previously mentioned, the data
governance organization should become flatter and
more agile in order to accommodate big data. This
will likely require extending existing business roles or
adding new ones. In general, companies should seek to
assemble a research-oriented team that includes a mix of
data scientists, senior architects and business users, and
to implement processes that allow them to collaborate.
They should also verify that the team’s skill sets align
with enabling technologies so that participants can make
the most of new and existing tools.
• Master data integration—Big data and master data are
highly synergistic, i.e., Big data uses master data, and
master data can be enriched based on big data. And
quite importantly, low-quality master data can impede
big-data analytics. To benefit from these synergies
and to enable big-data analytics, new master-data-
management roles may need to be added to the
existing data governance structure. These individuals
will be responsible for defining and executing processes
for using big data in order to enrich master data in
alignment with business needs. For instance, this might
involve leveraging the web or social media to increase
the quantity of data or to validate and complete master
data. To improve master data management, companies
may also wish to consider leveraging new technology
capabilities, such as solutions for sentiment and
network analyses, customer segmentation, preference
management and churn management, along with
business-friendly graph databases that scale more
naturally to large data sets.
• Metadata management—It is important for organizations
to assign a high-priority to the agile development of
building and maintaining a metadata infrastructure.
Why? Even though it is more challenging for big data,
metadata management is critical for deriving meaningful
insights from the flood of information. Once again, the
business will need to play a more active role in defining
the data “context,” since big data enters the organization
through many paths, and not just through a centralized
IT department. Thus, companies will likely need to
extend existing roles that sit within the business, such as
information stewards and metadata administrators, to
encompass metadata-management activities in relation
to big data. Processes may also need to be redesigned or
established, such as those for creating and maintaining
a consistent business glossary and data dictionary,
and those for facilitating communication between the
business and IT. To enable these expanded activities, the
analytics and visualization tools that decision-makers
leverage to gain insight should also provide capabilities
for metadata management at the point of usage.
Tools that enable automated capture, assignment and
publication of metadata should additionally be explored.
• Data security and privacy—The objectives of any data
security program are to guard against breaches, prevent
misuse of data, and to mitigate reputational, regulatory
and legal risks. However, big data is a prime target
for abuse since it rapidly introduces a huge volume of
heterogeneous information. This information could easily
be mishandled or misinterpreted, especially considering
that laws and regulations concerning data use vary
across geographies. To manage the security and privacy
aspects of big data, an expanded approach may be
required that includes assignment of security analysts
as well as integration with legal and risk-management
departments. The role of this integrated team is to
minimize security threats by designing processes for
identifying and managing sensitive data, while taking
the legal and legislative aspects of the data sources
into consideration. In addition to upgrading security
measures, companies will likely need to upgrade their
security technologies as well, with an eye toward
deploying adaptive and automated programs to

4
maintain visibility and control over internal and external
data, therefore protecting information assets. It is also
important to address data security at the information
consumption layer through analytics and visualization
tools. Self-service analytics should enforce security with
data segmentation at the application, metadata, and
data access levels all while providing the flexibility of
business driven big data analysis.
• Data quality—With big data, the data relationships are
not explicit and data may be loaded in “data lakes,”
or large-scale storage repositories. Because it is largely
unstructured, big data demands a different approach
in measuring, improving and certifying the quality
and integrity of the data. Companies may wish to
consider involving business stakeholders to determine
confidence intervals for the quality of big data and
appointing data stewards who are accountable to an
information governance council for improving metrics
over time. Processes will likely need to be redesigned
for understanding the needs of the business, exploring
incoming data, and then defining the data quality rules
for big-data elements that are critical to producing the
desired business outcomes. In terms of technology
enablers, sophisticated data cleansing solutions that are
based on machine-learning and predictive modelling
algorithms can be helpful in correcting deficiencies and
optimizing and updating quality rules. Additionally, data
visualization can be leveraged by analysts to help identify
data outliers and measure/monitor data quality.
• Data lifecycle management—The volume and variety of
big data make it hard to understand the regulatory and
business requirements that dictate what data to retain,
compress, archive or delete. Within the data governance
organization, responsibilities should be assigned for
defining retention and disposal rules, potentially
collaborating with the legal and risk departments in the
process. Ultimately, the retention schedule should be
expanded to include big data, and the associated legal
requirements should be documented and socialized
throughout the enterprise. In enabling data lifecycle
management, data compression techniques can be
helpful in reducing data bottlenecks, thus improving
performance and decreasing total cost of ownership.
Similar, but expanded
On the whole, an effective, big-data governance
organization differs from the traditional governance
structure in terms of its breadth (i.e., the people involved, as
well as their roles and responsibilities) and its collaborative
capabilities. This expanded organization essentially serves as
a cross-functional gatekeeper of all of the commercial and
clinical information that is delivered within the enterprise.
While Big Data Organization is similar to that of
the traditional there will be major shift in the Roles
and Responsibilities
As in traditional governance models, the various levels
within the transformed data governance organization are
responsible for:
• Finding it (data acquisition)—Includes the strategic
sourcing of data, managing those who provide or
produce it, and setting proper expectations on the
completeness quality, timing, consistency, and use
of data.
• Storing it (data integration)—Encompasses verifying that
the data is current, complete, consistent, compliant,
accurate, and actionable; owning the data dictionary
and business rules that govern importation and storage;
governing how data is accessed, used, and updated; and
managing stewardship operations.
• Delivering it (information access)—Comprises owning
access and authorization on reporting platforms;
ensuring report availability; creating “certified” self-
serve data access; ensuring consistent interpretation of
key performance indicators (KPIs); and responding to
ad hoc requests.
Data Governance
Working Group
Governance Steering Team
Data
Governance
Council
Governance Steering Committee includes executive level group from
security, risk and procurement who can make strategic decisions
considering the scope and attributes of the data to be managed.
As the big data changes rapidly, the responsibility of
management and implementation of big data is delegated
to the data owners and working group by the council in
alignment with the attributes determined by the council.
More specialized and varied skillset will be required
for executing and taking responsibilities for various
core data management roles, including data stewards,
metadata lead, data architect, data quality lead,
data scientists and data custodians.
Copyright © 2016 Deloitte Development LLC. All rights reserved.

Overall, the transformed governance organization should
seek to fulfill these fundamental responsibilities by defining
and assigning data owners, stewards and governors at every
level to ensure quality and consistency of information; and
by facilitating collaboration among all of the groups that
create, manage and use data. Within the life sciences sector,
these groups often encompass R&D and clinical, commercial
operations, contracting and managed markets, sales and
marketing, manufacturing and supply chain, finance, safety
and medical affairs, legal and compliance, IT and data
management, and advanced analytics.
Look ahead of the curve
As life sciences organizations expand their governance
approaches to accommodate big data, they will also
need to consider the technologies available to meet the
information needs of each these groups while supporting
key personnel in fulfilling their data-management roles
and responsibilities. To respond to the escalating demands
and opportunities associated with big data, life sciences
companies should look for an information management and
analytics platform that is ahead of the curve—meaning that
it encompasses several leading-edge capabilities that are
available, but not commonly found, in technology offerings
today. Ideally, it should have the ability to:
• Allow users to see data very quickly without first going
through extensive technical processes to store and
harmonize it
• Diagnose data integration issues quickly
• Access and integrate useful, reliable external data
• Enable individuals to explore data and discover insights
on their own
• Incorporate publicly governed data libraries, such as
those housing economic statistics, geographic data,
weather information, etc.)
• Create departmentally governed data libraries with
business-driven metrics that are managed by the business
and shared across departments
• Produce smart, intuitive visuals that provide
meaningful insights
• Share and collaborate with anyone, anywhere while
maintaining the security and integrity of the data set
• Satisfy a broad range of needs across the organization
with a single tool
• Use data visualization tools to support governance
processes by allowing users to monitor data quality and
identify anomalies
By evolving their data governance models and enabling
them with advanced information management technologies,
life sciences companies can begin to uncover the many
insights buried within big data. And, they can gain greater
confidence that the data sets they are using, and the
analyses they are performing, are accurate and assured.
Information governance enables better information
usage and increased data adoption
Conclusion
The opportunities and risks associated with big data are
much too large to be ignored. In order to capture the
potential of big data to fuel growth and enable innovation,
while managing the new types of risks that accompany
it, life sciences organizations may need to transform their
legacy data governance structures. Today, a centralized
governance organization is still an important part of
the solution, but more is needed. Transformative data
governance requires both distributed and centralized
participation, and it provides an environment for keeping
both big data and traditional data safe and compliant, while
simultaneously ensuring that the right users have access to
relevant insights in the necessary context to drive innovation.
Ability to access
data promplty
Ability to meet diverse
business information needs
Accommodates a
changing internal and
external data environment
Works with and complements
existing teams and processes Leverages across the business and
infrastructure and processes already
in place. Avoid redundancy
Availability Accessibility
Completeness
Visibility
EfficiencyIntegration
Scalability
Flexibility
Potential Benefits
of Transformative
Data Governance
Rules, methodologies
understood and known
by all consumers
Ability to control data
completeness though SLAs
with providers and defined
communication process
Built for business enablement,
not technology savvy

As used in this document, “Deloitte” means Deloitte Consulting LLP, a subsidiary of Deloitte LLP. Please see www.deloitte.com/us/about for a detailed
description of the legal structure of Deloitte LLP and its subsidiaries. Certain services may not be available to attest clients under the rules and
regulations of public accounting.
This publication contains general information only and Deloitte is not, by means of this publication, rendering accounting, business, financial,
investment, legal, tax, or other professional advice or services. This publication is not a substitute for such professional advice or services, nor should
it be used as a basis for any decision or action that may affect your business. Before making any decision or taking any action that may affect your
business, you should consult a qualified professional advisor. Deloitte shall not be responsible for any loss sustained by any person who relies on this
publication.
Copyright © 2016 Deloitte Development LLC. All rights reserved.
Data is the currency of the future. For more information on how transformative data governance can help your organization
to leverage this currency in the life sciences sector, please contact:
Contacts
Scott Barnes
Director
Deloitte Consulting LLP
scbarnes@deloitte.com
Anjan Roy
Specialist Leader
Deloitte Consulting LLP
anroy@deloitte.com
Peter Leddy
Director
SI Solutions
Qlik
Peter.leddy@qlik.com
Christopher Ferrara
Director
Life Sciences
Qlik
Christopher.ferrara@qlik.com
About Deloitte and Qlik
Through a strategic alliance, Deloitte and Qlik collaborate in helping organizations to improve the way they access, visualize, and interact with their
data. Together, Deloitte and Qlik offer powerful solutions for analytics and data discovery, leveraging Deloitte’s knowledge of leading practices in data
management and Qlik’s business intelligence and data visualization platform. Industries served include life sciences & healthcare, financial services, and
supply chain & logistics, among others. As a result of its significant experience with Qlik delivery projects, Deloitte has received several Qlik Partner of
the Year and Developer awards, including 2015 NAM System Integrator Partner of the Year, 2015 Benelux System Integrator Partner of the Year, and
2015 APAC Partner of the Year.

LS_WhitePaper_NextGenAnalyticsMay2016

More Related Content

What's hot (17)

Similar to LS_WhitePaper_NextGenAnalyticsMay2016 (20)

LS_WhitePaper_NextGenAnalyticsMay2016