the roadmap for successful big data adoption · milestones of big data adoption . 8 . unaware ....
TRANSCRIPT
This presentation, including any supporting materials, is owned by Gartner, Inc. and/or its affiliates and is for the sole use of the intended Gartner audience or other authorized recipients. This presentation may contain information that is confidential, proprietary or otherwise legally protected, and it may not be further copied, distributed or publicly displayed without the express written permission of Gartner, Inc. or its affiliates. © 2013 Gartner, Inc. and/or its affiliates. All rights reserved.
Svetlana Sicular
Twitter: @Sve_Sic
Blog: http://blogs.gartner.com/svetlana-sicular
The Roadmap for Successful Big Data Adoption
Gartner Definition of Big Data: High-volume, velocity and variety
information assets that demand cost-effective, innovative forms of
information processing for enhanced insight and decision making.
Gartner Research Circle 2013 Big Data Survey
720 Respondents
Worldwide
$3.2B Mean
Company Size
5,100 Mean
Employees
60% Mainstream
Adopters
18% Focused on
Running/Maintaining
Are They Investing?
30% Have
31% No plans
at this time
19% Plan to within
the next year
15% Plan to
within two years
5% Don't know
How Does That Compare to Last Year?
Note — Survey base increased from 473 in 2012 to 720 in 2013
27
15
16
11
30
19
15
31
5
Have invested
Within next year
Within two years
No plans
Don't know
2013 2012
0 10 20 30 40
Things Are Done Differently in Silicon Valley
Traditional IM • Requirements based
• Top-down design
• Integration and reuse
• Technology consolidation
• World of DW and ECM
• Competence centers
• Better decisions
• Commercial software
"Big Data" Style
• Opportunity oriented
• Bottom-up experimentation
• Immediate use
• Tool proliferation
• "World of Hadoop"
• Hackathons
• Better business
• Open source
Business Process: Information Centric
Information-centered tomorrow
Info. & Content,
Analytics, Metadata
Structured Process
Resource Optimization
Case Management
Business Applications
Analytics
Meta data
Information &
Content
Applications-centered today
Build bridges between the world in
information and the world of process
Process Application Information Information Process Application
The Data-Driven Enterprise
7
The Road Map: Typical Stages and Milestones of Big Data Adoption
8
Unaware
Experimental
Opportunistic
Strategic
Transformative
Aware
Time
Value
Initial implementation of the LDW
Traditional technology cannot meet all needs
Data-Driven Enterprise
Big data initiative is justified
Big data strategy planned
Stabilized big data infrastructure
Information governance is a must
Data products emerge
Big data is becoming the new normal
Ramp up (investments outstrip returns)
A milestone
9
Unaware
Experimental
Opportunistic
Strategic
Transformative
Aware
Time
Value
Initial implementation of the LDW
Data-Driven Enterprise
Big data initiative is justified
Big data strategy planned
Stabilized big data infrastructure
Information governance is a must
Data products emerge
Traditional technology cannot meet all needs
“Something about the way we do work today is
either broken or is about to break."
Stage 1: Unaware
Simultaneous Shifts Brought About by the Nexus of Forces
Customer
Borderless
Physical
Ubiquitous
Mobile
Enterprise
Highly Structured
Digital
Handful
Desktop
Business Focus
Organization
Scope
Devices
Workstation
Shifts
Stage 1: Unaware
11
Organizational Inertia Data
Growth
Traditional Technology
Cannot Meet All Needs
BI Immaturity
Outsourced IT
Changed Customer
Expectations
Actions to Progress to the Next Stage
Start information governance
Evaluate your BI maturity and plan BI improvements
Educate IT and businesspeople about big data opportunities to spark their interest in big data adoption
Stage 1: Big Data Myths
A Common Myth
Reality
12
Big data is about volume
Big data is high-volume, high-velocity and high-variety information assets that demand cost-effective, innovative forms of information processing for enhanced insight and decision making
Big data = Hadoop
Big data is a technology stack that satisfies the definition of big data. A logical data warehouse (LDW) is the target architecture for combining data warehousing and big data technologies
Our current EA and EDW solutions will work for a long time without architecture changes
Enterprise architecture (and the EDW as part of it) is a process not a goal, it reflects the organizational evolution necessary for the business survival under the forces of the nexus.
13
Unaware
Experimental
Opportunistic
Strategic
Transformative
Aware
Time
Initial implementation of the LDW
Data-Driven Enterprise
Big data initiative is justified
Big data strategy planned
Stabilized big data infrastructure
Information governance is a must
Data products emerge
Traditional technology cannot meet all needs
Value
"An understanding of when to use big data is
lacking right now."
Stage 2: Aware
In Search of Solutions
14
Doctor, I have a big problem
You have a BIG DATA problem!
Do You Have a Big Data Problem?
Stage 2: Aware
15
No Business Case for Big
Data Adoption
Skunkworks With New
Technologies
Big Data Strategy
Limited Information
About Actual Big Data
Deployments
Business Does Not
Understand Big Data
Opportunities
Digital Marketing in
the Cloud
Actions to Progress to the Next Stage
Select big data use cases that represent best business opportunities
Assess the state of your information governance and plan its improvements
Get educated about big data technology, people and process.
Stage 2: Big Data Myths
A Common Myth
Reality
16
All our unresolved problems are big data problems
Big data technologies are best for solving business problems where questions are not known in advance and the data types to answer these questions could vary and might need unstructured data
Everybody is implementing big data technologies and we are behind our peers
The majority of enterprises, including those touted by the media as successful adopters of big data technologies, are at Stages 2, 3
Big data is created outside our enterprise
“Dark data” — an underutilized data in existing data stores — is of greatest interest to all enterprises who investigate big data opportunities.
17
Unaware
Opportunistic
Strategic
Transformative
Time
Initial implementation of the LDW
Data-Driven Enterprise
Stabilized big data infrastructure
Information governance is a must
Data products emerge
Traditional technology cannot meet all needs
Experimental Aware
Big data initiative is justified
Big data strategy planned
Value
"Being highly focused on the business driver
is fundamental to success in big data."
Stage 3: Experimental
The Internet of
People
18
Opportunities are in Interconnectedness
The Internet of
Things The Internet of
Data
The Internet of
Ideas
@
Data Scientist Is Just a Part of the Solution….
A Multi-disciplinary Team:
• Mix of business and technology is necessary
• Techies are predominantly introverts but big data requires extroverted understanding of customers, partners and competitors
• The need for specialized skills that are non-traditional for the enterprise (such as linguists or anthropologists)
• Team members inspire and enrich each other.
Increase Maturity:
• Team and teamwork
• Elements of big data analytics
• Big data analytics process
• IT
• Team members can later represent specific functional groups or sub teams
The Goal Is to Share the Big Data Experience
Stage 3: Experimental
20
Wrong Evaluation
of a Big Data Initiative
Business Leadership
Justification of a Big Data Initiative
Lack of Skills
Absence of Information Governance
Identified Big Data
Use Cases
POCs
Actions to Progress to the Next Stage
Start forming a multidisciplinary big data team of business and technology professionals
Filter out best business ideas for big data, prove their value, select appropriate technologies and justify big data adoption in the well-informed business case.
Stage 3: Big Data Myths
A Common Myth
Reality
21
All big data opportunities are evaluated by the same criteria
Big data opportunities are of two types: game-changing and business extension. Evaluate them differently.
Big data technology is mature for the enterprise needs
Many technology gaps exist. They require workarounds or might not possible to mitigate.
A data scientist will implement a big data solution for our enterprise.
A multidisciplinary team with diverse skills can meet technology challenges and solve complex business problems of big data adoption.
What's Still Needed?
Security
Data Warehousing Tools
Governance
Distributed Optimization
Subproject Optimization Skills
23
Unaware
Experimental
Opportunistic
Strategic
Transformative
Aware
Time
Initial implementation of the LDW
Data-Driven Enterprise
Big data initiative is justified
Big data strategy planned
Stabilized big data infrastructure
Information governance is a must
Data products emerge
Traditional technology cannot meet all needs
Value
"When you integrate disparate data together,
it's not clear what the exposure is: We're afraid
of what we might discover and not having
control of it."
Stage 4: Opportunistic
Big Data Governance: Value, Reuse, Risk and Compliance
USE CASE
& Data
25
The Heat Map of Big Data Governance Priorities
What From To Business
Value Reusability Compliance Risk
Business Focus
Enterprise Consumer
Organization Highly Structured
Borderless
Data Confined Liberated
Devices Handful Ubiquitous
Scope Digital Physical
Workstation Desktop Mobile
Software Proprietary Open Source
Solutions Deterministic Probabilistic
Applications Monolithic "Fit for purpose" in App Stores
Technology Adoption
IT Business
High Medium Low
12 Dimensions of Big Data
• Volume — sources x produced data
• Variety — different types of structures
• Velocity — timing and streams
• Complexity — standards, rules, formats
Classification Contracts
Technology Pervasive Use
Perishability Fidelity
Validation Linking
Velocity Volume
Variety Complexity
• Perishability — shelf life
• Fidelity — reusability of data and results
• Validation — understanding of use case
• Linking — relationships between elements
• Classification — access controls
• Contracts — SLA
• Technology — fragmented tools
• Pervasive Use — usage levels
Stage 4: Opportunistic
27
Siloed Big Data Implementations
Business Extension Use Cases
Stabilized Big Data
Infrastructure
Lack of Specific Skills
Reactive Information Governance
Information Governance
Additional Use Cases Amplify the Value of Big
Data
Actions to Progress to the Next Stage
Revisit your initial assumptions and adjust your big data strategy
Make information governance an integral part of big data adoption
Stabilize infrastructure and operations for big data technologies
Start building an LDW to transform your business strategy and operations for the data-driven economy
Stage 4: Big Data Myths
A Common Myth
Reality
28
Hadoop is cheap to implement
Initial implementation is more expensive than expected. It involves a lot of unanticipated technical and non-technical difficulties.
Big data infrastructure usually takes advantage of commodity hardware.
Enterprises buy high-end hardware.
Programming for big data technologies is difficult.
Programming for big data technologies is easy for programmers.
29
Unaware
Experimental
Opportunistic
Strategic
Transformative
Aware
Time
Initial implementation of the LDW
Data-Driven Enterprise
Big data initiative is justified
Big data strategy planned
Stabilized big data infrastructure
Information governance is a must
Data products emerge
Traditional technology cannot meet all needs
Value
"When mining big data, you'll find unexpected
(but real) results. Don't start a project if you're
unwilling to deal with the findings."
Stage 5: Strategic
Logical Data Warehouse
Which Big Data characteristic is the
biggest issue for your organization?
Stage 5: Strategic
Technology Immaturity Leads
to Custom Solutions
Game-Changing Business Ideas
Logical Data Warehouse
Talent Attrition
Variety of New
Technologies
Actions to Progress to the Next Stage
Treat big data as the new normal.
Operationalize infrastructure like for any ongoing initiative. Develop support, provisioning and capacity planning.
Complete the initial stages of the LDW by implementing a variety of integration and virtualization tools.
Stage 5: Big Data Myths
A Common Myth
Reality
32
Organizations can complete building out their LDW in several years
The LDW will keep evolving. It is being built for extensibility, not for completion. Implement it in incremental releases that deliver capabilities for your roadmap toward a data-driven enterprise.
The ratio between unstructured and structured data is 80:20. Unstructured data contains 80% of information value.
80% represents share of the quantity of unstructured data, not necessarily value of it. Structured data has been refined and its density and quality are much greater than comparable amounts of unstructured data.
Companies can completely migrate off RDBMS
RDBMS technologies serve many purposes better that big data technologies.
33
Unaware
Experimental
Opportunistic
Strategic
Transformative
Aware
Time
Initial implementation of the LDW
Data-Driven Enterprise
Big data initiative is justified
Big data strategy planned
Stabilized big data infrastructure
Information governance is a must
Data products emerge
Traditional technology cannot meet all needs
Value
“During this transition, some leaders left the
company, some people were forced out, and
some people changed their mind.”
Stage 6: Transformative
Big Data (not to scale)
Fish Grows Bigger in a Bigger Tank
Big Data Causes a Mindset Shift
• Asking bigger questions
• Finding new answers
• New ideas
• Openness
• No return to small tank
The Emergence of Collaborative-decision-making Platforms
CDM Environment
• Reliable and secure
• Integrated with systems of record
• Complex decision support (workflow)
• Capture best practices Social Networking
• Examine relationships of decision makers
• Map information flows
• Intelligent social profiles
Decision Tools
• Simulations
• Optimization tools
• Scenario planning
• Mind mapping
• Brainstorming
• SWOT
• Predictive analytics
• Prediction markets
• Decision methodologies
Collaboration
• Shared workspace
• Communication (email, IM, phone, etc.)
• Web conferencing
People
• Involve right people to inform the decision
• Incorporate expert and diverse opinion
• Minimize bias
Information
Decision Tools Collaboration
Social Networking People
CDM Components
Information
• Access to any data source and decision input
• Search, content analytics, visualization tools
• Business intelligence content
• Assumptions and pattern detection and monitoring
Stage 6: Transformative
36
Technology Inability to Meet Leaders' Needs
Collaborative Decision-Making
Platform
Information Products
Non-Data-Driven Employees
Information Trust Models
Stage 6: Big Data Myths
A Common Myth
Reality
37
Only internet companies can be at this level
Any enterprise can become data driven. In fact, there are already such organizations virtually in every industry.
We will sure find a solution for all our problems.
There are problems which inherently do not have solutions. Do not persist if a solution does not seem to be possible. In other words, fail fast.
Our company will remain a leader for a long time
The business environment and technology change fast. Continuously monitor these changes and adjust your strategy to stay ahead of competition.
38
The Data-Driven Enterprise
Recommendations
Develop or adjust your big data strategy by using the big data adoption roadmap
An initial goal of big data adoption must be learning and growing big data expertise
Build your big data strategy by planning improvement targets for four dimensions at each stage: business, people, process and technology
Avoid common big data fallacies
It is extremely important to understand not only the stages but how to move from one stage to the next.
Recommended Gartner Research
The Road Map for Successful Big Data Adoption Svetlana Sicular (G00254965)
Decision Point for Practical Big Data Use Cases Svetlana Sicular (G00239633)
A Framework for Evaluating Big Data Initiatives Svetlana Sicular (G00246250)
Information Governance in the Age of Big Data Svetlana Sicular (G00251071)
How to Design and Implement the Next-Generation Data Warehouse Mei Selvage (G00219658)
To learn more about how Gartner for Technical Professionals can assist you, visit [email protected] or contact your
Account Executive.