big data initiative justification and prioritization framework

13
BIG DATA INITIATIVE JUSTIFICATION Framework to justify and prioritize big data initiative Authors - Neeraj Sabhnani ( Enterprise Strategy Sr. Consultant in Microsoft Services) Balarama V Raju ( Enterprise Strategy Sr. Consultant in Microsoft Services)

Upload: neerajsabhnani

Post on 22-Nov-2014

607 views

Category:

Technology


5 download

DESCRIPTION

Framework to justify and prioritize Big Data initiatives. The deck will also help in developing roadmap for big data projects

TRANSCRIPT

Page 1: Big data initiative justification and prioritization framework

BIG DATA INITIATIVE

JUSTIFICATIONFramework to justify and prioritize big data initiative

Authors -

Neeraj Sabhnani ( Enterprise Strategy Sr. Consultant in Microsoft Services)

Balarama V Raju ( Enterprise Strategy Sr. Consultant in Microsoft Services)

Page 2: Big data initiative justification and prioritization framework
Page 3: Big data initiative justification and prioritization framework

the amount of data generated by enterprises

is expected to grow by 48% this year and

90% of it will be unstructured data

Big Data is a top three priority at Walmart

70% of Big Data projects revolve

around customer facing ventures—

driving sales & boosting retention

Page 4: Big data initiative justification and prioritization framework

Limited implementation of big data

projects

Source – Survey results published in IBM report – Analytics real world use of big data

Page 5: Big data initiative justification and prioritization framework

Risks and Challenges for Big Data

Projects • Big data technology is evolving and many organizations

are waiting for it to stabilize

• Big data solution might not be needed for all problems,

existing analytical solutions might be well equipped to

provide business benefits

• Organizations that have not addressed the more

traditional requirements of storage, processing and

information architecture need to carefully weigh the use of

big data solutions against more traditional ones

Organizations need due diligence on benefits and risks before initiating

big data initiative and not just go with market hype

Page 6: Big data initiative justification and prioritization framework

Justification FrameworkStep 1

Business

Relevance

Step 2Technical

Complexity

Step 3Economic

Viability

Step 4

Pilot

Success

Step 5Implementation

and Adoption

Organizations can use this justification framework for due diligence of big

data initiatives and for developing the roadmap

Page 7: Big data initiative justification and prioritization framework

Business Relevance

Step 1

Identify

beneficiary of

data analysis

(organization

department)

Step 2

Identify

Organization/dep

artment goals &

objectives

Step 3

Identify functional

use cases

Step 4

Map functional

use cases against

business goals

Page 8: Big data initiative justification and prioritization framework

Typical Functional Use Cases

Customer

Insights

• Customer insights can help to identify valuable customers, help to attract more and better customers, retain valuable customers longer. Successful enterprises are able to attract more profitable customers as compared to competitors, drop undesired customers and retain their best customers by knowing them better than their competitors do

• Typical sources for customer information : � Customer information through channels – stores, web, phone, catalogs� Web logs having customer click stream information showing customer preferences,

buying patterns, testing website features to attract more visitors.� Third party information � Publically available data � Information from social media

Product

Marketing

• Industries are facing challenges to reduce design cycle times and costs, satisfy global regulations, and satisfy customers that expect high-quality, well-designed products

• Typical sources for product information:� Product use data � Product feedback sent to manufacturers� Customer reviews� Social media data � Publicly available data � Data from patents organization

Page 9: Big data initiative justification and prioritization framework

Typical Functional Use Cases

Operations

Objective is to reduce operations cost through monitoring of devices and processes for failures and problems, issuing SLA alerts for running out of capacity, or troubleshooting and preventing application outages• Typical data sources are :

� Logs- web, application, transactions etc.� RFID data� GPS data

Fraud

detection &

prevention

• Social data is widely used to detect fraud. Medical claims, insurance claims, online retail or Web click fraud are areas where big data analytics can play an important role through social media data

• Big data technology gives a high-granularity view of the social networks and other relationships, therefore resulting in a substantially clarified picture of fraud activities

Risk

Management

• Organizations can increase the sophistication of risk calculation by using more data (longer time span) and additional data from multiple sources

• Typical data sources are :• Social data � Credit history� Assets� Web logs, event logs� Publicly available data

Page 10: Big data initiative justification and prioritization framework

Technical Complexity

1

2

3

4

5

Complexity Level

Single Dataset Simple Analysis

Single DatasetMultiple type of

Analysis

Linked Dataset Simple Analysis

Linked DatasetMultiple type of

Analysis

Linked Dataset with

transactional data,

unstructured data

Multiple type of

Analysis

Page 11: Big data initiative justification and prioritization framework

Sample Use Case –

Customer Insights

Customer Insights Use Case

AttractingNew customers

Customer Retention

Innovation Product Expansion

Technical ComplexityMapping

Customer sentiment

analysisX X X X 2

Customer segmentation X X X 3

Customer lifetime value X X 4

Customer churn X 4

Customer campaign X X X 3

Recommendation

enginesX X X 2

Personalized website

optimizationX X X 2

Business Use

Case Business Goals & Objectives

Business RelevanceTechnical

Complexity

Unless there is business urgency for specific goal , prioritize uses cases meeting maximum

business goals and having least technical complexity .For above example roadmap can be :

Customer sentiment analysis, Recommendation engine and Personalized website optimization

Page 12: Big data initiative justification and prioritization framework

Economic ViabilityDifferent evaluation for initiative types

• Game Changers

• Business modifiers/extenders

Game Changers

• Have potential to provide

long term and significant

impact

• Impact of big data not

known in beginning

• Cost Benefit analysis might

not be appropriate method

for initiative selection

Business modifiers/extenders

• Cost benefit analysis can be

used for evaluation

• Typical initiatives can cover

efficiency improvement, cost

reduction, market expansion

etc.

• Rigorous analysis required

to determine if existing

technologies can work or

will require big data

For cost benefit analysis, consider all costs

• Infrastructure cost(hardware , software)

• Implementation cost

• Operations(ongoing) cost

Page 13: Big data initiative justification and prioritization framework

Pilot Success, Implementation &

Adoption

• Success for pilot is not just about technical

implementation but more about realized business benefits

from insights provided.

• Pilot or POCs might be needed for different use cases as

analysis requirements might vary across use cases.