How to Create and Manage a Successful Analytics Organization

Preview:

Citation preview

Mario Faria

1

How to Create and Manage a Successful Analytics Organization

Mario Faria

fariamario@hotmail.com +1 - (425) 628-3517

@mariofaria

Mario Faria

2

Who am I ?

•  MIT recognition as one of the 1st Chief Data Officers and Data Scientist

Leaders in the world (just Google “Mario Faria Chief Data Officer”) •  20+ years working with Information Technology, Management

Consulting, Financial Services, Retail, CPG and Private Equity •  Proven expertise in Data Management, Data Science, Analytics, CRM

and Supply Chain Management •  Speaker at several conferences on the subject in USA, Europe and Latin

America •  Contributor to magazines and publications •  Big Data Advisor TPN at the Bill and Melinda Gates Foundation •  Member of the MIT Data Science Initiative •  Helping companies cross the Big Data Chasm

Mario Faria

3

Objectives of this webinar

•  Provide insights on how you should successfully create a Data organization

•  With that in place, you will be able to work effectively with Big Data projects

Mario Faria

4

My mission : To help the data community evolve with sustainability

Mario Faria

5

By being a consultant, I want to say 3 things ...

Mario Faria

6

The 3 things:

•  Situation : where the market is at this point •  Complication : current issues with Data

Management, Big Data and Analytics •  Solution : what I recommend you to do and how

to do it

Mario Faria

7

Situation

Mario Faria

8

How we got here

Mario Faria

9

Evolution of Business Intelligence

Mario Faria

10

The 4 driving factors that are changing the technology industry as

we know it •  Social •  Mobile •  Cloud •  Information

Mario Faria

11

This brave new world we are living in

•  How does success look like in a world where consumers are now marketers ?

•  Where a trillion data points are available, alive and transforming decisions (preference / purchase) and relationships as we speak ?

•  How to understand, connect and consistently engage with consumers and customers creating loyalty and recommendations ?

Mario Faria

12

Mario Faria

13

“The balance of power in the 21st century is influenced by the ability to leverage information assets” – Gwen Thomas, CEO of The Data Governance Institute

Mario Faria

14

Data is about •  People •  Technology •  Processes •  Modeling •  Statistics •  Communication •  Decisions •  Actions A data-driven culture is a disruptive factor for entire industries

Mario Faria

15

SQL

MAPREDUCE

HADOOP

CLOUDSCALE

MPI

BSP

PREGEL

DREMEL

PERCOLATOR

What is Big Data?

Mario Faria

16

Mario Faria

17

Mario Faria

18

What is Analytics ?

“The extensive use of data, statistical and quantitative analysis, explanatory and predictive models, and fact-based management to drive decisions and actions” – Thomas Davenport

Mario Faria

19

Analytics is transforming data assets into

competitive insights, that will drive business

decisions and actions, using people, processes

and technologies

Mario Faria

20

Analytic Maturity Curve

Mario Faria

21

Analytics is not just about : •  Large volumes •  Greater scope of information •  Real time access to information •  New kind of data and analytics •  Data influx from new technologies •  Non-traditional forms of media •  Variety of sources

It all of the above, plus a transformation in processes and culture, and it is a disruptive factor for entire industries

Mario Faria

22

Analytics is about customer centricity

•  Supply Chain forecasting •  Behavioral analysis •  Operations improvement •  Marketing targeting / decisions •  Real-time pricing / promotions •  Customer experience analysis •  Customer insights •  Customer lifecycle management •  Fraud prevention and analysis •  Network monitoring

Mario Faria

23

Predictive Analytics

•  Prediction is powered by the world's most potent, booming unnatural resource: data

•  Predictive analytics is the science that unleashes the power of data

Dr.Eric Siegel

Mario Faria

24

Big Data & Analytics

=

Human Behaviour

Mario Faria

25

Data Monitoring Centers

Mario Faria

26

Complication

Mario Faria

27

Land of Confusion

Mario Faria

28

Who owns the Data inside an organization ?

Mario Faria

29

Some problems, at this point, in most organizations

•  Data is fragmented and scattered •  Silos of information hanging around •  Like the truth, data has many versions •  The Data Lifecycle is a complex process •  Data projects being managed by IT •  A formal process to manage data is a

requirement in order to do Analytics

Mario Faria

30

The problem : data is an abstract concept

Mario Faria

31

The complexity of the Data Life Cycle

The Big Data

Technology Players

Mario Faria

33

The evolution path to Big Data

Mario Faria

34

Confusion between Big Data and Hadoop

•  Hadoop is being wrongly treated as a synonym of Big Data

•  Hadoop is one of the technologies to be used at Big Data projects

•  Hadoop is a great technology for storing unstructured data in an expensive and scalable manner, in a high granularity

•  What Linux did to Operating Systems, Hadoop is bringing to Information Management

Mario Faria

35

The Hadoop Ecosystem : growing everyday

Mario Faria

36

The Big Data Fragmented Tech Vendors : data life cycle process view

Mario Faria

37

Understanding Hadoop/MapReduce

Usage Output/Input

(records)

Job Input Size

GB PB

Best case scenario

Mario Faria

38

An analogy of using MapReduce

Traditional usage

MapReduce usage

Mario Faria

39

And, unfortunately, technology alone will not change the previous results

To succeed in Data & Analytics, an organization will be required to change some of its current internal processes

Mario Faria

40

The catch : just a few companies (users and consulting) understood the nits and grits about Analytics : it requires you to moving from a simple data management vision (tactical) to an information management vision (strategic)

Mario Faria

41

Solution

Mario Faria

42

Find a real object that people can relate to

Mario Faria

43

The Data Value Chain

Mario Faria

44

The Deming Model : Production Viewed as a System

Mario Faria

45

What is Data Quality ?

•  Quality is a customer perception •  A few dimensions: freshness, coverage,

completeness, accuracy •  It is a never ending job

Mario Faria

46

Usage of wrong data can destroy credibility

Mario Faria

47

A Few Quality Programs

TDQM

TIQM

Mario Faria

48

More and more, Data Leaders are being hired to think strategically think about all the steps from getting raw data and making it useful to

business users

Mario Faria

49

Foundations of the Analytics team responsibilities

•  Data Strategy •  Data Analytics •  Data Insights •  Data Architecture •  Data Governance •  Data Quality •  Data Acquisitions •  Data Operations •  Data Policies •  Data Security •  Data Protection

Chief  Data  Officer  /    Head  of  Analy6cs  /    Data  Scien6sts  

Mario Faria

51

Chief Data Officer (CDO) / Chief Analytics Officer (CAO) /

Lead Data Scientist

Mario Faria

52

The role of a Chief Data Officer or Lead Data Scientist

A data scientist is the one who looks for insights

The insight is operationalized in BI/DW products, by data architects

The insight is shared with the enterprise The CDO or Lead Data Scientist is the

executive responsible and accountable for the data life cycle inside the organization, managing the people involved in the data activities, such as acquisitions, analytics,

processes, governance, quality, technology and budget

Mario Faria

53

Mario Faria

54

Why should not IT be managing this transition ?

Because data projects are business projects, not IT projects and the CDO/Data

teams are the bridge between IT and Business Units

Mario Faria

55

The Chief Data

Officer Role

Mario Faria

56

The 3 Architectures a Company needs to succeed

Business Architecture

Technology Architecture

Data Architecture

Mario Faria

57

Why do you need a Chief Data Officer ?

Mario Faria

58

Why do you need a Chief Data Officer ?

•  Data is about business, it's not about IT

•  Data is an economic asset, so you need a senior person to handle the data initiatives.

•  As an economic asset, data needs: control, show value and monetization

•  There is now way you can do Advanced Analytics unless you have some data management practices in place.

Mario Faria

59

“Organizations are about to be swamped with massive data tsunamis. The Chief Data Officer is responsible for engineering, architecting, and delivering organizational data success” – Peter Aiken, PhD

Mario Faria

60

Mario Faria

61

A Chief Data Officer is the executive responsible to

manage these areas

Mario Faria

62

•  A good CDO can implement a data & analytics organization with success

•  A great CDO has the ability to turn raw data into large revenue streams for the business

•  Components such as technology and methodologies are important, but they are just enablers

•  The CDO focus is delivering enterprise value to the business (not writing code or SQL scripts)

From good to great CDO

Mario Faria

63

The evolving CDO role will challenge structure, scope and power relationships between executive committee members. The scarcity of information leader talent will require executive leaders to develop it as much as hire it.

Mario Faria

64

At the end, on Big Data, a CDO and the team should

•  Support the data initiatives, using the assets from different sources, with quality as a requirement

•  Drive business insights, so the users can act promptly

•  Execute his/her tasks fast, in real-time if possible

Mario Faria

65

The main drivers for Analytics projects

•  Make more money •  Reduce current costs •  Improve efficiency

Mario Faria

66

What it takes to make Analytics projects drive results

•  Data – understand what they have and how to be creative when it comes to using internal and external data

•  Models – focus on developing models

that predict and optimize •  People – transform their organizations

with tools and effective training so that managers can take advantage of Big Data's insights.

Mario Faria

67

To start an Analytics Team inside, there are 4 main things to consider

People Technology

Process to implement the

Practice Methodology for

the Delivery

Mario Faria

68

From good to great, an analytics team must have:

•  Passion for analytics and data •  Never stop learning •  Always be there for tough analytics

questions •  Ask questions until everything makes sense

and you are satisfied with the answers and analyses

•  Learn how to develop prototypes quickly •  Be an advocate for building a strong

foundation in corporate analytics •  Be a "bridge builder" between IT and

business users

Mario Faria

69

Looking ahead in the near future …

Mario Faria

70

Which companies will thrive in 2015? •  The ones which will understand how to adapt faster to

this new scenario •  The ones which will have successful Analytics

implementations •  The ones with great human capital, which understand

how to leverage their resources and with proven methodologies to embrace this change

Mario Faria

71

Is your company going to lead, influence or follow when using data

and analytics to drive results ?

What does it take to succeed in

this Analytics journey ?

Mario Faria

73

Major points on how to structure an Analytics program

•  Upper management buying and support •  Do not reinvent the wheel : use and abuse of best

practices that already exist •  Communicate always and be transparent •  Quick wins And …

Mario Faria

74

Hire the best and most eager resources you can find

Mario Faria

76

“Business are complex systems, optimizing a single element rarely creates lasting value”- Peter Drucker, the father of modern management

Q&A

Mario Faria

78

Thank you Mario Faria Data Strategy Advisor http://www.linkedin.com/in/mariofaria/ Founder of the Digital Mad Men www.slideshare.com/fariamario Twitter : @mariofaria fariamario@hotmail.com +1 (425) 628-3517

Recommended