Transform Raw Social Media Data into Real Competitive Advantage

There’s real competitive advantage buried in today’s deluge of social media data. If you know how to analyze it, you can increase your relevance to customers, establishing yourself as a trusted supplier in a cutthroat environment where consumers rely more than ever on “public opinion” about your products, services, and experiences.

Social Media Analytics is the complete insider’s guide for all executives and marketing analysts who want to answer mission-critical questions and maximize the business value of their social media data. Two leaders of IBM’s pioneering Social Media Analysis Initiative offer thorough and practical coverage of the entire process: identifying the right unstructured data, analyzing it, and interpreting and acting on the knowledge you gain.

Their expert guidance, practical tools, and detailed examples will help you learn more from all your social media conversations, and avoid pitfalls that can lead to costly mistakes.

You’ll learn how to:

• Focus on the questions that social media data can realistically answer
• Determine which information is actually useful to you—and which isn’t
• Cleanse data to find and remove inaccuracies
• Create data models that accurately represent your data and lead to more useful answers
• Use historical data to validate hypotheses faster, so you don’t waste time
• Identify trends and use them to improve predictions
• Drive value “on-the-fly” from real-time/ near-real-time and ad hoc analyses
• Analyze text, a.k.a. “data at rest”
• Recognize subtle interrelationships that impact business performance
• Improve the accuracy of your sentiment analyses
• Determine eminence, and distinguish “talkers” from true influencers
• Optimize decisions about marketing and advertising spend

Whether you’re a marketer, analyst, manager, or technologist, you’ll learn how to use social media data to compete more effectively, respond more rapidly, predict more successfully…grow profits, and keep them growing.

Foreword    xviii

Preface: Mining for Gold (or Digging in the Mud)    xx

Just What Do We Mean When We Say Social Media?    xx

Why Look at This Data?    xxi

How Does This Translate into Business Value?    xxii

The Book’s Approach    xxiv

Data Identification    xxiv

Data Analysis    xxv

Information Interpretation    xxvi

Why You Should Read This Book    xxvii

What This Book Does and Does Not Focus On    xxix

Acknowledgments    xxxi

Matt Ganis    xxxi

Avinash Kohirkar    xxxi

Joint Acknowledgments    xxxii

About the Authors    xxxiv

Part I: Data Identification

Chapter 1: Looking for Data in All the Right Places    1

What Data Do We Mean?    2

What Subset of Content Are We Interested In?    4

Whose Comments Are We Interested In?    6

What Window of Time Are We Interested In?    7

Attributes of Data That Need to Be Considered    7

Structure    8

Language    9

Region    9

Type of Content    10

Venue    13

Time    14

Ownership of Data    14

Summary    15

Chapter 2: Separating the Wheat from the Chaff    17

It All Starts with Data    18

Casting a Net    19

Regular Expressions    23

A Few Words of Caution    27

It’s Not What You Say but WHERE You Say It    28

Summary    29

Chapter 3: Whose Comments Are We Interested In?    31

Looking for the Right Subset of People    32

Employment    32

Sentiment    32

Location or Geography    33

Language    33

Age    34

Gender    34

Profession/Expertise    34

Eminence or Popularity    35

Role    35

Specific People or Groups    35

Do We Really Want ALL the Comments?    35

Are They Happy or Unhappy?    37

Location and Language    39

Age and Gender    41

Eminence, Prestige, or Popularity    42

Summary    45

Chapter 4: Timing Is Everything    47

Predictive Versus Descriptive    48

Predictive Analytics    49

Descriptive Analytics    53

Sentiment    55

Time as Your Friend    57

Summary    58

Chapter 5: Social Data: Where and Why    61

Structured Data Versus Unstructured Data    63

Big Data    65

Social Media as Big Data    67

Where to Look for Big Data    69

Paradox of Choice: Sifting Through Big Data    70

Identifying Data in Social Media Outlets    74

Professional Networking Sites    75

Social Sites    77

Information Sharing Sites    78

Microblogging Sites    79

Blogs/Wikis    80

Summary    81

Part II: Data Analysis

Chapter 6: The Right Tool for the Right Job    83

The Four Dimensions of Analysis Taxonomy    84

Depth of Analysis    85

Machine Capacity    86

Domain of Analysis    88

External Social Media    88

Internal Social Media    93

Velocity of Data    99

Data in Motion    99

Data at Rest    100

Summary    101

Chapter 7: Reading Tea Leaves: Discovering Themes, Topics, or Trends    103

Validating the Hypothesis    104

Youth Unemployment    104

Cannes Lions    2013    110

56th Grammy Awards    112

Discovering Themes and Topics    113

Business Value of Projects    114

Analysis of the Information in the Business Value

Field    115

Our Findings    115

Using Iterative Methods    117

Summary    119

Chapter 8: Fishing in a Fast-Flowing River    121

Is There Value in Real Time?    122

Real Time Versus Near Real Time    123

Forewarned Is Forearmed    125

Stream Computing    126

IBM InfoSphere Streams    128

SPL Applications    129

Directed Graphs    130

Streams Example: SSM    131

Step 1    133

Step 2    134

Step 3    134

Step 4    135

Steps 5 and 6    136

Steps 7 and 8    136

Value Derived from a Conference Using Real-Time

Analytics    138

Summary    139

Chapter 9: If You Don’t Know What You Want, You Just May Find It!: Ad Hoc Exploration    141

Ad Hoc Analysis    142

An Example of Ad Hoc Analysis    144

Data Integrity    150

Summary    155

Chapter 10: Rivers Run Deep: Deep Analysis    157

Responding to Leads Identified in Social Media    157

Identifying Leads    158

Qualifying/Classifying Leads    160

Suggested Action    161

Support for Deep Analysis in Analytics Software    163

Topic Evolution    163

Affinity Analysis in Reporting    165

Summary    167

Chapter 11: The Enterprise Social Network    169

Social Is Much More Than Just Collaboration    170

Transparency of Communication    171

Frictionless Redistribution of Knowledge    172

Deconstructing Knowledge Creation    172

Serendipitous Discovery and Innovation    172

Enterprise Social Network Is the Memory of the Organization    172

Understanding the Enterprise Graph    174

Personal Social Dashboard: Details of Implementation    175

Key Performance Indicators (KPIs)    177

Assessing Business Benefits from Social Graph Data    183

What’s Next for the Enterprise Graph?    185

Summary    186

Part III: Information Interpretation

Chapter 12: Murphy Was Right! The Art of What Could Go Wrong    189

Recap: The Social Analytics Process    190

Finding the Right Data    193

Communicating Clearly    195

Choosing Filter Words Carefully    198

Understanding That Sometimes Less Is More    198

Customizing and Modifying Tools    201

Using the Right Tool for the Right Job    204

Analyzing Consumer Reaction During Hurricane Sandy    204

Summary    209

Chapter 13: Visualization as an Aid to Analytics    211

Common Visualizations    212

Pie Charts    213

Bar Charts    214

Line Charts    216

Scatter Plots    218

Common Pitfalls    219

Information Overload    219

The Unintended Consequences of Using    3D    220

Using Too Much Color    221

Visually Representing Unstructured Data    222

Summary    225

Appendices

Appendix A: Case Study    227

Introduction to the Case Study: IBMAmplify    228

Data Identification    228

Taking a First Pass at the Analysis    234

Data Analysis    241

A Second Attempt at Analyzing the Data    243

Information Interpretation    244

Conclusions    247

Index    249