23
© Copyright 2010 Hewlett-Packard Development Company, L.P. 1 © Copyright 2010 Hewlett-Packard Development Company, L.P. Social intelligence: applying new insights to business Michael Procopio @MichaelProcopio LinkedIn.com/in/MichaelProcopio

Social media & web analytics innovation procopio-2012-04

Embed Size (px)

DESCRIPTION

My presentation from the Social media & web analytics innovation conference April 2012 in San Francisco put on by theiegroup.com. I cover examples of gaining insights on products from social media conversations

Citation preview

Page 1: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    1© Copyright 2010 Hewlett-Packard Development Company, L.P.   

Social intelligence: applying new insights to

business

Michael Procopio @MichaelProcopioLinkedIn.com/in/MichaelProcopio

Page 2: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    22 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

Customer Intelligence Transformation Model

Grow and Retain your Customers

Build Brand Loyalty Beyond Reason

Improve Marketing

Effectiveness

360 Custome

r ViewInfrastruct

ure

Syndicated

Pyschographic

Social Media

User Survey

Primary Research

Unstructured DataInDirect

Channel

Transactions Demographi

c

Touch History

Dynamic Preferences

Digital Bread

Crumbs

Information Asymmetry

Page 3: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    33 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

#2 A strong negative polarity in media signal for a product may lead to increased support tickets.

Research questions

Hypothesis #1 A strong media signal for a product may lead to increased sales/registrations.

We developed two hypothesis to test the business benefits for the project

Focus

Page 4: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    44 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

“I don’t know what people are complaining about regarding the software but it installed seamlessly and is intuitive in its operations. Even though I am dissatisfied with the paper tray all together I am happy with this printer.”

Attributes

Measures

Making sense of unstructured data

Sentiment Attributes

Page 5: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    55 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

Minimizing the knowledge gap with merging unstructured with structured data

Product Profile

Product ID

STRUCTURED DB

UNSTRUCTURED DB

Product ID

# purchases # tickets # service orders

product launch date

2,131,517 856 98 10/1/2010

Cartridge Paper Tray

Price Printer

0 +1 -1 +1

Product Understandin

g

Perception • Sentiment• Intentions• Suggestions• Influencers

Behavior

• Sales• Frequency• Value• Service

Page 6: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    66 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

Findings – Hypothesis A

Products Data Reviewed

Photosmart printer 1

Photosmart printer 2

384 Consumer reviews from Amazon, CNET and HP Forums

6 Professional reviews from CNET, PC World

515K purchases

Hypothesis

Drivers#1 A strong media signal for a

product may lead to increased sales/registrations.

Consumer & Prof Reviews, Consumer & Prof Ratings,Rating Source*

Data

Page 7: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    77 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

Findings – Hypothesis #1 – Photosmart A

Analysis Results A strong media signal for a product

may lead to increased sales/registrations. r=0.58

r=0.72 (excludes holiday season)

•Medium to high level of Correlation

•Media signal from consumer forums tends to follow purchase trends

Purchases Weighted Rating (Consumer + Professional)

Page 8: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    88 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

Findings – Hypothesis #1Photosmart B

r=0.59r=0.96 (excludes holiday season)

•Medium to high level of Correlation

•Media signal from consumer forums tends to follow purchase trends

Analysis Results

Purchases

Weighted Rating (Consumer + Professional)

A strong media signal for a product may lead

to increased sales/registrations

.

Page 9: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    99 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

Hypothesis Status

Partially Confirme

d

A strong media signal for a product may lead to increased sales / registrations.

Hypothesis

Analysis/Findings

• Moderate Correlation between overall media signal and sales trends

• Strong Correlation when adjusted for seasonality

Possible Business Application• Opportunities for Outbound

Marketing Optimization (marketing mix – spend, audience, channel)

• Increased sales and operations forecast accuracy by monitoring and measuring consumer interactions in digital media

Page 10: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    1010 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

Findings – Hypothesis #2 – Photosmart A

r =—0.62 (Overall product quality)

r =—0.95 (Print Quality)

•High Correlation

•Correlation strength increases as the attribute granularity increases (e.g. Print Quality)

Analysis Results

Consumer negative sentiment

Tickets/Purchase

A strong negative polarity in media signal for a product

may lead to increased support tickets

Page 11: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    1111 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

Findings – Hypothesis #2 – Photosmart B

r =—0.9 (Overall product quality)

r =—0.7 (Print Quality)

•High Correlation

Analysis Results

Consumer negative sentimentTickets/Purchase

A strong negative polarity in media signal for a product may lead to increased support tickets.

Page 12: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    1212 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

Findings – Hypothesis #2Hypothesis Drivers

A strong negative polarity in media signal for a product may lead to increased support tickets.

• Consumer Sentiment Score

• Product Attributes

Products Data Reviewed

Photosmart Printer 1Photosmart Printer 2

384 Consumer reviews and sentiment from Amazon, CNET and HP Forums 54K Service tickets

Data

ConclusionsHypothesis Status Analysis/Findings Possible Business Application

Confirmed

Strong correlation between negative media signal and customer support calls.

Operational Efficiency: i.e. managing the Customer support resources based on consumer sentiments

Proactive Product quality validation and response

Page 13: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    1313 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

DEMO 1

Page 14: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    1414 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

DEMO 2 – VALUE FOR MONEY

Page 15: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    1515 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

DEMO 3 – REMOVE TWITTER

Page 16: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    1616 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

DEMO 4

Page 17: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    1717 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

DEMO 5 – REMOVE VALUE FOR MONEY

Page 18: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    1818 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

DEMO 6 – BRAND POSITION

Page 19: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    1919 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

DEMO 7 – TIE TO INTERNAL DATA

Page 20: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    2020 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

DEMO 8 – FIND ADVOCATES

Page 21: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    2121 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

Situation

ENHANCE WARRANTY & CLAIMS MANAGEMENT WITH SOCIAL INTELLIGENCEPRIMARY EUROPEAN AUTOMOTIVE MANUFACTURER

The warranty & claims, quality management and product development departments gathered information related to quality issues and product defects through the warranty process. •Time delay in getting data•Labor intensive to get data•DWH system not flexible enough for ad hoc analysis to answer “unknown questions”.

Social Intelligence solution

Page 22: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    2222 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

Solution Results

Enhance existing search and analysis platform with Social Media data

Ease of use front end, like “Google” search.

Enable users to do intuitive root cause analysis, identifying cause connections between warranty cases and Social Media discussions

More effective analysis of structured and unstructured information

Identifying potential defects and quality issues over a month earlier through social media information

Understanding customer needs to be considered in future developments

Social Intelligence solution

Page 23: Social media & web analytics innovation procopio-2012-04

© Copyright 2010 Hewlett-Packard Development Company, L.P.    2323 © Copyright 2010 Hewlett-Packard Development Company, L.P.   

Thank You

Your questions?

@MichaelProcopio

LinkedIn.com/in/MichaelProcopio