The Insight Value of
Social Sentiment
Seth Grimes
Alta Plana Corporation
@sethgrimes
IIeX – Atlanta
June 13, 2016

Dion Hinchcliffe
https://www.enterpriseirregulars.com/58013/social-media-marketing-predictions-for-2013-part-1/

Three types of (social) listening:
• Strategic
Market research & consumer insights.
• Reactive
Customer engagement.
• Retroactive
Customer experience.
Listening to what?
The Enterprise POV
Claire Michard,
http://www.dictanova.com/wp-
content/uploads/2016/02/logo-rvb-sur-
blanc-BD.png

http://www.greenbook.org/grit
Social

“Sentiment analysis is the task of identifying positive
and negative opinions, emotions, and
evaluations.”
– Wilson, Wiebe & Hoffman, 2005, “Recognizing
Contextual Polarity in Phrase-Level Sentiment
Analysis”
… and not just from text.
Facts & Feelings
© Jorge Royan / http://www.royan.com.ar / CC-BY-SA-3.0

What are people saying? What’s hot/trending?
What are they saying about {topic|person|product} X?
... about X versus {topic|person|product} Y?
How has opinion about X and Y evolved?
How has opinion correlated with {our|competitors’|general}
{news|marketing|sales|events}?
What’s behind opinion, the root causes?
(How) Can we link opinions & transactions?
(How) Can we link opinion & intent?
Who are opinion leaders?
How does sentiment propagate across channels?
Can we predict the impact of messages?
Can we find our own optimal design/messaging/channels?
From Facts & Feelings to Insights

For social sentiment:
• Text analytics
…via natural language processing (NLP).
• Image analysis.
• Video analysis.
• Digital (content) metrics.
Start with text…
Mechanisms

Data Type Examples Applications
Verbatims from
quantitative surveys
Brand trackers,
customer/ employee
engagement studies
Advert/Product testing
likes & dislikes
Enterprise Feedback
Management/Voice of
the customer
programmes
What are people saying?
Auto-coding of verbatim into themes
What is good and bad?
Measure and track sentiment overall and by themes
What should we improve first?
Text based impact/driver analysis for action prioritisation
Real-time analysis of customer feedback
Prioritise actions and interventions in timely fashion
Qualitative surveys
Online communities,
focus group transcripts,
public consultation
transcripts
What is being said and in what context?
Extract key words and patterns in data
Can we map the key themes?
Visualise patterns and relationships between concepts
What should wider analysis focus on?
Guide/focus analysis
Social media data
User Generated
Content from social
networks, data from
forums/ blogs/review
sites, news sites etc...
What are people saying?
Auto-coding of comments into themes
What is good and bad?
Measure and track market sentiment
What is being said about our competitors?
Competitive intelligence
Is there anything new we should be worrying about?
Identify early signals of user/customer opinion
Feedback/data held
by organisations
Call centre
logs/recordings, data
from website, email
data etc..
What have people said in the past?
Audit of existing data
Is any of our existing data
relevant for our current task?
Find relevant data for further analysis
What are people saying?
Identify recurrent themes
https://www.ipsos-
mori.com/researchpublicatio
ns/publications/1766/A-
Guide-to-Text-Analytics.aspx

New people measures:
• What do they feel when they experience X?
• How do they talk about their experiences and feelings?
• Intent.
Timely measures:
• Emerging issues; early warning.
• Leading indicators, anomalies, and trends.
Correlation, confirmation & root cause.
Voices:
• The public; own and competitors’ prospects, customers,
lapsed customers, and employees.
• Personas, affinity communities, influence networks.
Eight Insights

Experiences

Feelings/Talk

Correlation

Root Cause

Competitive Voices

Personas & Affinities
http://www.slideshare.net/Affinio/affinio-analysis-
of-dota-vs-codworldleague

SentimentSymposium.com
IIEX code = 20% off

The Insight Value of
Social Sentiment
Seth Grimes
Alta Plana Corporation
@sethgrimes
IIeX – Atlanta
June 13, 2016