Here is a POV regarding the state of Social Media analysis and how it compares/ contrasts to that of more traditional Market Research. This was done in the context of Healthcare / Pharma.
1. The Social Media Business Process – SM vs. MR
Social Media Listening Outputs Market Research Outputs
Conversation Pre Qualified / Sorted
Cloud Panel or Focus Group
Brand / Indication
Targeted Survey /
Share of Voice / Interview
Conversation 60%
40%
20%
Psychographics
Influencer
Ethnographies Personas Attitudinal Insights
• On-Topic Post Demographics
• Author Reputation Relative
• Author Reach Authority
• Platform Authority
Social Media Listening can serve as a loose proxy for more traditional market
• Platform Reach
research in that the attitudinal insights derived from observed conversations can
• Content and Context Sphere of be developed in near real-time – as associated to MarCom events or in the absence
Resonance with the Influence / Reach of them. While the granularity as expressed in the outputs from market research are
Community / Audience
a level or depth of analysis deeper than that of Social Media – we can derive similar
insights from Social Media based upon the same qualitative research
methodologies – it’s just that the digital medium is inherently different and the
analysis required is exhaustive.
2. Quantitative Methods vs. Qualitative Methods
Social Media Listening is a blend of both Quantitative and Qualitative
research methodologies. As the medium is dependent upon sorting, filtering,
and segmenting the digital space and distributed / syndicated content (volume
as well as significant spam in many cases) - Quantitative methodologies are the
backbone. It’s an exercise in separating the signal from the noise to identify
authoritative, on-topic conversations, the associated influencers, and the various
indications surrounding sentiment.
Traditional Market Research is also an exercise in both Qual and Quant, but
with an emphasis on the more qualitative approach and output – with pre-
defined / pre-vetted (targeted) questions and the applicable (targeted) results.
These results help to inform Brands / MarCom regarding the attitudinal
inclinations of a target audience in association to a set criteria, set of
circumstances, or product/brand attributes. Part of the target audience vetting
process allows the researcher to set or obtain demographic information,
psychographic information, ethnographic information, attitudinal indications, etc.
3. The State of Influence 2010-2011, Part I
Influence has many meanings in Social Media measurement circles, but the
basic factors of influence are:
1. A timely, relevant, resonant on-topic post with appropriate strategic
keywords as associated to the Brand, Product, Initiative, Campaign, etc.
2. Content and contextual relevance and resonance within:
• A particular community
• A wider audience (potential viral effect)
3. The relative authority of the on-topic post author:
• Reputation / Trust within the community
• Reputation / Trust within wider circles in traditional and/or social media
• Reputation / Trust within a group of peers
• Reputation / Trust within an area of expertise (SME or subject matter
expertise).
4. The relative authority of the particular platform or outlet from one or more of
these various perspectives:
• Syndication / Distribution / Reach
• Platform / Outlet Popularity
• Semantic nature of platform technology
• Platform / Outlet Reputation in industry, market, community etc.
4. The State of Influence 2010-2011, Part II
There are many characteristics and factors that create Influence in Social as well
as Traditional Media. The problem with creating a standardized system for
Influencer Identification is that any and all standardization systems will be
obsolete from one analysis period to the next or worse, be completely arbitrary or
subjective. This is the nature of digital influence as represented by the infamous
and often foolishly sought-after or aspired-to viral effect. I cringe at the words,
“Can you make my campaign Go Viral?”
Influence can happen on a granular, post-by-post basis – in the absence of many
of the characteristic “authority factors” as represented on the previous slide. We
need only look backward a few years to see that an intrepid blogger on a
Wordpress platform could quite easily “hijack” a brand strictly on the basis of
platform authority due to a search engine friendly, semantic structure and the
coincidental, optimized use (SEO) of strategic keywords.
Influence or Influencer Identification is therefore an exercise in moment-by-
moment (period-by-period) analyses by the individual analysts qualified to make
the proper evaluations. Influence, much like Sentiment resists a standardized,
mechanical, automated approach to assessment. And Influence, much like
Sentiment, requires a rigorous, qualitative analysis on a regular basis.
5. Sentiment Analysis in 2010/2011
Sentiment Analysis has been approached in the various social media
measurement and monitoring/listening tools based upon a pre-defined set of
keywords selected for their attitudinal indications regarding sentiment. With
the advancement of Boolean and nearness operators in social media
sentiment analyses – in support of the Net Promoter Score (NPS)
methodology – sentiment is approaching a nominal degree of accuracy.
Traditional market research approaches sentiment analysis in the way that
gives a much higher degree of accuracy according to current capabilities. In
traditional market research, a focus group (n=xxx) is surveyed for attitudinal
indications (sentiment) from a mainly qualitative perspective.
In social media, the same Best Practice can be followed. We can agree upon
a reasonable sample (n=xxx), and perform a qualitative analysis in order to
determine accurate sentiment as applied in an index/scale. The effort is the
same as in traditional market research and requires a skilled analyst to
comb-through the conversations and make the appropriate evaluations.
6. Social Media Actionable vs. Anecdotal Insights
Actionable Insights
Engagement Non-Engagement
Regulatory Issues /
Restrictions
Influence the Inform Brand
Influencer Lexicon
Leverage Influencer Inform MarCom Limited Regulatory
Issues / Restrictions
Platform Strategy & Tactics
Inform Creative
Limited Regulatory
Concepting
Issues / Restrictions
Editorial Process
& Calendar
Information without action is simply overhead (anecdotal)
7. Patient Journey Optimization, Part I
Stage in Patient Journey Example Terminology / Keyword Phrasing
Awareness Opportunity: hyperactivity, impulsive behavior,
Symptoms Language short attention span
Awareness Opportunity: adhd, attention deficit, adhd
Interest / Diagnosis / Disease Info treatment / medication
Engagement Opportunity: concerta, concirta, concerta side
New Rx / Treatment / Efficacy effects, concerta trial
Engagement Opportunity: concerta dosage, taking concerta,
Compliance / Adherence / Retention concerta newsletter
Source: “Five Trends in Digital Healthcare in 2010” by Amy Cowan, Head of Healthcare | Google
8. Patient Journey Optimization, Part II
Drivers / Assets Landing Pages / Assets
Create Composite Reports
to Develop Full Traffic /
Appropriately Targeted Referral Attribution and Appropriately Segmented
Creative, Content, or Insight Content or Instance of
Messaging (e.g. Display, Conversion (Brand Web site,
Search, or Content Microsite, Unbranded Site, etc.)
% CTR /
Awareness Conversion Awareness
% CTR /
FICO DB / Business Web Analytics /
Education Objects / eCRM DB Omniture SiteCatalyst
Conversion Education
% CTR /
Rx Conversion Rx
% CTR /
Adherence Conversion Adherence
Leverage Campaign ID
Display & Search Social Media Listening / According to Journey Stage
Methodology
(Instrumented URLs)
Trafficking / DART / Radian6 / Alterian SM2 Disposition (%) – where does
DFA / CheckM8 / Atlas / MyBuzzmetrics / the Brand need to reinforce?
and/or T-Code
SERMO / etc.
By “Front-Loading” or Segmenting the Brand creative and
content and appropriately tracking the referrals or
engagement instances (e.g. clicks), we can develop the full
attribution model for campaigns and content in order to begin
Search / SEO Market Syndicated to surgically apply SEM, SEO, SMO, Display, CRM,
Tools Research Research Creative Concepting, Editorial, etc.