Social Data Intelligence: Integrating Social and Enterprise Data for Competitive Advantage

Social Data Intelligence:
Integrating Social and Enterprise Data for
Competitive Advantage
By Susan Etlinger
With Charlene Li and Jessica Groopman
Includes input from 34 ecosystem contributors
A Market Definition Report
July 25, 2013

The State of Social Analytics
Making Social Data Actionable
Building a Data-Driven Organization
Six Dimensions of Analytics Maturity
Conclusion
Methodology
Endnotes
About Us
The average enterprise-class company owns 178 social accounts, while 13 departments — including marketing, human resources, field
sales, and legal — are actively engaged in social media.1 Yet social data, such as posts, comments, and reviews, are still largely isolated
from business-critical enterprise data collected from Customer Relationship Management (CRM), Business Intelligence (BI), market
research, email marketing, and other sources. The challenge is that, without consistent governance and strategy in place, social media
will become yet another silo within the organization, yielding isolated findings that lead to missed opportunities and risk.
Altimeter Group interviewed 34 organizations that are actively using social data in conjunction with enterprise data to make investment
and resourcing decisions, better understand customer needs, and discover unanticipated risks and opportunities. While integration of
social and enterprise data is still nascent and rife with challenges, we noted a clear progression of maturity in the application of social
analytics, from organizations who treat social data in an ad-hoc and siloed manner to those who view it as a business-critical asset, with
the discipline needed to make it strategic and sustainable. Those who approach data strategically already see measurable results in the
areas of opportunity and risk identification and improved customer experience, and they are able to identify and act on insights that they
may not otherwise have discovered.
Increasing the sense of urgency is the fact that much social data meets the criteria for “big data,” which encompasses signals from
sensors, mobile devices, digital assets, transaction records, global positioning systems, and a host of other sources — many generating
data in real time — that have not yet been invented or broadly implemented.
As social data continues to proliferate, the challenges faced by organizations will only compound. To make the most of this critical
resource — and avoid costly missteps — organizations must define a common approach to data collection, institute processes to share
and interpret it, and, most importantly, set clear criteria for action. The questions you must answer, sooner rather than later, are how
proactive you wish to be and how much investment and organizational disruption you are willing to face to realize the promise of next-
generation social data intelligence.
Executive Summary
............................................................................................................
........................................................................................................
................................................................................................
..............................................................................................
.........................................................................................................................................
......................................................................................................................................
............................................................................................................................................
............................................................................................................................................
Table of Contents
3
9
13
18
22
23
25
26

Social Data Is Not an Island
During the past year, Altimeter Group has seen an increase in the number of
organizations that are actively seeking to look beyond fans and follower counts and view
— even integrate — social data with enterprise data to understand if and how they relate
to each other. They want to:
• Detect signals that could affect the health, reputation, and financial performance of their
organization.
• Understand what people are saying about their products, services, and brand.
• Take action on this knowledge, and use the insight they gather to make critical business
decisions.
While we are still in the very earliest stages, social data is beginning to enter the
enterprise mainstream. Organizations need to know whether what they detect in
social channels is similar or different to what they collect in customer satisfaction
scores, market research, or through business intelligence and customer relationship
management platforms. They need not just social data, but social data intelligence,
which we define as follows:
Insight derived from social data that organizations can use confidently, at scale, and in
conjunction with other data sources to make strategic decisions.
The goal of social data intelligence is not simply to justify the existance of social
business initiatives, but to recognize at a strategic level that signals from the outside
world — as seemingly chaotic as they may be — now constitute critical input into
organizational decision making.
Viewing Social and Enterprise Data in Context
Because social data is mostly unstructured and — more to the point — still relatively
new, it tends to be largely separate from enterprise systems. While some organizations
are building social data warehouses or exploring other approaches to true technical
integration, most are just starting to look at social data intelligence as a potential driver
for enterprise decision making.
For the purposes of this report, “integrated” is to be taken broadly to mean that an
organization is viewing the output of separate systems in context of each other, rather
than integrating the data in a strict technical sense. Integrated systems and platforms
will evolve over time as enterprise systems become better adapted to social data, as
The State of Social Analytics
3
Social Data
Intelligence
Insight derived from social
data that organizations can
use confidently, at scale,
and in conjunction with
other data sources to make
strategic decisions.

organizations refine and automate their workflows,
and as tool ecosystems and APIs mature.
In an Altimeter Group survey of 27 enterprise-class
organizations, 70% stated that they are viewing
social data in the context of other enterprise data
to make business decisions.2
While this is a small
sample size and should be taken anecdotally, it’s an
early indication that social data is beginning to be
considered a legitimate source of intelligence.
The most common and logical initial point of
integration is business intelligence, as it is an
4
We integrate social data with ... (check all that apply):
45%
40%
35%
30%
25%
20%
15%
10%
5%
0%
Business
Intelligence
(BI)
Market
Research
Customer
Relationship
Management (CRM)
Email
Marketing
Sensor
Data
42%
35%
27% 27%
4%
Fig. 1: Organizations Begin to Integrate Social and Enterprise Data
Case Study: Parasole — Using Social Data to Protect Product Quality
Parasole Restaurant Holdings, a Minneapolis-based restaurant company with 1,400 employees, is known
for creating exceptional brand experiences. One of its brands, Manny’s Steakhouse, is celebrated for the
superior quality of its steaks; the company spends approximately $5 million each year on steaks alone.
Recently, Parasole and Manny’s noticed a steep uptick in negative feedback about its steak in both customer
satisfaction surveys and social media.
The company used the New Brand Analytics platform to track social mentions and discovered a sudden
change in sentiment related to its meat quality. Says Kip Clayton, VP of Marketing and Business Development
analytics-driven function to begin with (Figure 1). But
it’s important to note that market researchers and
customer experience and support teams are also
paying close attention to social data to identify trends,
both attitudinal and behavioral, that could influence
their experience and, ultimately, their loyalty.
The following case study illustrates how one
organization used social data in the context of
enterprise data to discover the extent of — and
mitigate — a business risk.

Challenges of Measuring and Integrating Social
Data
Social data presents methodological, interpretive,
and technical challenges that are quite different from
what has come before; when it comes to collecting,
computing, analyzing, and acting on social data,
organizations are in new territory. Unlike the classic
department-based model of business, in which teams
analyze clearly defined data sets, social media comes
from outside the organization and does not respect
boundaries such as database structures, processing
windows, or organizational design.
As to be expected with any disruptive trend, social
media creates challenges related to people, process,
and technology. Altimeter found the following to be
the three most common issues organizations face as
they try to interpret social data in the context of the
business.
1. Social Data Serves Multiple Internal Constituents —
and Interests
One of the characteristics of social media is that it
breaks silos; that is to say, it forces groups who may
not regularly work together to collaborate, or at least
to share information on an unprecedented basis.
Altimeter research indicates that, today, at least 13
departments are actively engaged in social media
throughout enterprise-class organizations (Figure 2).3
Far beyond marketing and corporate communications,
or even customer support, this participation extends
throughout the business: human resources, product
development, information technology, and legal.
Furthermore, while an email, survey, or phone call
is clearly directed to one individual or a small set of
people, social media is a two-way mechanism, and
the intent, intended recipient, and desired action is not
always clear or repeatable. As a result, social data can
have many stakeholders, and the processes to identify
the data, put it in context, benchmark it, share it, act
on it, and document the process appropriately can
vary dramatically.
The impact for social strategists — and analysts — is
that many more stakeholders within the organization
become actual and potential customers of the social
data intelligence. This can create tensions as social
data analysts strive to serve multiple constituents with
different needs. Here are several examples of how
individual departments use social data:
• Community managers and customer service
representatives need data to facilitate
engagement and case management on a one-to-
one basis.
5
at Parasole, “We noticed a dramatic increase in complaints across the board; this influx turned up the heat
very quickly, prompting us to act. Without seeing the comments in aggregate, this situation could have gone
on for months before we would be able to see the full extent of the problem. When manufacturing a product
that is sold on the same day, taking a year to identify it can be deadly. But by coupling social data with
personal experiences, we were able to pinpoint the precise dates, times, and incidences of the faulty product.”
Parasole and Manny’s identified six suspect samples, lined them up, tasted them, and immediately discovered
the root of the problem. The company acted quickly; it cut ties with the meat supplier, provided employee
training to smooth the transition, and updated its employee incentive programs to integrate social ratings and
reviews into the accountability process.
In this instance, Parasole used social data opportunistically — in conjunction with customer satisfaction surveys
and verbal feedback — to identify a quality issue with its meat supplier. While this was essentially a manual and
ad-hoc process, it highlights the opportunity of social data to supply context — in this case, about unsatisfied
customers that the company otherwise may have missed, or seen too late.

• Marketing and digital teams need performance
measures that help them understand how
campaigns are performing, often in real time,
so they can make adjustments that improve
performance.
• Risk and compliance, legal, and human
resources need the capability to discover quickly
when problematic terms or issues are escalating
that might present a risk to the company.
• Market researchers need to view social data
in the context of existing enterprise measures
of satisfaction, such as Net Promoter Scores,
customer satisfaction, and other surveys, to add
context to the information they’re gathering about
customer attitudes, behaviors, and intentions.
• Product developers and product marketers need
to understand customers’ likes and dislikes — not
only those expressed directly to the brand, but
also those shared with friends and followers — so
that they can learn more about them, address
product issues before they become critical, and
even incorporate their findings into product
roadmaps.
Given the variety of requirements, it’s not surprising
that no single tool or process can address every
scenario. Many organizations struggle with tool
proliferation, multiple analytical approaches, unclear
expectations, and strain on analytics resources while
having to constantly respond to a steady stream
of ad-hoc requests for different data cuts. This all
undermines the team’s ability to deliver a holistic — or
at a bare minimum — consistent analysis of the social
data.
The greatest danger, however, is that fragmentation
of social data undermines the organization’s ability
to reach what Filippo Passerini, Group President
of Global Business Services and CIO of Procter &
Gamble, calls “one single version of the truth”; a
foundational element for any business strategy,
and one that is increasingly important as social
data proliferates and is used throughout the
enterprise.4
Without a common version of the
truth, organizations are vulnerable to a variety of
disasters: missing a crisis, halting a highly successful
campaign, overlooking a customer service issue, or
underestimating a competitive threat.
6
Fig. 2: Thirteen Departments Are Actively Engaged in Social Media
In which of the following departments are there dedicated people (can be less than one FTE) executing social?
Marketing
Corporate Communications/PR
Customer Support
Digital
Social Media
HR
Product Development/R&D
Advertising
Experience
IT
Executive
Legal
Market Research
0% 10% 20% 30% 40% 50% 60% 70% 80%
73%
66%
40%
37%
35%
29%
16%
16%
15%
14%
11%
9%
8%

2. Because Social Data Is Different, It Requires New
Analytical Approaches
Many organizations continue to use traditional
marketing and advertising metrics, such as
impressions and Advertising Value Equivalency
(AVE), to gauge the value of their social programs.
The default assumption is that social media is a new
kind of broadcast channel through which to distribute
content and messages.
But social media — and social data — are highly
complex, and using traditional metrics to understand
it can distort its meaning, and ultimately, its value.
“The fundamental problem we have,” says Matt
Hixson, CEO, Tellagence, “is that few people are doing
genuine new research to understand social data.
They are using old models that were never designed
for social networks and are trying to fit the data into
them.” Aggregating social data with enterprise data
before understanding the native characteristics of
social data and networks is counterproductive, and
can compound interpretive problems.
Why is social data so different from what has come
before? One key reason is that it is a type of “Big Data,”
which Gartner Group has defined in terms of “The
Three Vs.”5
• Volume: Social data is characterized by
unprecedented volumes. Well-known brands can
reach hundreds of thousands of brand mentions
per day, while conversations on popular topics
number in the millions.6
• Variety: It’s comprised of multiple data types:
structured (stars and ratings), unstructured
(comments, posts), images, sound, and video.
Different types have different life spans; for
example, the average (unpromoted) Facebook
post or tweet has a fleeting life span, while a
Pinterest pin or a video can last forever.
• Velocity: Social data can be blindingly fast. One
common Twitter metric is “Tweets Per Second”
(TPS), which, for context, reached a peak of more
than 10,000 TPS in the last few minutes of Super
Bowl XLVI and more than 25,000 TPS during the
2011 Japan broadcast of the anime movie Castle
in the Sky. (In contrast, the Syfy premiere of
Sharknado generated 5,000 tweets per minute.)7
This explains why evaluating social data in terms of
traditional advertising metrics, such as “impressions,”
can set a team up for failure. Because of the inherent
volatility of the social web, mentions may be
extremely high one month and low the next, for a host
of reasons that a social strategist is unable to control
(for example, new product introductions or service
issues).
Furthermore, measuring impressions or ad value
equivalency without understanding the nature of
the signal — a share, a like, a positive or negative
comment — distorts the meaning of the signal,
leading to poorly informed decisions and, ultimately,
risk. Before we aggregate social data with other
enterprise data sets, we need to better understand the
nature of social networks and the data they create.
Here are seven core characteristics that make social
data unique — and require a different approach to data
collection and analysis (Figure 3):
When viewed in this context, it becomes clear
that social data is highly heterogeneous and rich
with meaning in a way that traditional marketing
and advertising metrics are simply not tuned to
account for. Understanding these signals, and
their relationships, is a prerequisite to the ability to
use social data strategically and in context of the
business.
3. Social Data — and Analysts — Lack Enterprise
Credibility
An open secret among organizations working with
social data is the credibility gap between social and
enterprise data, and, as a result, between social and
enterprise analysts. “Put bluntly,” says a representative
at a data analytics vendor who preferred not to
be quoted by name, “we see challenges where the
analysts on the social media team don’t have the heft
to make it with other data geeks in the organization.”
There are several drivers of this phenomenon:
7

Social data is new. Organizations do not have the
same comfort level with using social data for decision
making as they do with enterprise data from sources
such as customer relationship management and
business intelligence systems, so they may dismiss
critical signals that can provide context or identify
issues that they may not have detected otherwise.
Social data lacks standards. Unlike other forms of
enterprise data, such as web analytics and market
research data, there are no analytical standards for
social data. As a result, social data is — quite literally
— open to interpretation. Organizations struggle to
analyze it and align their findings across multiple
stakeholders. Furthermore, social analytics tools have
not yet reached a comparable level of maturity to the
types of business intelligence tools that executives
trust to inform decision-making.
The role of social data analyst is new. The role of
an enterprise data analyst is established and highly
professionalized, while the role of social data analyst
is newer, with no established standards with regard to
required expertise, particularly in quantitative analysis.
The lack of credibility of social data—and social
analysts—creates significant risk for organizations.
For analysts to be of strategic value, they must be able
to surface insights in the data that lead to decisions
and action. The question, however, is where to start,
and without accepted approaches, processes, and
standards, they lack the credibility and organizational
gravitas to make their case effectively. The solution
is to prioritize metrics in a way that stakeholders can
support, even if their priorities differ from others in the
organization.
The following pages will examine how to address
these three challenges. The first step is to identify and
prioritize which social data and metrics matter the
most to your organization.
8
Fig. 3: Key Characteristics of Social Data
Characteristic Description Considerations
Content Type
Social Action Type
Content Lifespan
Platform Demographic
Strength and Type
of User Relationships
Data Type(s)
Location Awareness
When measuring engagement, factor in the energy
expended to produce the social content. For example,
it’s more time-consuming to create a two-minute
video than a 140-character tweet.
Consider the result of the social action. A like on
Facebook puts content from that brand into your
news feed, while a share ensures that your friends
will see it as well.
Platforms, such as Facebook or Twitter, that deliver
content in the form of timelines or feeds have a shorter
lifespan than those that deliver content on pages or
channels, such as Pinterest or YouTube.
Posts, comments, tweets, photos, Vines, and
video require different levels of energy to create,
and can signal different levels of engagement.
A Facebook post is visible for up to a few
hours, while a Pinterest pin is permanent
(until removed).
Demographics vary. Snapchat has proportionately
high adoption by teens, while demographics
on other platforms are more evenly distributed.
Consider whether relationships among
users are based on social graph versus
interest graph.
Structured data, unstructured data, or a mix.
Availability of location data, via mobile,
GPS, or other sensor.
Re-tweets, favorites, shares, +1s, fans, likes, and
votes up and down carry nuances of
meaning and intention.
Facebook, Twitter, Tumblr, LinkedIn, and Snapchat
have somewhat different demographics.
Some platforms, such as Facebook and LinkedIn,
are generally based on relationships, while others,
like Youtube and Pinterest, are generally based
on interests.
Tweets and comments are narrative and therefore
unstructured, while star reviews are structured.
Product reviews are typically a mix of both.
Access to location data depends on the platform
being used as well as the user’s behavior. Some
platforms, such as Foursquare, inherently use
location data, while others, such as Facebook
Twitter, and Instagram, make is optional.

Define Your Core Social Metrics
A common dynamic among organizations Altimeter
interviewed was the delicate balance between
what data is readily available versus what should
be measured in a perfect world to best understand
the dynamics of the customer experience, the sales
process, brand reputation, and the many other areas
that social data can illuminate.
After all, we have relatively clean data on volume
metrics; it’s not hard to count likes or shares. It is far
more challenging, however, to connect them with
revenue generation, especially for organizations
with multi-channel operations, long sales cycles, or
ingredient brands.8 In some cases, analysis may
be a Herculean task involving multiple people, data
sources, Excel pivot tables, and manual calculations;
in others, it may rely on your social analytics tool’s
ability to deliver the desired metric on a dashboard.
The key is to focus at first on a few key metrics
that you believe are practical to deliver and have
the most impact on your business. Then formalize
those metrics with processes and dashboards
before expanding. Says Chris Kahle, Manager, Web
Analytics at Caesar’s, “The good news is that we can
track everything. The bad news is that we can track
everything. That will be the challenge for everyone.”
To address this issue, work backward from business
goals; they will be your “North Star” as you determine
what is and isn’t important to measure. Following are
some key recommendations:
1. Tie Metrics to Business Strategy. To make
social media more relevant to your organization,
you must tie it to something that matters to
the C-suite, primarily in the areas of revenue
generation, brand lift, and customer experience.9
Connect your social media strategy to business
objectives by forming a hypothesis. For example:
a) If your goal is to understand the impact of
social media on revenue, reverse-engineer the
process by starting at revenue-generating events
and connect them back to social content. There
are six primary ways to do this, documented
in Altimeter’s Report, The Social Media ROI
Cookbook.10
b) If your goal is to improve customer experience,
set a hypothesis that sentiment in social data
should correlate roughly to sentiment in other
data collection channels, such as surveys or Net
Promoter Score (NPS). This will help sensitize you
to fluctuations in the data, and look for drivers
that may indicate whether or not your hypothesis
is valid.
2. Define “Synthetic” Metrics. In addition to raw
volumes, you will be asked to demonstrate
increases in more nebulous phenomena, such
as “engagement,” “reach,” and “influence.” These
are “synthetic” metrics because they are a
combination of multiple metrics (often from
different data sources). None has a universal
definition because each business is different. But
they do tell you how people are interacting with
your organization’s people and content, how far
content is spreading, and which individuals or
groups influence the conversation.
a) Define these “synthetic” metrics so they reflect
what is important to your business and are
clear and consistent across all groups who use
social data. “Engagement” is the most important
to define early, as it is most widely discussed.
Document the ingredients you will use to
calculate engagement. Is it time on site? Shares?
Tweets and Retweets? Facebook PTAT (people
talking about this)? Or some other combination?11
b) Once you have defined “engagement,” follow
the same process to agree on definitions for
9
Making Social Data Actionable

“reach” and “influence,” as well as other metrics
that require some degree of consensus as to their
impact on your business. Measure your progress
by the extent to which you are able to drive
alignment around these definitions. One litmus
test is the ability to provide a consistent answer
to the question: “How do we define engagement
for our organization?”
3. Prioritize Your Metrics. While you should expect
ad-hoc requests to be a way of life, the most
strategic thing you can do is to use the “Social
Metrics Scorecard,” discussed in the next section,
to set expectations internally as to what data
you have readily available, the resources required
to analyze it, the value of the metric, and its
priority. This will establish that the analytics team
is thinking strategically about what is credibly
measurable, what is measurable with caveats,
and what is not yet measurable within the
organization.
The following case study describes how Caesar’s
Entertainment is using social data to better
understand customer touchpoints. The approach
at Caesar’s is a useful example of how one
organization tailors its measurement strategy to key
business drivers. Their approach is to focus first on
understanding the journey that customers take with
the brand — whether online or offline in their hotels,
casinos, or restaurants — and how that experience
may vary across segments, media types, and brands.
By prioritizing the most important metrics, Caesar’s
expects to better understand its customers’ unique
preferences and inform future investment decisions.
Metric Prioritization
Altimeter’s Metrics Scorecard provides a starting
point to help you evaluate and, more to the point,
communicate your organization’s capability to deliver
these desired metrics (Figure 4). This is a sample only;
use this template to catalog your existing social data
10
Case Study: Caesar’s — Integrating Social Data to Understand the Customer Journey
Objective
Caesar’s Entertainment, Inc. is an American public gaming company that owns and operates more than 50
casinos, hotels, and golf courses under several different brand identities throughout the world. As a result,
making sense of the vast array of data, systems, brands, audience segments, and transaction history — in
addition to a host of other data streams — is no easy task.
Caesar’s realizes that the value of its data lies in its ability to provide insight into the customer journey
across multiple channels and touchpoints. But customers don’t move linearly or predictably in an
increasingly digital world. Says Chris Kahle, Web Analytics Manager for Caesar’s, “We were looking at
attribution as direct vs. indirect. Social Direct means someone comes to any one of your social assets,
interacts with that asset, and then most companies drive that traffic directly back to their core site to
book (convert). With Social Indirect, people come on, then go talk to someone, then come back through
another channel, then head over to Google where they might come in on paid search or via email. If you
don’t have all channels integrated — including social — you won’t be able to see that.” As a result, when the
company looked at direct attribution numbers for the impact of social media on revenue, it didn’t seem as
though social was having much of an impact. What Caesar’s couldn’t see was the indirect, or multi-point,
attribution history.

11
Solution
Caesar’s deployed Adobe Social to better understand the customer journey and is now undergoing a mass
data integration project across the company, beginning with its customer relationship management and
enterprise data warehouse systems. In addition, the company is aggregating data across advertising
channels, such as display, email, organic, search, and affiliate, and is driving toward one integrated
marketing channel in order to gain a holistic and complete view of each prospect or customer’s relationship
with Caesar’s. Kahle explains, “By making sure all customer data is integrated, if customers interact with
us, we can see that. We’re also aiming to be able to track movement through all channels so we can
give credit to each for how it drives marketing and budget.”The goal is to understand both online and
offline touchpoints along the customer journey and how they may vary across segments, media types,
and brands. Part of Caesar’s strategy includes integrating online and offline data; for example, to enable
hotel concierges to use customer preferences to inform recommendations during their stay and enter
information after check-out to better personalize their future interactions.
Caesar’s is also leveraging data to inform smart engagement throughout the customer’s lifecycle. It is using
previous purchase and engagement history — both online and offline — to build preference models and is
tying together pre-purchase and rewards data to drive loyalty, enabling customers to earn points toward
rooms, discounts on shows, and other perks. The goal is to enable Caesar’s to offer the right rewards at the
right time through the right channel, tailored to customers’ specific interests.
From an organizational point of view, the company is gaining tremendous insights into the behavioral
preferences and tendencies of its audience segments to inform more efficient, strategic, and timely
investments in marketing and customer service offerings. Says Kahle, “Best practices are all theories right
now. We have the tool set to develop best practices to know what works and what doesn’t. First we want
to understand the entire model; then we’ll start building preference models based on customer profiles and
previous purchase history.”
Results
Although this project is still in the early stages, Caesar’s is already detecting insights at both a customer
and organizational level. Kahle is cautiously optimistic: “I think our customers will be receptive to it,” he
says. “They are already very engaged.” As the company continues to integrate social data with other
enterprise data sources, the focus will shift toward scaling this effort more broadly across the organization,
enabling Caesar’s to prioritize the insights they receive in a strategic and actionable way.
metrics, rate them, and determine which ones your
organization is most able to deliver and which will
require additional criteria to be met before they can be
adopted and shared.
Below is a short process overview of how to use this
sample metrics scorecard.
1. List the core set of metrics you would like to
evaluate.
2. Score them as follows, on a scale of one to five,
where one is lowest and five is highest:
a) Value: How useful this metric is to your
organization. For example, number of fans
and followers may be relatively low value,

12
Fig. 4: Sample Metrics Scorecard
while impact to revenue or brand health will be
much higher, as they correspond to existing
organizational metrics.
b) Capability: Your organization’s ability to deliver
this metric. It’s important to note that your
capabilities today are highly dependent on the
availability and stability of social network APIs, as
well as the analytics tools ecosystem. While there
will always be constraints, measurement will
become increasingly consistent as the industry
matures.
c) Resource: The time and staff power it will
take to deliver this metric. API changes and
gaps among tools mean that many analysts
are required to calculate core metrics manually
or using complex Excel pivot tables. Manual
calculations should be scored at the low end,
while automatic and/or readily available metrics
should be scored at the higher end.
d) Dependency: The degree to which other
metrics or future decisions rely on it. For example,
a metric describing conversation drivers provides
the foundation to understand many other
metrics, including share of voice of a particular
conversation, brand lift, or potential risk.
This will give you an overall score, which will constitute
an average of Value, Capability, Resources, and
Dependency. The final score will indicate which should
be your initial priority metrics versus those that require
additional conditions to be met (for example, the ability
to integrate web and social analytics platforms) before
you can deliver them at scale.
Once you have prioritized your metrics, it’s time to
focus on the organizational and process aspects of
integrating social data effectively into the fabric of the
enterprise.
This is a sample scorecard; these numbers are intended as placeholders and will vary according to business context
Category Metric Average
Value
Average
Capability
Average
Resources
Average
Dependency
Overall
Score
Brand Health
Brand Health
Brand Health
Brand Health
Revenue
Generation
Revenue
Generation
Revenue
Generation
Revenue
Generation
Revenue
Generation
Share of relevant
conversation
Sentiment over time
Sentiment drivers
Top Influencers,
Top Detractors
Relationship of Generation
Social to Awareness
Relationship of Social
to Consideration
to Conversion
to Loyalty
to Advocacy
Scale of 1-5: 1 is very low and 5 is very high
5 3 4 5 4.25
4 2 2 4 3
5 4 5 5 4.75
5 3 3 4 3.75
4 4 2 3 3.25
5 1 3 2 2.75
5 1 1 5 3
4 3 3 4 3.5
5 2 1 4 4

13
From Social Data to Organizational
Intelligence
For social data — any data — to be relevant and
credible to the business, it needs to move past
counting (how many?) to intelligence (what does it
mean?) to action (what should we do?). While this
process may be reasonably well established for
traditional enterprise data, it is still nascent for social
data. To ensure that social data becomes a strategic
driver for decision-making, analytics teams must
establish a set of expectations and processes early on
that demonstrate its integrity, reliability, and utility.
The Social Data Integration Maturity Map
The Social Data Integration Maturity Map categorizes
the phases of social analytics maturity across
six dimensions, from the initial ad-hoc, on-the-
fly approach to a mature, strategic, holistic, and
sustainable set of business processes that scale
across the organization (Figure 5).
Following is an overview of each maturity stage, its
primary characteristics, and the value created for the
organization.
Stage 1: Ad-Hoc
During the Ad-Hoc stage, the organization is beginning
its social data collection efforts, typically in one
department or work group, or several isolated work
groups. At this point, the organization typically begins
to track volume metrics: numbers of fans, followers,
and likes, with the objective of demonstrating
momentum within social media programs.
Organizations should use this phase as a learning
opportunity to understand as much as possible about
normal versus peak conversation volumes, how
people are talking about the brand in social channels,
where they congregate, what topics they discuss,
and which social media efforts indicate momentum.
This early learning stage is critical, as it will build
organizational confidence and competence and form
the basis for understanding the nature of social data
as it relates to other aspects of the business.
The next step is to show how social media supports
business strategy, rather than simply “social for
social’s sake” value propositions.
Stage 2: Formalized
During the Formalized stage, the organization
has taken what it learned during the Ad-Hoc
stage and has extended its efforts. It is beginning
to create workflows and processes for social
data, and has begun to scale analysis and create
common approaches and measures of success
across departments such as marketing, corporate
communications, and customer service.
Most importantly, the organization desires to
understand the business impact of its social media
efforts and move beyond volume metrics (social for
social’s sake) to understand how social data maps
to business outcomes such as brand awareness,
customer experience, and revenue generation. But,
while the organization has become more advanced
with regard to understanding the business relevance
of social media, social data still remains largely
separate from other data sets in the organization.
The extent to which organizations execute on Stage
Two will set a precedent for how the success of future
programs and campaigns will be evaluated and will
strongly influence later efforts to secure additional
budget and resources for subsequent social media
programs. For example, teams able to identify
significant opportunities or risks — such as that noted
by Parasole in its examination of social data related
to its meat quality — are able to reach consensus on
Building a Data-Driven Organization

Learning
Business Relevance
Learning
Business Relevance
Holistic View
Learning
Business Relevance
Holistic View
Scale
Learning
One department
or workgroup only
Across the
enterprise
Across 4
or more departments
Across
1-3 departments
Metrics are
volume-based
Social metrics
mapped to
business outcomes
Social metrics viewed
in context of other
business metrics and
mapped to business
outcomes
All social metrics
mapped to business
outcomes
Ad-hoc
Social data only
Social and enterprise
data sets viewed
in tandem
All relevant
customer-oriented
data sets
Undefined
Defined for
priority metrics
Clearly defined, with
known benchmarks
and shared KPIs
Defined for most
data, with a process
for establishing
new metrics
Processes defined
among social teams
Processes defined
among social and
other business teams
Entirely programmatic
and integrated into
established processes
and workflows
No benchmarks
for social metrics
Benchmarks
emerging, some
formally defined
Most metrics
benchmarked
Benchmarks
established
and shared
Social data only
Stage 4
Holistic
Stage 3
Integrated
Stage 2
Formalized
Stage 1
Ad-hocDimension
Desired
Outcome
1. Scope
2. Strategy
3. Context
4. Governance
5. Metrics
6. Data
14
Fig. 5: Social Data Maturity Map
the value of social data intelligence much earlier and
avoid costly and time-consuming “analysis paralysis.”
Organizations have mastered Stage Two when any
social program, campaign, or content can pass the
“So what?” test: How does this drive our business
forward? What value does it provide to shareholders,
customers, and stakeholders? What have we learned
that will influence future decisions?
The next step is to integrate that understanding with
other enterprise data for a more holistic view of the
business.
Stage 3: Integrated
Organizations moving into the Integrated Stage
already understand that social media can drive
business outcomes. Now they must determine how
findings from social data relate to other enterprise
data, such as CRM, business intelligence, or market
research. During this stage, organizations begin to
view social and enterprise data side-by-side to better
understand whether trends in one are detectable
in the other, whether social data provides context
for enterprise data, or vice versa. Some may even
incorporate econometric data such as interest or
inflation rates, gas prices, or other external factors
that can influence a buying decision or the market
landscape as a whole.
During this stage, the resource, organizational, and
tool challenges become more acute, as a holistic view
of data requires organizational alignment, shared
objectives and approaches, and rationalization of
what is generally a large and heterogeneous data (and
tool) set. One additional factor on the horizon is the

15
emergence of real-time (or as some refer to it) “right-
time” analytical capabilities, such as those offered by
SAP HANA, which will increase not only the potential
for actionable insight but the pressure to respond
effectively within hours, minutes, and even seconds.12
The truth is, this is hard work, and it requires the
ability to facilitate the integration of multiple groups,
agendas, and requirements. As a result, the most
critical criterion for mastering Stage Three is executive
demand and support for a holistic view of data and
the intelligence needed to inform strategic decision-
making.
If this executive involvement does not exist, or if social
data intelligence is not considered strategic enough
to champion, the organization will never achieve true
insight, even assuming the best possible people,
resources, and intentions are available. Without an
executive mandate for holistic intelligence, groups
become vulnerable to charges of empire building, and
the organization is at risk of feudal warring among
analytics teams.
Organizations will know they have mastered Stage
Three when they have a consistent framework for
business analytics, the ability to view and analyze
data from multiple streams in tandem, and the
processes needed to support intelligence across the
organization.
The next step is to do all of this efficiently and at
scale.
Stage 4: Holistic
While organizations in the Integrated stage have
the analytical and governance heft to view multiple,
heterogeneous data streams in context, the effort and
time required to do so is still significant, leading to real
trade-offs between cost and insight. In Stage Four,
the organization is now able to view multiple signals
in context and respond to them fluidly. The source
of the data — social, business intelligence, CRM — is
irrelevant; it’s the intelligence that matters.
Another key characteristic of the Holistic stage
is the ability to scale — to take what was learned,
operationalized, and integrated in Stages One,
Two, and Three and put them to use across the
organization in a rational, repeatable, and efficient
manner. More importantly, any signal that illuminates
a specific business issue — customer experience,
risk management, brand reputation — can be viewed
holistically by whoever needs it, when they need it
(the “right time”), and without regard to data source or
type.
While no companies we interviewed have yet reached
all criteria for Stage Four, some, such as Caesar’s
and Symantec (see case study below), are already
well on the way. Caesar’s is mapping the customer
experience with the goal of understanding desires and
preferences across its properties at all stages of their
journey, while Symantec has built out a strategic and
comprehensive governance model.
The following case study illustrates how Symantec is
addressing the challenges of meaningfully integrating
social data with enterprise business processes and
data. Note that, while many of the processes the
company is using are still manual, the organization
has established a clear vision for how it will solve
both granular issues (integrating social signals into
customer care) and more visionary ones, such as how
it will use social data to inform a deeper and more
strategic ongoing relationship with customers.

Case Study: Symantec — Using Social Data Across the Organization
The following case study illustrates how Symantec is addressing the challenges of meaningfully integrating
social data with enterprise business processes and data. Note that, while many of the processes the company
is using are still manual, the organization has established a clear vision for how it will solve both granular issues
(integrating social signals into customer care) and more visionary ones, such as how it will use social data to
inform a deeper and more strategic ongoing relationship with customers.
Ultimately, the company vision is to integrate social data intelligence deeply into the organization to empower all
20,000 employees around the globe to engage with its 2.5 million customers.
Objective
Founded in 1982, Symantec provides security, storage and systems management solutions to help customers
secure and manage their information and identities independent of device. The company uses social data to optimize
business value across the customer journey, as well as to drive revenue, improve efficiencies and mitigate risk.
Approach
Today Symantec uses Salesforce Marketing Cloud to harvest social data — including posts, brand mentions,
and comments — from across the web and sends it to a central team within the marketing organization
that determines the business function best equipped to serve the customer. The central team, known as the
Social Business Team, has established processes and workflows to route incoming queries and mentions to
approximately 300 trained employees based on which product or issue is mentioned.
Symantec has established specific tracks for specific products, but most notably classify what they call Actionable
Internet Mentions (AIMs) into seven buckets, falling into different business functions and corresponding to various
phases of the customer journey. The seven classifications are:
1. Case: Request for help resolving real-time issue
2. Query: Question that doesn’t require support resource
3. Rant: Insult that merits brand management consideration
4. Rave: Praise from Symantec brand advocate
5. Lead: Pronouncement of near-term purchase decision
6. RFE: Request to enhance a product with a new feature
7. Fraud: Communication from an unauthorized provider of Symantec products
These seven categories incorporate workflows for Symantec’s top 15 product lines and span business functions
that include marketing, customer support, engineering, PR, product management, and legal. Following are some of
the ways Symantec harvests social data:
• Customer Experience. Symantec aims to optimize customer experience at every touchpoint. For
example, if a customer mentions the name of a product in a social post, that case is automatically
assigned and routed to the appropriate support resource trained in social and on the specific product.
Symantec also leverages existing content to streamline the process for both customer and employee,
routing to support team members with a deep knowledge of product content who can answer the
question or direct the customer to an existing thread in Symantec’s online community.
16

17
For example, one of Symantec’s goals (and key metrics) is the conversion of “ranters” into “ravers.” The PR
team is trained to surface a potential product issue hidden in the “rant,” re-tags the case accordingly, and
conducts a “warm transfer”of the customer to the proper support staff to resolve the issue.13
“Raves” are also
used to improve experience and are routed when appropriate to product marketing to say thank you, inquire
about customer references, or invite the customer to become a blogger or forum advisor. For other customers,
submitting ideas for innovation (or “Requests For Enhancements” — RFEs), Symantec routes suggestions to its
product management team to help instruct development roadmap and priority.
• Lead Nurturing. Symantec also uses social data intelligence to generate and nurture leads, both for
consumers and businesses. If the customer is comparing with a competitor, questioning renewal, asking
for product specifications, or expressing frustration with a competitor’s product, listening tools enable
Symantec to route these insights into its lead pipeline and engage in the most appropriate manner based
on the customer’s comment.
• Risk Mitigation. Symantec has also discovered a way to mitigate risk when analyzing data to build
content for marketing. “The fraud protection value of social media monitoring came as a surprise to
me: As we ran our product monitoring queries, we were alarmed to find a number of posts promoting
and linking to illegal download sites,” explains Tristan Bishop, Symantec’s Director of Social Business. In
addition to training its legal department to handle these posts, Symantec now actively monitors for fraud
and helps preserve the integrity (and limit the potential for negative posts) around the product.
Results
Since rolling out this workflow, Symantec has resolved numerous support cases, converted many ranters into
ravers, generated hundreds of business and consumer leads, rapidly identified key areas to prioritize for product
development, and uncovered hundreds, if not thousands, of fraudulent product pilots.
In the longer term, Bishop hopes to standardize all customer data within the same CRM system to provide full
context for every employee. “When a Symantec employee interacts with a customer, we hope they’ll be able to
view Symantec’s entire relationship with that customer: The customer’s sales history, their support history, and
their social likes and shares of Symantec products and content. By giving our frontline staff this context, we can
empower them to create a superior customer experience.”

18
In Altimeter Group’s interviews with and surveys of
enterprise-class companies, we noted a number of
consistent themes that signal a level of maturity in
social data analysis that is a prerequisite for strategic
decision-making across the business. The following
six dimensions lay out the most critical factors that
you must evaluate, formalize, and — over time — scale
across your organization.
Use this model to gauge your organization’s social
analytics maturity level within each dimension and
identify the steps you must take to reach the next
stage (refer to Figure 5). It’s important to understand
that while you may be at one stage for one
dimension (such as Strategy), you may be at another
stage within a different dimension (such as Process).
The goal is to review all dimensions, locate your
current level, and note the steps needed to reach the
next stage.
Below is an overview of these six dimensions of
analytics maturity.
1. Scope
Definition: The number of internal groups that work
with social data and the scope of data to be measured:
which platforms, which data points, and why.
Defining scope early on — what you will do initially,
what you won’t do, and why — is your first priority.
Why? Because the process of defining scope will
become a forcing function to align multiple, disparate
stakeholders; will establish analytics as a strategic
discipline rather than a tactical “help desk”; and will
help mitigate the “report monkey” phenomenon, in
which analysts are forced to react blindly to a constant
stream of ad-hoc requests.15
While initially you may only be able to address a small
portion of what you (and your stakeholders) want
to measure, take the time to lay the groundwork for
a consistent, methodological approach across the
organization. The following are the key factors of
scope:
• Inventory: A clear understanding of all
departments actively engaged in social media
(market research, HR, ecommerce), whether and
how they are using social data, what metrics
they are tracking and reporting on, and any
inconsistencies across departments.14
• Documented Methodology: Documentation of
the confidence levels you are able to deliver
on various social platforms. For example, if
sentiment analysis is important to your business,
you must be transparent about the accuracy level
you are able to achieve with your tool set.16
Set
expectations that trial and error will be the norm,
as you will be accorded a very short learning
curve before you’re expected to deliver social data
insights.
• Documented Success Criteria: Document
success criteria for delivery of metrics. This
includes clear formulae, consistent use of
tagging, and access to web analytics dashboards
to track the impact of social content.
You will know you’ve mastered “Scope” when you can
easily answer questions such as: What social data do
we have at our disposal? What do we track? What is
our methodology for social data? What are the critical
success factors to scale this across the organization?
2. Strategy
Definition: The extent to which social data — and
metrics — is in alignment with strategic business
objectives across the organization.
Most organizations begin measuring social media
focus first using volume metrics: how many likes,
fans, followers, brand mentions, etc. But counting
can only take you so far. If you cannot demonstrate
a connection to the outcomes that executives in
Six Dimensions of Analytics Maturity

19
the C-suite care about — brand reputation, revenue
generation, operational savings, customer satisfaction,
and so on — you’re at risk of creating a silo (worse, a
cost center) that you are unable to justify.
In some organizations, business strategy may involve
multiple dimensions of business value; in others,
it comes down to one, simple, over-arching goal.
Says Ryan Smith, Digital Marketing Manager, Royal
Canin, “The really important thing for our vision of
the business is to improve the understanding of our
brand and what we stand for.” In Royal Canin’s case,
this consists of using social media to drive brand
awareness and preference in concert with other
marketing activities.
The prioritization of metrics will evolve over time to
support the evolution of business strategy. In the early
stages, you may start with tying social media metrics
effectively to only one specific goal, such as marketing
optimization or brand awareness. But you will be able
to chart your organization’s maturity by the extent to
which every social media initiative throughout your
organization — no matter how small or short-term — has
a clear set of goals and measures that illustrate success.
3. Context
Definition: The extent to which the organization is able
to view social data in various contexts to understand
what is typical, what is unusual, and the drivers for
each.
When you start to measure social data, you will have
no historical record on how social assets or campaigns
performed in the past. And while you may have
performance benchmarks for non-social campaigns,
you’ll need to understand the actual performance
of your social programs and campaigns to evaluate
impact.
Typically, you will need a healthy sample of data
(minimum three to six months of completed
campaigns) before you can get a feel for what is
“normal” and begin to craft defensible KPIs. But you
can start measuring and learning about the drivers
of results on day one and use this insight to set
expectations internally. In addition:
• Understand how social data changes in the
context of time. For example, net sentiment may
be an inexact metric at a single point in time, but
significant fluctuations are worth investigating
and worth digging deeper to better understand
the source of the change.
• Multiple outliers gain significance. A single
exception at a point in time has little meaning,
but over time, may signal a significant risk or
opportunity.
• Look at existing metrics — brand lift, aided and
unaided awareness, Net Promoter Score, or other
customer satisfaction metrics — to set a point
of comparison for what you may find in social
data. For example, if you introduce a campaign to
drive awareness of a new brand or service, does
social data confirm or refute your existing brand
lift metrics? In Parasole’s case, the social data not
only confirmed customer satisfaction feedback
and verbal complaints, it demonstrated that the
problem was much more widespread than the
company would otherwise have realized.
While it’s important to be aware of what your
competitors are doing, don’t over-focus on competitive
analysis or searching for industry benchmarks until
you have a good feel for your data and for how your
customers behave in social channels. Says Jen Evans,
CEO, Squeeze CMM Atom Agency, “The digital
behavior of every audience is completely different;
you can’t rely on general research to understand your
customers.”
The markers of maturity for the Context dimension
are simple: the existence of clear benchmarks against
past history, enterprise data, and, in some cases, the
competition.

4. Governance
Definition: The extent to which the organization has
developed, socialized, and formalized processes
related to workflow, collaboration, and data sharing.
In the early stages, governance processes tend to be
ad-hoc; they become formalized and integrated into
business process as the use of social data matures.
When you begin to measure social media, you will
run into numerous situations for which you have no
existing processes or policies. While you will not yet
have established governance for new workflows, you
can begin to note exceptions that require attention
down the line. Following are key areas of focus.
• Data Sharing: Social media creates an
unprecedented need to share data across the
organization. But not every organization is
uncritically enthusiastic about sharing data
among teams, whether for process, resource,
or cultural reasons. Sharing data can have
unintended consequences: exposing problems to
peers, management, and employees before there
is a common understanding of the problem or
a proposed solution, which can create tensions
and pressure among and between groups. This
can be especially acute in more “command-and-
control” organizations, where information sharing
is less the rule than the exception.
While it’s critical to respect your organization’s
processes and cultural norms, this is an
opportunity to begin preparing your organization
to take best advantage of social data (and
eventually other types of big data) that do not
respect departmental boundaries or cultural
conventions. To do this, you must explicitly
demonstrate the value of sharing data, and the
risks inherent in suppressing it.
Tristan Bishop of Symantec experienced this
first-hand. “Symantec helps consumers and
organizations secure and manage their information-
driven world. Because we focus on eliminating
risks to information, technology, and processes,
we had to internally clarify the “risk versus reward”
ratio of social engagement before launching the
[social analytics] program.” For Symantec, the risk
of organizational tension was outweighed by the
customer experience, risk reduction, and other gains
they achieved by sharing social data.
• Executive Support: Bringing the right executive
sponsors and champions to bear is critical to
your success. To do this, you need executives
who are comfortable with the challenges of
social data and are willing to publicly support
the degree of research and investment needed
to pursue unexpected findings and work
across departments. Says Jen Evans, “To run
an effective social analytics pilot, you need a
multidisciplinary focus with significant tolerance
for changing direction as the findings warrant.
You need an executive champion, clear goals and
objectives, and clear expected outcomes. It takes
a special type of executive to do this.”
The ideal executive for this role is someone who
has a vested interest in customer, operational,
or financial insights; that is, a CMO, COO, CFO, or
head of strategy. He or she needs to champion
the integration of social data in the belief that it
will yield greater insight than separate data sets
can on their own. But to support this champion,
analytics teams must do their homework: They
must establish a clear and credible discipline for
social data within the organization to provide the
executive with the justification needed to push for
alignment, change, and, ultimately, budget.
As the organization matures, this champion’s role
will evolve from an evangelical to a facilitative role
as multiple organization heads now understand
the value of a holistic view of data.
Organizations will reach Governance maturity
when their social data measurement processes
are documented, socialized, and understood
throughout the organization and when workflows
are clear, automated, and scalable.
20

5. Metrics
Definition: The extent to which metrics have been
defined and socialized throughout the business.
The key to metrics maturity is definition and
prioritization, as discussed in depth in the “Making
Social Data Actionable” section. In the early stages,
organizations focus on a small set of metrics, define
them, and socialize them throughout the organization.
As they gain context and confidence, they establish
clear benchmarks, which set the stage for defining
KPIs and establishing future metrics.
A marker of maturity for metrics is the ability to
articulate not only the ingredients and data sources,
but also the process and criteria by which a metric
is established and evaluated. For example, ad value
equivalency may be a practical metric early on as the
organization becomes accustomed to social media
and its impact. But over time, you should expect to
see metrics emerge that tell you more about the
impact of social media on the business.
As metrics mature fully, it will become clear which
ones drive insight and decision-making and which
should be used as KPIs to measure performance.
For example, impressions are generally too volatile
a metric in social media to establish as a KPI, while
ROI on a marketing campaign will inform not only
future decisions but performance evaluation as well.
A mature analytics organization clearly understands
and can justify the value of these metrics across the
organization.
6. Data
Definition: A strategic approach to the data and
platforms at your disposal.
There are three main areas of focus to ensure a
strategic approach to social data:
• Know Thy Social Data. Note the difference
between a social action (a share, like, pin, retweet)
versus social text (a tweet, a post, a comment),
because each carries different signals. A person
who “likes” a Facebook page will see that brand
in his feed (unless he hides it), while a comment
may simply become a one-to-one transaction.
“This data is only going to get more complicated,”
says Matthew Knell, Director, Social Media, AOL.
“We need to understand where content is coming
from: a fan page, a comment, a share button, etc.”
• Know Thy Platforms. Make sure someone in your
organization is spending the necessary time to
stay on top of social media platforms and their
metrics. You’ll need to know what they are, what
they do, what samples you have access to, and
how much historical data you are able to acquire.
This is critical to understanding how to view
the data and what conclusions to draw from it.
Stay abreast of API and Terms of Service (TOS)
changes that may affect your data samples and
results. Tools such as Brandle, which provides
online presence management, can ensure that
you have an inventory of all points of presence
that generate social data.
• Warehouse Thy Social Data. Qualcomm has
built a social data warehouse to provide a
shared service to inform decision-making, while
Caesar’s is integrating social data with Customer
Relationship Management and its enterprise data
warehouse.
Your organization’s approach to social data will evolve,
especially as it becomes integrated with enterprise
and other data streams. Over time, social data may
be folded into an organizational “big data” initiative,
including transaction data, sensor data, and other
data sets. Keep in mind that the work you do now
to establish the integrity of social data intelligence
will influence the extent to which it is viewed as
credible and integral to the business later on. So your
investments now will pay off down the line, even if the
organization of the future looks markedly different from
today.
21

Conclusion
22
Although technical integration of many disparate data
streams is on the horizon, today’s opportunity is to see
social data as a lens for insight and a crucial input into
business decisions and strategy. This will be a journey
in itself, but a worthwhile one, as organizations strive to
make data actionable and even predictive. It will create
expected and unexpected outcomes. We are already
seeing the following patterns emerge:
• A view from the customer in, not the organization
out. One of the opportunities of social data
intelligence is to better understand how
consumers, citizens, patients, volunteers, and
business customers think and feel at different
stages of their journey with us. Right now, what
information organizations have is stored in CRM,
business intelligence, ERP, and supply-chain
systems, social media dashboards, search,
apps, and devices, in a format that is (arguably)
useful for individuals in the organization but too
fragmented to provide a holistic view from a
customer perspective. Chris Kahle of Caesar’s
is already working toward this future. “We are
trying to understand the customer’s entire journey
and all the touchpoints so we can optimize the
experience.”
• Social data is the dress rehearsal for “Big Data.”
The volumes, variety, and velocity of social data,
combined with innate interpretive challenges,
multiple stakeholders, and utility across the
business mean that organizations that embrace
social data and seek to understand it are already
addressing issues they will inevitably face as other
real-time, heterogeneous, and unprecedented
data streams become available, integrated, and
scalable.
• Big data drives organizational change. One of
the most interesting aspects of a data-driven
organization is the impact to decision-making,
especially at the executive level. An October 2012
article in the Harvard Business Review discusses
the “HiPPO” phenomenon, meaning that the
traditional mode of decision-making in many
organizations is based upon the “HiPPO,” or
“highest paid person’s opinion”16
.
• As organizations gain access to better quality,
better filtered, curated, and interpreted data,
clashes between the intuitive or experience-
based and the data-driven decision makers are
inevitable. The challenge will be to anticipate and
transparently plan for the cultural implications of
better, faster, and more transparent data, while
preserving the human factors that cannot be
quantified.
• Making the real-time enterprise real. Location
data, social media, multi-channel experiences,
sensors, and transactional data are all realities
today. Propensity modeling demonstrates the
statistical likelihood of event “A” versus event
“B,” given a specific set of conditions. With the
advent of new technologies, such as SAP’s HANA
platform, real-time insights become possible —
SAP and Adobe are already working on an early
integration of HANA and Adobe Marketing Cloud.
It’s real, and it’s happening now.
Social or not, big data raises customer expectations
and competitive pressure. It breaks silos. Admittedly,
we are at the very early stages of social, enterprise,
and big data integration, but organizations such as
Caesar’s, Symantec, and others are already thinking
deeply about these issues and preparing their
organizations for the changes to come, taking social
data and turning it into organizational intelligence.

23
Interviews with 11 brands that are currently
integrating social with enterprise data.
Interviews with 23 agencies and vendors of social
media or social analytics technology.
Survey of 27 end users of technology on how they
are integrating social data in the context of other
enterprise data sets. Note that, because of the
small sample size, these findings should be viewed
as anecdotal rather than quantitative.
Ecosystem Input
This report includes input from market influencers,
vendors, and end users who were interviewed or
briefed by Altimeter Group during the course of this
research. Input into this document does not represent a
complete endorsement of the report by the individuals
or companies listed below.
Brands (11)
AOL, Matthew Knell, Director, Social Media
Caesar’s, Chris Kahle, Web Analytics Manager
Caterpillar, Kevin Espinosa, Social Media Manager
e-Toro, Adi Yagil, Social Media Strategist
Interscope Records, Lee Hammond, VP Digital
Parasole, Kip Clayton, VP Marketing and Business
Development
Parasole, Sarah Nearison, Marketing Manager
Parasole, Randy Stanley, Digital VP
Qualcomm, Michael Whelan, Senior Manager, Marketing
REI, Paolo Mottola, Digital Engagement Program Manager
Royal Canin, Nigel Blackwell, IT Manager,
Royal Canin, Ryan Smith, Digital Marketing Manager
Symantec, Tristan Bishop, Director of Social Business
Third Door Media (Search Engine Land; Marketing Land),
Monica Wright, Director of Community
Vendors Agencies (23)
Adobe, Lawrence Mak, Product Marketing
Adobe, Jordan Enright-Schultz, Product Manager
Adobe, RJ Simonian, Sr Product Manager for Analytics
Adobe, Carmen Sutter, Product Manager, Listening
Ai-One, Olin Hyde, VP of Business Development
Anametrix, Pelin Thorogood, CEO
Awareness, Sal Gilberto, Marketing Program Manager
Brandle, Chip Roberson, CEO
Converseon, Constantin Basturea, VP, Strategy
Coveo, Dianne Berry, VP Marketing
Coveo, Jessica Hohn-Cabana, Senior Director,
Corporate Marketing
Crimson Hexagon, Wayne St. Amand, VP of Global
Marketing
Digital Marketing Works, Aaron Zwas, Director, Emerging
Technologies
Infegy, Justin Graves, CEO
Janrain, Bill Piwonka, VP, Marketing
LittleBird, Marshall Kirkpatrick, CEO
MixPanel, Suhail Doshi, CEO
NetBase, Lisa Joy Rosner, CMO
New Brand Analytics, Cindy (Cynthia) Kazan, Strategic
Marketing Public Relations
Revinate, Michelle Wohl, VP Marketing
Salesforce, Gordon Evans, VP Marketing, Salesforce
Marketing Cloud
SAP, Amit Sinha, VP Database and Technology
Marketing
Squeeze CMM Atom Agency, Jen Evans, CEO
Simply Measured, Adam Schoenfeld, CEO
Methodology

24
Spiceworks, Adam Schaeffer, Manager, Public
Analyst Relations
Spiceworks, Adam Weinroth, Executive Director of
Vendor Marketing
Tellagence, Sarah Buchanan, Director of Marketing
Tellagence, Matt Hixson, CEO and Co-
FounderTellagence
uberVU, Mark Pascarella, CEO
Acknowledgements
With deepest thanks for insight and/or support from:
David Clark, Megan Costello, Alistair Croll, Sebastian
Hempstead, Margaret Francis, Rob Johnson, Israel
Mirsky, Chris Moody, Filippo Passerini, Blake Robinson,
Jud Valeski, Asha Hossain, Cheryl Knight, Vladmir
Mirkovic, Alec Wagner, Charlene Li, Rebecca Lieb, Chris
Silva, Brian Solis, Alan Webber, Christine Tran, Jaimy
Szymansky, Jon Cifuentes, Julie Viola, and Susan Wu.

25
1
Source: Jeremiah Owyang, Social Business Readiness.
http://www.altimetergroup.com/research/reports/
social-business-readiness
2
Based on a quantitative survey of 27 organizations
with revenues in excess of $100M who currently use
social data in the context of enterprise data. The small
sample size for some of the answers (disclosed on the
individual charts) reflects the nascency of this trend.
As a result, Altimeter recommends viewing these
findings as suggestive rather than indicative. Altimeter
will repeat this survey over time to track the evolution
of social data integration in the enterprise.
3
Jeremiah Owyang, Social Business Readiness, op. cit.
4
Source: Speech at Brandworks University, May 21,
2013.
5
Source: Gartner Group: http://www.gartner.com/it-
glossary/big-data/
6
IBM has said that 90% of the data in the world was
created in the past two years alone (source: http://
www-01.ibm.com/software/data/bigdata).
7
Source: Twitter.
8
An ingredient brand is one in which the product
is used in other products, but generally not as a
standalone entity. For example, Intel’s chipsets are an
ingredient brand.
9
For recommendations on how to map social data with
business outcomes and measure revenue impact.
See A Framework for Social Analytics and The Social
Media ROI Cookbook: http://www.altimetergroup.com/
research/reports/the-social-media-roi-cookbook.
10
The Altimeter Report The Social Media ROI Cookbook
is available at http://www.altimetergroup.com/
research/reports/the-social-media-roi-cookbook.
The six ways to map social activities to revenue
generation are brand health, marketing optimization,
revenue generation, operational efficiency, customer
experience, and innovation.
11
Simply Measured has a good analysis of the pitfalls
of correlating Facebook PTAT with engagement. For
more information, see: http://simplymeasured.com/
blog/2013/05/16/why-ptat-doesnt-correlate-with-
engagement-on-facebook
12
The reason HANA, and “real-time” analytics is so
disruptive is that it speeds up the rate at which data
becomes available and therefore the organizational
requirement to act on it. Organizations that do
not have scalable processes, standards, and
governance in place to act on weekly or monthly
reports, or criterial for what data should be acted on
at what time, will be even more challenged when the
pace of data increases to days, hours, minutes, and
seconds.
13
A “warm transfer” is the customer service standard
for transferring one person to another within a call
center. The key is to ensure that the second party
is available and all parties are fully prepared and
aware of the reason for the call before the transfer
is complete. For a more detailed definition of “warm
transfer,” see https://www.8x8.com/Resources/
Learn/HintsandTips/CallTransferTips.aspx.
14
“Report Monkey” is a disparaging term for
someone whose job it is to respond continuously
to ad-hoc report requests, rather than perform a
more strategic analytical role. The lack of a clear
analytical foundation for social data can exacerbate
this situation, leading to employee turnover, missed
opportunities, and risk — especially if there is a lack
of analytical standards.
15
Altimeter defines this as having at least one person
(not necessarily a full-time equivalent, or FTE) who
is tasked with listening and/or engagement within
the department.
16
Most sentiment analysis tools deliver 60%-65%
accuracy in English out of the box; machine learning
and more sophisticated text analytics can raise the
accuracy level to 80% or higher. Some tools, such as
NetBase, Tracx, and Adobe, go further by enabling
analysts to categorize emotion type and intensity.
But this will never be perfect; even humans disagree
in their interpretation of sentiment. Choose the best
tools you can afford, and be transparent about their
accuracy.
176
“Big Data: The Management Revolution,” Andrew
McAfee and Erik Brynjolfsson, Harvard Business
Review, October 2012.
End Notes

Susan Etlinger, Industry Analyst
Susan Etlinger (@setlinger) is an industry analyst at Altimeter Group, where she works with global companies
to develop both social media listening and measurement strategies that support their business objectives.
Previously, Susan was a Senior Vice President at Horn Group. Susan is a published translator and has a
bachelor’s degree in rhetoric from the University of California at Berkeley.
About Us
Altimeter Group provides research and advisory for companies challenged by business disruptions, enabling them to pursue new
opportunities and business models. We share our independent research on business disruptions via research reports, webinars, speeches,
and more. We also offer advisory services to business leaders who wish to explore the specific implications of these disruptions within their
organizations.
Contact us: 650.212.2272 • sales@altimetergroup.com • altimetergroup.com • @altimetergroup
Authors
Jessica Groopman, Senior Researcher
Jessica Groopman (@jessgroopman) is a Senior Researcher with Altimeter Group where she researches the
evolving challenges and opportunities of disruptive technologies on businesses, primarily focusing on how
data is shifting business models and behaviors. She has experience conducting business, technological, and
anthropological research.
Charlene Li, Founder and Analyst
Charlene Li (@charleneli) is Founder of Altimeter Group and author of the New York Times bestseller, Open
Leadership. She is also the co-author of the critically acclaimed bestselling book, Groundswell, which was named
one of the best business books in 2008. She is one of the foremost experts on social media and technologies and
a consultant and independent thought leader on leadership, strategies, social technology, interactive media, and
marketing.
Open Research
This independent research report was 100% funded by Altimeter Group. This report is published under the principle of Open Research and is
intended to advance the industry at no cost. This report is intended for you to read, utilize, and share with others; if you do so, please provide
attribution to Altimeter Group.
Permissions
The Creative Commons License is Attribution-Noncommercial-Share Alike 3.0 United States at http://creativecommons.org/licenses/by-nc-sa/3.0.
Disclosures
Your trust is important to us, and as such, we believe in being open and transparent about our financial relationships. With permission, we
publish a list of our client base on our website. See our website to learn more: http://www.altimetergroup.com/disclosure.
Disclaimer
ALTHOUGH THE INFORMATION AND DATA USED IN THIS REPORT HAVE BEEN PRODUCED AND PROCESSED FROM SOURCES BELIEVED
TO BE RELIABLE, NO WARRANTY EXPRESSED OR IMPLIED IS MADE REGARDING THE COMPLETENESS, ACCURACY, ADEQUACY, OR
USE OF THE INFORMATION. THE AUTHORS AND CONTRIBUTORS OF THE INFORMATION AND DATA SHALL HAVE NO LIABILITY FOR
ERRORS OR OMISSIONS CONTAINED HEREIN OR FOR INTERPRETATIONS THEREOF. REFERENCE HEREIN TO ANY SPECIFIC PRODUCT OR
VENDOR BY TRADE NAME, TRADEMARK, OR OTHERWISE DOES NOT CONSTITUTE OR IMPLY ITS ENDORSEMENT, RECOMMENDATION, OR
FAVORING BY THE AUTHORS OR CONTRIBUTORS AND SHALL NOT BE USED FOR ADVERTISING OR PRODUCT ENDORSEMENT PURPOSES.
THE OPINIONS EXPRESSED HEREIN ARE SUBJECT TO CHANGE WITHOUT NOTICE.

Social Data Intelligence: Integrating Social and Enterprise Data for Competitive Advantage

More Related Content

What's hot

Viewers also liked

Similar to Social Data Intelligence: Integrating Social and Enterprise Data for Competitive Advantage

Recently uploaded

Social Data Intelligence: Integrating Social and Enterprise Data for Competitive Advantage