The big story behind your big data: Six Practices for Making an Impact with Text Analytics

The Big Story Behind
Your Big Data:
Six Practices for Making
an Impact with Text Analytics
By Beth Benjamin and Joachim B. Lyon

MEDALLIA.COM 2
Susan, an expectant mother near her due date,
walks into BABIES“R”US®, a leading retailer of
baby products. A sales associate approaches her
as she looks around and asks if she’s joined the
online registry, which is offering a promotion that
includes a 10 percent back eGift card on registry
purchases. Anticipating the several thousand dol-
lars she expects to spend on baby gear, Susan
eagerly signs up. She’s told she’ll receive her
eGift card approximately ten weeks after her
baby is born.1
But the eGift card never arrives. She calls the
customer contact center and an agent opens a
ticket, promising someone will get back to her.
A week passes. Nothing happens. She calls again.
This time the agent tells her that, regrettably, her
registry was started three days before the pro-
motion went into effect. Frustrated, Susan asks to
speak to a manager, only to be told that customer
service managers “can’t fix this—only someone
in the gift card department can.” So, she asks to
speak to someone in the gift card department. The
agent replies: “That’s not possible.”
Susan posts her exasperated version of events on
Facebook and tags BABIES“R”US® as well as one
of its competitors, buybuy BABY®. Within an hour,
an online agent from BABIES“R”US® calls Susan.
The issue is resolved, and Susan receives her
eGift card.
More Touchpoints, More
Channels, More Complexity
What can we learn from this customer experience
snafu? One lesson is the increasing complexity
that companies face in their efforts to provide
a smooth and consistent customer experience
across a growing number of touchpoints, channels,
and technologies. Because these touchpoints are
often managed by different business units and
influenced by different functions, creating a seam-
less experience can be challenging. However,
while the challenges are significant, so too are the
rewards. Estimates suggest that customers who
use multiple channels spend two to four times
more annually than single-channel customers.2
Another lesson is that customers today commu-
nicate an enormous amount of feedback digitally.
Eighty-five percent of Millennials who responded
to a recent Medallia survey said they write online
reviews or post on social media after they’ve had
a great or terrible customer experience.3
Some
digital feedback is solicited directly from customers
through online surveys, but much of it is unsolicited
and easily shared using review sites, blogs, and
social media platforms—what Wharton professor
Eric Clemmons calls “word of mouse.”4
Intro........................................................................................................................................................................................
Customer Experience for Competitive Advantage: The New Frontier......................................................
The Organizational Challenges of Text Analytics...............................................................................................
Six Practices to Maximize the Value of Text Analytics......................................................................................
What’s Next for Text Analytics?..................................................................................................................................
2
7
10
12
18
TABLE OF CONTENTS

MEDALLIA.COM 3
Technology for the
Unstructured World
Most consumer feedback shared openly online or
solicited through company surveys is in the form
of free-flow text, often referred to as “unstructured
data.” Unstructured data is different from other
types of data because it cannot be organized into
numeric fields or analyzed using traditional
business intelligence software. While the exact
amount of unstructured data is unknown, it is unde-
niably huge.5
IDC estimates that unstructured data
is growing at a rate of 62 percent per year and
predicts that by 2020, 93 percent of all data in the
digital universe will be unstructured.(See Figure 1)6
To be sure, unstructured data provides a gold mine
of information for companies to use in their efforts
to win the battle for customers. Customer commen-
tary, shared as open-ended comments in surveys,
online review sites, blogs, chat logs, social media,
and a host of other sources, can produce unique
insight into the customer experience and new ideas
for product and marketing innovation. However,
because unstructured data is text heavy and irregu-
lar, making sense of what is being said and how it’s
being said—positively or negatively—is not for the
faint of heart.
Companies that underestimate the value of unstruc-
tured data do so at their peril. While extracting
meaning from text and other unstructured sources is
complex,companiesarequicklyrealizingthevalueof
turning digital customer chatter into actionable busi-
ness intelligence. Powerful new technologies have
evolved to analyze large volumes of text that were
previously unprocessable by computers. As a
result, companies are now able to transform masses
of unstructured information—documents, audio,
video, and images—into quantifiable data that
can be processed and analyzed by business
decision makers.
Text analytics is one of these powerful
new technologies.
20%
80%
7%
93%
Unstructured Data
Structured Data
2020
Source: Merrill Linch, 1998: Headwaters Group 2015
1998
Figure 1: Growth of Unstructured Data

MEDALLIA.COM 4
Using Text Analytics to Quantify
Customer Buzz
Using statistical, linguistic, machine learning, and
visualization techniques, text analytics software
automates the collection and interpretation of cus-
tomer commentary—verbatim words and phrases
that collectively represent what customers are say-
ing about their experience in the marketplace. Text
analytics enables companies to more accurately
interpret what their customers are saying, their wants
and needs, how well those needs are being met,
and what new products, features, or services might
appeal to them. Without text analytics, extracting
meaning from customer comments, social reviews,
and other digital dialogue requires manually sifting
and coding thousands and thousands of records, a
process that is not only time consuming and costly,
but also prone to human error.
Text analytics can be applied to sources across the
entire value chain (See Figure 2). As companies
gain familiarity with the technology’s capabili-
ties, many have started to apply text analytics to
some of their biggest operational challenges.
BUSINESS
GOALS
BRAND
HEALTH
INNOVATION
COMPETITIVE
INTELEGENCE
REVENUE
GENERATON
OPERATIONAL
EFFICIENCY
CUSTOMER
EXPERIENCE
Source: Medallia 2016
Figure 2: Opportunities for Driving Value with Text Analytics

MEDALLIA.COM 5
An independent study of 220 text analytics users,
conducted by IT consultancy Alta Plana, finds that
text analytics is being applied to more than 16 major
business needs (See Figure 3).7
Topping the list of business applications are cus-
tomer experience programs trying to create a
seamless shopping experience and better under-
standing of their customers’ needs and preferences.
With growing pressure to demonstrate more imme-
diate impact, customer experience professionals
have come to realize that it’s not enough to simply
know more from their data. They have to opera-
tionalize the knowledge they gain so that their
insights can be transformed into actions that drive
business results.
Figure 3: Business Applications of Text Analytics
Percent respondents citing application
Source: Alta Plana, 2014
Voice of customer /
Customer experience
Research
Brand/product /
Reputation management
Competitive Intelligence
Search, information access,
or question answering
Customer CRM
Content management/publishing
Online commerce including
Shopping, price.
Life sciences or
Clinical medicine
E-discovery
Insurance, risk management,
or fraud
Others
2011
2014
2009
39%
38%
38%
33%
29%
27%
25%,
16%
15%
14%
13%
11%
0% 10% 20% 30% 40%

MEDALLIA.COM 6
Text analytics makes more data actionable. It gives
business leaders the ability to extract customer
insights from data sources that were previously
too costly to analyze. These insights can be used
to drive higher satisfaction ratings, greater
customer retention, increased sales, opera-
tional improvements, new products and services,
and other performance outcomes (See Figure 4).
Given the potential for significant business impact,
customer experience management may be the first
frontier where text analytics strikes gold.
Take the case of a global financial services
company. When the company discovered that
customers were dissatisfied with the quality of
their service calls, it initially thought it would have
to overhaul its infrastructure and hire new agents,
a multimillion-dollar investment. But, by using text
analytics to dig deeper into thousands of customer
surveys and open-ended comments, the company
discovered that the quality problems were more
localized than it originally thought and not always
attributable to its agents. With this insight,
company leaders decided that it would be more
cost effective to revise training programs for
certain call centers and revamp the company’s risk
assessment algorithms. By analyzing
customers’ open-ended comments, the company
not only improved service-call quality, it also saved
millions of dollars in the process.
Figure 4: Impact of Text Analytics
0% 10% 20% 30% 40% 50% 60% 70%
Reduction in required staff/higher staff
Improved new-customer acquisition
Higher customer retention and loyalty
Ability to create new information products
Increased sales to existing customers
Higher Satisfaction ratings
Higher search ranking, web traffic
Fewer issues reported and/or service complaints
Lower average cost of sales, new & existing
More accurate processing
Faster processing of claims/requests/casework
Achieved
Measure
Plan to Measure
Source: Alta Plana, 2014
19% 33%16%
21% 33%12%
15% 31%19%
17% 37%8%
17% 34%9%
13% 29%12%
14% 29%12%
12% 29%9%
14% 30%6%
11% 28%7%
11% 24%10%

MEDALLIA.COM 7
Customer Experience for
Competitive Advantage:
The New Frontier
Consumer decision making has changed dramati-
cally over the past 15 years. As online resources
and access to the Internet have increased, con-
sumers are using social review sites, blogs, social
media, and other online outlets to share their
experiences and learn from other consumers.
And they appear to trust online resources: A 2013
Nielsen survey found that nearly 70 percent of
global consumers trust consumer opinions posted
online, making online reviews one of the most
trusted sources of brand information, second only
to recommendations from friends and family.8 As
customers are able to evaluate a broader range
of options and make better purchasing decisions,
their expectations are growing and changing more
rapidly than ever.9
In the face of these changes, companies leading
the way in customer experience management are
building capabilities that go well beyond customer
service. They are carefully tracking each custom-
er’s experience, understanding the reasons for
that experience, and then rapidly responding to
what customers say and want. Combining strategic
needs with operational capabilities, customer expe-
rience leaders are using customer feedback and
new technologies like text analytics to fuel learn-
ing, motivate change, and develop new sources of
competitive advantage.
In the past, relatively few companies tried to differ-
entiate on the basis of customer experience. Icons
like Apple, Nordstrom, and Zappos were excep-
tions rather than the rule. That has changed.
A 2014 Conference Board survey of global business
leaders, found that corporate executives worldwide
ranked “customer relationships” as one of their
top five “critical challenges” and “sharpening their
Text Analytics in Action
Improving customer interactions. Text
analytics can be applied to open-ended
comments in customer surveys to identify
opportunities for improving the service ex-
perience. In addition, using text analytics
to capture customer feedback from review
or social media websites helps companies
listen to customers more fully, engage
with them, and drive service optimization.13
For example, using text analytics, one
major retailer discovered that customers
were consistently complaining about messy
dressing rooms, something the retailer had
failed to ask in its regular surveys. The re-
tailer addressed the problem, and custom-
ers became more satisfied with their shop-
ping experience.
understanding of customer needs” as their number
one strategy for meeting the challenge.10
Other studies confirm this new customer experi-
ence imperative. In a 2015 Gartner technology and
strategy survey, for example, 89 percent of compa-
nies said that customer experience would be their
“primary basis” for competition in 2016, up from 36
percent in 2011. Similarly, Gartner predicts that by
2017, 50 percent of product investment projects will
be redirected to customer experience innovations.11
A Walker Research study of large multinational B2B
companies finds that customer experience will
overtake price and product as the key brand dif-
ferentiator by 2020.12
To remain competitive, companies have little choice
but to embrace the game-changing technologies
that can help them gain a deeper understanding of
their customers. Text analytics applied to customer
feedback will not replace traditional surveys, but it
can yield more information than traditional surveys

8MEDALLIA.COM
other feedback loops. According to Forrester, com-
panies that get the most from text analytics are
those that take advantage of its always-on nature.
Analysts can look for themes and patterns when-
ever issues arise, unconstrained by a predefined
survey structure.15
The customer voice dominates
Traditional approaches to collecting customer
feedback make assumptions about what’s
important to customers and what’s not, like the
and a more refined understanding of that informa-
tion. More importantly, it opens the door to radical
new discoveries that would never have been pos-
sible using numeric ratings alone.
Text analytics applied to unstructured customer
feedback complements traditional feedback analy-
sis in a number of ways.
Ultimate flexibility
While surveys are administered at discrete times,
many sources of unstructured feedback flow into
companies continuously in the form of online
reviews, blogs, emails, social media posts, and
81%Unstructured
Data
Developing omnichannel solutions.
Ensuring a consistent multi-channel experi-
ence has become a leading priority for
companies. Text analytics makes it possible
to integrate and analyze data collected
across the full set of customer touchpoints,
including online surveys and reviews, con-
tact centers, website transactions, and in-
store interactions. A Wharton study found
that customers using the most channels—
on average 2.5—tend to be the wealthiest
shoppers across segments and spend the
most. These multi-channel shoppers are
also the most engaged with social media
and the least likely to return to the same
retailer for their next purchase.14
Using text
analytics, companies can look for themes
within and across channels to develop solu-
tions that may provide a more seamless
and attractive experience for these chal-
lenging shoppers, perhaps increasing their
loyalty and repeat purchasing behavior.
“courtesy of the service agent” or “the time it took
to resolve the issue.” While these certainly may be
important, they may not be what customers really
care about. Open-ended questions, social media
posts, and the other text-based responses let
customers talk about anything that is important to
them, in words that express their preferences and
emotions.

MEDALLIA.COM 9
Numeric ratings tell only part of
the story
Numeric ratings indicate what customers may
have experienced and how much it influenced
them, but they don’t tell us why. Text analytics
uncovers details customer survey ratings can’t
explain. Importantly, customers’ comments can
differ substantially from their ratings in terms of
sentiment, weight of importance, or content. A
Cornell study of hotel reviews found that nega-
tive comments influence a guest’s numeric rating
more than positive comments do.16 In other words,
when guests take the time to comment on a poor
experience, it tends to have a greater impact on
their rating than when they comment on a positive
experience. This uneven weighting indicates that
a simple average of positive and negative numeric
ratings may not be an accurate representation of a
customer’s actual experience or opinion.
Opportunities for innovation emerge
Companies may find unexpected ideas for new
product or service offerings through analysis of
unstructured data. When clothing retailer Tommy
Bahama examined open-ended comments from
customer surveys, executives saw a recurring
theme. When asked what the retailer could do to
make their experience better, customers repeat-
edly mentioned “complimentary margaritas.”
Seeing a way to reinforce the brand and make
their customers’ experiences more memorable, the
executives decided to test the concept in some of
their larger stand-alone stores. If they had relied
solely on customer survey ratings, they would
have missed this novel opportunity to provide
new value.
Generating market insights. Text analytics
provides a portal into the broader market.
When customers have the freedom to
discuss any aspect of their experience—not
just the service attributes listed in a sur-
vey—companies can discover entirely new
information they never would have looked
for otherwise. A restaurant chain using text
analytics to understand local food crazes
can aggregate data across markets to
identify broader trends, such as those that
occurred around artisan fries and spicy
sauces.17 Clearly, when companies are able
to identify emerging trends and preferenc-
es faster than their competitors, they can
gain significant market advantages.
Empathy at scale
When companies deal with millions of consumers,
there is a temptation to stereotype or depersonal-
ize customers. To simplify the masses, customers
become numbers, survey scores, transactions, or
data points. While summary statistics may be use-
ful for making sense of data, they don’t work well
for building empathy with the customer or under-
standing the emotional impact of their experience.
By identifying themes and surfacing stories, text
analytics can bring the customer experience to life
and motivate decision makers to take actions that
are more connected to their customers.

10MEDALLIA.COM
The Organizational
Challenges of Text
Analytics
Text analytics provides companies with the poten-
tial to mine customer feedback for insights hidden
deep within millions of customer comments. But
realizing that potential is not a given. While text
analytics brings big data capabilities to customer
experience management, early adopters are still
figuring out how to use these new capabilities most
effectively. According to a 2014 Forrester report,
many practitioners believe “it’s easy to get mired
in text analytics without delivering value to their
organization.”18
It is not unusual for companies to struggle when
adopting big data technologies. Although tech-
nologies like text analytics are designed to glean
intelligence from data, the technology alone rarely
creates value. It’s only when business processes
and technology align that companies realize the
benefits of their investment.
For example, BABIES“R”Us® customer support
agents, trapped in rigid service protocols and
boxed in by corporate silos, were powerless to send
a young mother her gift card. Even if the agent had
fixed Susan’s problem on the first call, the bigger
issue—a flawed promotional launch—would have
remained unresolved. Text analytics might be able
to surface problems that cross channels, and even
suggest root causes, but insights from analytics
alone can’t fix execution problems. For real change
to happen, managers must believe that the insights
are valid and must have the resources, influence,
and support needed to convert insights into action.
According to researchers at MIT’s Center for Digital
Business, while the technical challenges of using
big data analytics are very real, the managerial chal-
lenges are even greater.19
Customer experience
professionals using text analytics to drive strate-
gic and operational change face many of the same
issues. The most common challenges include:
Structural impediments
Technical specialists trained in text analytics are
often housed in a centralized function, separated
from the day-to-day realities of the business. They
may appear out of touch or insular. As a result,
frontline managers and employees don’t always
readily embrace their insights.
Data skeptics
People accustomed to traditional survey analysis
are sometimes skeptical of text analytics because
it doesn’t provide the definitive, quantifiable
insights that structured feedback data yield. Also,
because of the nature of human language, text
analytics will never be 100 percent accurate. While
accuracy above 85 percent is usually sufficient to
identify core themes in customer feedback, errors
will inevitably reduce confidence among those
who are unfamiliar with the technology.
New role requirements
Companies must be willing to redesign roles and
processes to accommodate and support their
new analytic capabilities. They must embrace
text analytics as a game changer in managing the
customer experience and be willing to search for
new opportunities. Customer experience profes-
sionals must take an active role in leading this
process. They must move beyond simply identi-
fying insights and monitoring brand compliance
to actively leading a multifunctional process of
data-driven discovery and decision making. This
requires new leadership and facilitation skills.

MEDALLIA.COM 11
Executive roles will also change. Executives who
typically rely on experience and intuition to guide
a company’s direction may have to reformulate how
they make decisions and where they can provide
the most value. While their experience will remain
valuable, their most important role may no longer
be coming up with the right answers, but knowing
the right questions.
Cultural misalignment
Embracing innovative ideas from customer
feedback requires an organizational culture
that encourages exploration and experimenta-
tion. When customer experience management is
viewed primarily as a mechanism for maintaining
consistency and brand compliance, there may be
relatively few avenues for innovation.

MEDALLIA.COM 12
Six Practices to Maximize
the Value of Text
Analytics
None of these execution challenges are insurmount-
able. In fact, they mirror challenges that companies
have faced for centuries when implementing new
technologies that create powerful new capabilities.
To realize the enormous potential of text analytics,
customer experience professionals are doing what
generations of managers have done before them:
they are learning how to use the technology most
effectively to improve the customer experience,
and they are adapting work practices, roles, and
decision-making processes to support and enable
the technology in generating value for the business.
To understand how some companies are suc-
cessfully leveraging text analytics to strategically
improve the customer experience, the Medallia
Institute interviewed customer experience profes-
sionals at 12 companies actively using text analytics.
While the companies varied in their specific prac-
tices, they all had considerable experience using
text analytics to better understand their customers
and drive improvement efforts. The interviews were
supplemented with examples from other compa-
nies as well as previous studies examining similar
technology adoption.
Our interviews revealed six practices that success-
ful customer experience teams are using to get the
most value from text analytics.
Practice 1: Become an indispensable
business partner
Companies succeed with text analytics because
their leadership teams value text analytics insights
and use those insights to make important decisions.
This doesn’t happen by chance. Effective customer
experience teams know how to frame big issues for
senior management, and they spend a great deal
of time actively engaging with managers who have
a deep understanding of the company’s products
and services.
Being an indispensable business partner means
working closely with line and functional manag-
ers to identify issues and jointly solve problems.
Without business direction and a well-informed
model, unstructured text analysis can be a big
waste of resources. The most effective customer
experience teams start with customer satisfaction
patterns that they know will attract the attention of
senior leaders. Then they collaborate directly with
managers who understand the business and have
the relevant domain expertise to guide the next
level of investigation.
For example, when the customer experience team
at an enterprise software company discovered that
customers were complaining about a new product,
the text analytics team immediately reached out
to managers in the product function. The product
group offered specific questions that helped to
clarify the problem and its origin: Which configura-
tions were reporting problems and how frequently?
Which features appeared to be buggy? Were the
problems happening with specific screens or loca-
tions or were they occurring more broadly? Working
together, the analysts and managers zeroed in
on the key questions and sources of information
needed to shape the investigation and, ultimately,
generated the insights required to take appropriate
action.
Being an indispensable business partner also
means proactively providing value to the business.
The customer experience manager at a major cruise
line used text analytics to resolve an unexpected
Effective CX teams know how
to frame big issues for senior
management.

MEDALLIA.COM 13
source of customer frustration. Surprised by a sud-
den drop in customer satisfaction ratings for one
of the company’s newly renovated ships, he noti-
fied the ship’s general manager. Working together,
the analyst and the GM used text analytics to form
hypotheses, investigate possibilities, and identify
the source of the complaints. Their investigation
revealed that the renovation had reduced the
size of the dining facilities in one part of the ship
and made it less convenient to get to other dining
options. With this information in hand, the com-
pany quickly came up with
a solution that allowed
customers to reach their
dining destinations more
easily, stemming the flood
of negative reviews. By
combining text analyt-
ics with the operating
expertise of line managers, the program manager
provided significant value to the business by iden-
tifying the root cause of a critical problem before it
impacted the ship’s bottom line.
Customer experience business partners also know
how to make a strong financial case. One customer
experience team we spoke with performed a set
of analyses that showed how text analytics could
be used to predict customer satisfaction scores.
The team also worked with its internal champions
to demonstrate that customer satisfaction scores
were directly associated with contract renewal
rates, which, in turn, influenced revenue. The anal-
yses made a concrete case for the financial value of
text analytics.
Practice 2: Bust silos with
cross-organizational dialogue
As customers increasingly interact with companies
across multiple channels, customer experience
problems rarely fit squarely within any one business
unit or function. Our findings suggest that customer
experience teams tend to be most effective when
they use text analytics results to stimulate dialogue
and debate across departments.
Take, for example, a US insurance company. During
strategic planning discussions, text analytics helped
spark the realization that the product development
and sales organizations would have to collaborate
more effectively for the company to increase sales.
As part of the planning exercise, executives drew
on text analytics combined with other analyses
to arrive at a shared understanding of the factors
causing coordination
gaps. Armed with these
new insights, the execu-
tive team made the bold
decision to integrate the
two groups within a single
reporting structure and
common leadership.
Using text analytics to convene managers and
stimulate dialogue across the organization serves
several purposes. First, it reveals when problems
are more pervasive than first thought. Managers
reviewing text analytics together may discover
other units struggling with similar challenges. This
brings more attention to common issues and stimu-
lates information sharing and joint problem solving.
Second, it brings together groups with multiple per-
spectives vital to generating ideas for subsequent
analysis. By convening representatives from across
the organization to make sense of text analytics
results, the customer experience team creates an
iterative process of discovery that leverages rele-
vant expertise. The process ensures that insights
are used not only to identify the root cause of cus-
tomer problems, but also to drive decisions that
lead to actions and resolution.
Third, cross-organizational dialogue builds owner-
ship and commitment to implementing solutions.
Referring back to our enterprise software com-
pany, once text analytics revealed the source of
CX teams are most effective
when they use text analytics to
stimulate dialogue and debate
across departments.

MEDALLIA.COM 14
the buggy software, the customer experience team
didn’t just inform the product group and walk away.
The team continued meeting with functional leads
to gain agreement that the problem was legitimate
and that responsibility for it belonged with product.
And they didn’t stop there. They worked with the
team to determine who would be accountable for
resolving the issue and by when.
Some of the most effective customer experience
teams create formal governance structures and
regular forums to review insights from text analyt-
ics and other analyses. These governing bodies
bring leaders together to discuss customer issues
and expedite decision making. Several companies
in our study created
“customer experience
councils” or “cham-
pions networks” that
convened stakeholders
from around the com-
pany to review critical
themes and topics identified through text analyt-
ics. This collaboration ensured not only that the
customer experience team addressed business
questions that managers cared about, but also
that strategic, cross-functional challenges were
resolved with input from diverse stakeholders.
Practice 3: Build empathy through
compelling stories
One of the biggest benefits of text analytics is that
it engages senior decision makers with customer
feedback, thereby increasing their empathy with
the customer and the customer’s experience. There
is no guarantee that insights generated by the text
analytics team will influence strategic or opera-
tional decisions. These decisions often require the
backing and support of senior executives, so get-
ting their buy-in is essential.
Text analytics can surface compelling stories that
vividly illustrate customer frustrations and the
consequences of a bad experience. The most com-
pelling stories capture and convey emotion. It’s one
thing for executives to see satisfaction scores for a
baby registry hovering at four out of ten points. It’s
quite another to read a customer’s angry complaint
on Facebook as she tells her friends to shop at the
competitor.
In many ways, text analytics incorporates principles
similar to those used in design thinking, a human-
centered approach to innovation. Like text analytics,
design thinking seeks to understand the customer
experience through the eyes of the customer, in
their own words, and
in the context of their
natural surroundings.
design thinkers recog-
nize that connecting
executives with cus-
tomers’deepemotional
experiences is a better way to spur action than
simply presenting them with graphs, charts, and
statistics. For example, when a product designer at
GE wanted to make a magnetic resonance imaging
system less frightening for sick children, he could
have presented lots of PowerPoint slides with
charts and numbers. But the story of a frail young
girl with tears running down her cheeks, having to
be sedated before she would lie still and alone in
the huge noisy machine, generated far greater sup-
port across the organization than any quantitative
analysis ever could.
Likewise, with text analytics, customer experi-
ence professionals can weave together convincing
stories of customer experiences to persuade exec-
utives and managers that change is needed.
Combining big data with “big stories” makes it
possible for companies to create the seemingly
impossible—customer empathy at scale.20
Combining big data with “big stories”
makes it possible for companies to
create the seemingly impossible—
customer empathy at scale.

MEDALLIA.COM 15
•
•
•
Practice 4: Use data to validate and
innovate
Text analytics is a powerful tool for discovery and
for challenging assumptions. It can identify sources
of customer frustration that point the way to prom-
ising opportunities for change and innovation. But
doing something new requires taking risks, and
managers are often reluctant to take risks when
they don’t trust the underlying analytics.
Despite the promise of text analytics, many man-
agers need reassurance that it’s actually producing
valid results. Experienced text analytics teams
build confidence in their analyses by replicating
known results that managers already accept. As
one data analyst said, “We don’t want to give them
some surprising insight without building a level of
confidence first”
the European telecom provider that discovered,
through text analytics, mounting frustration from
customers who complained that they had to call the
company to get a copy of their bill and then wait
for it to arrive via regular mail. With the insight that
customers wanted something more convenient, the
company moved quickly to introduce digital copies.
The new digital billing process cost the company
less than $10,000, saves approximately five dollars
on every transaction, and has cut postage fees by
50 percent. And customers are much happier.
Analyzing both structured and unstructured data
simultaneously can be one of the most powerful
ways to gain deep insights from customer feed-
back. Many companies mentioned the value of
incorporating other data into their decision mak-
ing, like NPS® or operational measures such as
first-time problem resolution. When comments and
quantitative ratings are combined, along with data
on customer demographics, segmentation, and
financial outcomes, companies can uncover critical
relationships and linkages that can lead to all sorts
of opportunities.
As an example, the same telecom provider
combined customer survey ratings and text ana-
lytics to identify which aspects of the service
experience best predicted a customer’s likeli-
hood to churn. Analyzing comment topics, along
with survey ratings, analysts determined the rela-
tive impact of the various experience metrics on a
customer’s likelihood to terminate a contract. The
company estimated that customers with the worst
experience—as measured by ratings and com-
ments—were five times more likely to churn than
those with the best experience. As a result, the
company made some innovative changes to its
contracts, which ultimately produced a 30 percent
increase in new customers and a 20-point jump in
NPS among existing customers.
Once managers believe text analytics produces
results that are reliable, they quickly recognize its
potential for innovation. Part of the unique value of
text analytics is its ability to surface customer ideas
that companies may otherwise miss. Consider
Strategies for building confidence
in text analytics
Track topics that are measured in other
customer surveys, then show that the
results from both the traditional survey
and text analytics tell the same story.
Look at text analytics results in relation to
a well-defined journey map. Demonstrate
that the topics with the highest comment
volume and impact correspond directly to
the journey map’s touchpoints.
Compare text analytics results to those
obtained through manual coding to high-
light consistent themes.

MEDALLIA.COM 16
Practice 5: Stimulate organizational
learning
Organizational learning occurs when insights from
one part of the organization combine with insights
from other parts of the organization to inform
broad-based improvements. But many large com-
panies still collect and analyze customer feedback
in silos. Fully integrating feedback across the orga-
nization continues to be a big challenge.21
Text analytics makes it possible to draw insights
from unstructured feedback captured across dis-
parate channels, functions, and business units.
By integrating customer feedback from these dif-
ferent sources, and combining structured and
unstructured data, sophisticated companies can
proactively detect important patterns and relation-
ships that they might otherwise miss.
Some organizations expand their learning oppor-
tunities by embedding specialized text analytics
capabilities in key decision-making pockets across
the company. For example, the five-person cus-
tomer experience team at another large telecom
company trained 20 “power users” to enhance the
central team’s reach and impact. Based in groups
like corporate marketing, product development,
and operations, power users help to explain text
analytics insights to their home departments and
show how local results relate to broader trends in
the organization.
Some organizations apply text analytics to both
their customer and employee feedback data to get
a more complete picture of the customer experi-
ence. Drawing on feedback from employees, a
regional manager at one large retailer discovered
a significant disparity among stores when it came
to employee perceptions. Employees at some
stores were clearly more satisfied with their abil-
ity to deliver an exceptional customer experience
than employees at others. To promote the diffu-
sion of best practice among stores, the manager
introduced a rotational program that encouraged
employees to move from store to store. The pro-
gram was so successful that other regions soon
followed.
To promote the diffusion of best practice among
stores, the manager introduced a rotational pro-
gram that encouraged employees to move from
store to store. The program was so successful that
other regions soon followed.
Practice 6: Operationalize text analyt-
ics by engaging local users
Engaging managers with text analytics at the local
level creates many benefits. By educating and
encouraging local managers to use text analyt-
ics to explore their own customer data, customer
experience teams can begin to operationalize
unstructured feedback much as they do structured
feedback. While the true power of text analytics
requires feedback from a large number of customer
transactions, the themes and insights become most
useful when they inspire action at the local level.
The companies in our study encouraged frontline
managers to use text analytics in several ways. In
most cases, companies leveraged insights gener-
ated centrally to stimulate more specific exploration
locally—for example, at a given location, property,
or call center. Some of the most effective text ana-
lytics teams develop exercises or prompts to guide
local managers in their exploration efforts, helping
them to look for specific issues or challenges that
have surfaced more broadly.
Some organizations apply text
analytics to both their customer
and employee feedback to get
a more complete picture of the
customer experience.

MEDALLIA.COM 17
Take the case of a large multinational retailer. To
get local managers to engage with issues deemed
important to the success of the brand, the trends
and insights team sends a quarterly newsletter to
store managers summarizing themes that have
emerged company-wide. Frontline managers have
access to text analytics for their individual store data
but not for feedback collected at other stores. The
newsletter highlights topics trending nationally and
encourages managers to use specific search terms
to review the topics in their own data and come up
with solutions that might address the issue in their
particular store.
Other companies choose to distribute text analytics
insights more extensively. The lead customer expe-
rience manager at a B2B cloud computing company
decided to leverage text analytics to empower indi-
vidual employees responsible for solving customer
problems. Each of the company’s 6,000 employ-
ees, across virtually every function, receives a daily
email digest with customer experience feedback,
including text analytics specific to the employee’s
account or role.
Employees can customize the content of the digest
by adding topics that interest them. For example,
a marketing manager might follow the operations
digest to monitor how customers are responding to
a new product release. When an engineer chose
to follow feedback reported to a customer support
team, he noticed that the team was struggling with
database problems. He contacted the customer
experience manager and was quickly connected to
the support team to help resolve the issue.
Good things happen when line managers can use
the results of text analytics to improve the cus-
tomer experience. Because customer experience is
often very local and context-specific, empowering
managers with text analytics gives them another
tool to interpret and respond to feedback unique to
their region or location. Consider this example from
a large retailer. A regional manager overseeing sev-
eral retail stores noticed that the words “coffee” kept
popping up in his analyses. Digging a bit deeper, he
discovered that customers were talking about how
theywishedtheycouldgetacupofcoffeewhilewait-
ing for their phones to be activated. Recognizing an
opportunity to improve the experience, the regional
manager asked store managers to offer customers
coffee while they waited, and customer satisfaction
scores skyrocketed.
Using text analytics, companies
can leverage insights generated
centrally to stimulate more
specific exploration locally.

MEDALLIA.COM 18
What’s Next for Text
Analytics?
In 2015, MIT Sloan Management Review and
Deloitte published a study of companies at various
stages of “digital maturity.”22
The authors con-
cluded that the most digitally mature businesses
deploy technologies like text analytics to improve
customer experience and increase efficiency. But
they don’t stop there; they use these technologies
to transform their businesses and move ahead
of the competition.
Transformation is never easy. Customer experi-
ence professionals must raise their game and
reexamine the role they play in creating strategic
change. No longer can they afford to look for new
insights solely through the lens
of traditional surveys and struc-
tured data. Instead, they must
embrace new technologies and
create a larger playing field in
their organizations. They must
move from support function
to strategic business partner,
working closely with managers across the company
not only to surface and resolve customer problems
but also to leverage the voice of the customer in
driving innovation and change.
Despite some growing pains, market research pro-
jections today strongly suggest that within a few
years, most companies will be using some type of
text analytics. Allied Market Research, a Portland,
Oregon–based technology research firm, projects
that text analytics technology will reach a com-
pounded annual growth rate of 25 percent over the
next four years, creating a $6.6 billion global text
analytics market by 2020.23
DMG Consulting calls
the adoption of text analytics “a requirement for any
company with a social media analytics program.”24
If these projections are right, widespread adop-
tion of text analytics will likely increase consumer
expectations still further and raise the bar for man-
aging the customer experience. As we’ve noted
elsewhere, if you’re not competing directly with
a disruptor, chances are your customers have
been served by one, and those experiences are
shaping expectations.25
Text analytics has the potential to become a far
more powerful driver of differentiation, innovation,
and growth in customer experience management
than the traditional methods that most companies
have become accustomed to. They allow custom-
ers greater freedom
to express their wants
and needs. And they
open up a deeper
level of customer
dialogue for the com-
panies that serve
them. Companies that
begin the journey now to deploy text analytics tech-
nologies creatively, confidently, and strategically
can look forward to those capabilities becoming
a sustainable source of competitive advantage for
years to come.
Customer experience
professionals must move from
support function to strategic
business partner.

MEDALLIA.COM 19
1
BABIES “R” US website (United States) Registry page
https://babyregistry.babiesrus.com/home?ab=BRU_
Header:Utility2:Baby-Registry:Home-Page
2
See the following sources for estimates. https://bakerretail.
wharton.upenn.edu/wp-content/uploads/2015/04/multi-
channel_shopping_exec_summary_Apr_2012.pdf; http://mays.
tamu.edu/center-for-retailing-studies/wp-content/uploads/
sites/18/2015/04/Kushwaha-and-Shankar-JM-2013.pdf; https://
idc-community.com/retail/retailomnichannelstrategies/john-
lewis-multichannel-shoppers-spend-35-times-mo
3
Millennials: Your Most Powerful Brand Advocates, Medallia,
Jan. 2015. http://www.medallia.com/resource/millennials-your-
most-powerful-brand-advocates/
4
Eric Clemmons, “Finding the New Market Sweet Spots: The
Art and Science of Being Profitably Different in the Era of the
Informed Customer”, Wharton School of Business, University of
Pennsylvania, October 11, 2015 p. 14 http://opim.wharton.upenn.
edu/~clemons/files/Find_New_Market.pdf
5
Merrill Lynch, Enterprise Information Portals, 1998
6
International Data Group, Inc. (IDG), as cited in the
Headwaters Group, accessed April 2, 2016. http://www.
theheadwatersgroup.com/your-unstructured-data-is-sexy/
7
Seth Grimes, “Text Analytics 2014: User Perspectives on
Solutions and Providers,” Alta Plana, July 9,2014
8
Nielsen Online, “Under the Influence: Consumer Trust in
Advertising,” Sept. 17, 2013, accessed, March 15. 2016, http://
www.nielsen.com/us/en/insights/news/2013/under-the-
influence-consumer-trust-in-advertising.html ,
9
Itamar Simonson and Emanuel Rosen, Absolute Value: What
really influences customers in the age of (nearly) perfect
10
The Conference Board CEO Challenge 2014. People and
Performance: Reconnecting with Customers and Reshaping
the Culture of Work,” The Conference Board, Research Report
R-1537-7-14-RR
11
Gartner Surveys Confirm Customer Experience is the New
Battlefield, Oct. 23, 2014 http://blogs.gartner.com/jake-
sorofman/gartner-surveys-confirm-customer-experience-new-
battlefield/
12
Customers 2020, Walker Research, http://www.walkerinfo.
com/customers2020/
13
Seth Grimes, “Text Analytics 2014: User Perspectives on
Solutions and Providers,” Alta Plana, July 9,2014
14
“Understanding the Multi-Channel Shopper,” Jay H. Baker
Retail Center, The Wharton School, University of Pennsylvania,
accessed April 2, 2016,
https://bakerretail.wharton.upenn.edu/wp-content/
uploads/2015/04/multi-channel_shopping_exec_summary_
Apr_2012.pdf15Customers 2020, Walker Research, http://www.
walkerinfo.com/customers2020/
15
Jonathan Brown, Harley Manning and Carla O’Connor,
“How to Use Text Analytics in Your VOC Program,” Forrester
Research, February. 25, 2014
https://www.forrester.com/How+To+Use+Text+Analytics+In+You
r+VoC+Program/fulltext/-/E-RES110422
16
Hyun Jeong Han, Shawn Mankad, Nagesh Gavirneni, and
Rohit Verma, “What Guests Really Think of Your Hotel: Text
Analytics of Online Customer Reviews,” Cornell Hospitality
Report, February, 2016.
17
Paolo Lorenzoni, “Why Sandwich Wraps are Sexier than
Cronuts,” Fast Company, Sept. 5, 2014, accessed April 2,
2016, http://www.fastcodesign.com/3033546/food-week/why-
sandwich-wraps-are-sexier-than-cronuts
18
Jonathan Brown, Harley Manning and Carla O’Connor,
“How to Use Text Analytics in Your VOC Program,” Forrester
Research, February. 25, 2014
https://www.forrester.com/How+To+Use+Text+Analytics+In+You
r+VoC+Program/fulltext/-/E-RES110422
19
Erik Brynjolfsson and Andrew McAfee, “Big Data, The
Management Revolution,” Harvard Business Review, October
2012
20
Tom Kelley and David Kelley, Creative Confidence:
Unleashing the Creative Potential in All of Us, Crown Business,
2013:13-18
21
Judith Lamont, “Delving into Customer Thoughts: Text
Analytics Provides Insights,” KMWorld, July/Aug, 2014: 13
22
Gerald C. Kane, Doug Palmer, Anh Nguyen Phillips, David
Kiron, Natasha Buckley, “Strategy Not Technology, Drives
Digital Transformation,” Sloan Management Review, Summer
2015
23
Apurva Sale, “Global Text Analytics Market, 2013-2020,”
Allied Market Research, Jan. 2015
24
DMG Consulting Abstract: 2015-2016 Speech and Text
Analytics Product and Market Report, accessed Mar. 1, 2016
http://www.dmgconsult.com/services/speech/abstract.asp
25Operationalizing Experience Management in the Age of the
Customer: The Future of Sustainable Advantage, Medallia, 2015
Sources

20MEDALLIA.COM
About Medallia
Medallia® is the Customer Experience Management company that is trusted by hundreds of the world’s leading brands.
Medallia’s Software-as-a-Service application enables companies to capture customer feedback everywhere the customer is
(including web, social, mobile, and contact center channels), understand it in real time, and deliver insights and action
everywhere—from the C-suite to the frontline—to improve their performance. Founded in 2001, Medallia has offices in Silicon
Valley, New York, London, Paris, Hong Kong, Sydney and Buenos Aires. Learn more at www.medallia.com.
Medallia is a registered trademark of Medallia, Inc. Net Promoter, Net Promoter Score and NPS are registered trademarks of Bain
& Company, Inc., Fred Reichheld and Satmetrix Systems, Inc. Other names may be trademarks of their respective owners.
Follow us: medallia-inc @Medallia Medalliablog.medallia.com
Beth Benjamin
Beth Benjamin is the senior director of Medallia’s CX Strategy Research group.
Prior to coming to Medallia she held positions at the Stanford Graduate School of
Business, the RAND Corporation, and three management consulting firms. She has
a PhD in business from the Stanford Graduate School of Business, MA in industrial-
organizational psychology from the University of Maryland, and BA in psychology
with an emphasis in industrial and labor relations from Cornell.
Joachim B. Lyon
Joachim Lyon is an organizational behaviorist who has spent over 8 years unearthing
stories about the future of work, organizations, and professions. He has most recently
held a position in the Organizational Design practice at innovation consultancy
IDEO, and is completing a PhD in organizational behavior in Stanford’s School of
Engineering. He holds an MA in philosophy from University of Edinburgh, Scotland,
and a BA in cognitive science with a focus on organizational cognition from the
University of California, San Diego.

The big story behind your big data: Six Practices for Making an Impact with Text Analytics

Recommended

More Related Content

What's hot (20)

Similar to The big story behind your big data: Six Practices for Making an Impact with Text Analytics (20)

Recently uploaded (20)

The big story behind your big data: Six Practices for Making an Impact with Text Analytics