3. We’ve reached another turning point.
For as long as humans have been able to communicate, we’ve been evaluating, recommending
and sharing. After thousands of years of word of mouth, these conversations began to take place
online, where it is now digitally archived.
Naturally, smart businesses got hungry. Here was a chance to capture what their customers — real
people actually using their products and services – were saying about them. But, as has happened
so many times before, technology eventually outpaced our capacity to truly understand its impact.
To companies that were hungry for social data, suddenly it seemed there was simply too much of it.
Navigating the flood of social content and arriving at meaningful insights about customers became
a challenge — and still is — for many businesses.
Social data isn’t slowing down, but we’re now able to get ahead of it, to make sense of it.
We’ve arrived at a turning point at which social data can be captured, analyzed—and most
importantly, acted upon—to drive true, customer-centric change across companies, and even
entire industries.
But this report isn’t about social data. It’s about the people behind the data. We’re excited to
share with you what we’ve learned about them from the hundreds of billions of conversations
our technologies have been a part of.
The Conversation Index you’re about to read is the first of many to come. Now, on to the insights…
Best,
Erin Nelson (@erinclaire)
Chief Marketing Officer, Bazaarvoice
4.
5. Table of Contents
The Conversation Index .......................................................................................................2
Consumer language reflects economic uncertainty ...................................................................5
All over the map: sentiment by region, country ......................................................................6
The Friday Facebook flood ................................................................................................. 10
Women fuel word of mouth ............................................................................................... 12
Controlling spending, leading conversations ........................................................................ 14
Summer warms traveler sentiment ..................................................................................... 18
Poor service destroys product sentiment ............................................................................. 21
One in five spread the love ................................................................................................22
Pivot language reveals product opportunities .......................................................................25
As price and complexity increase, consumer… ......................................................................26
UGC served at 3 kilohertz, volume soars ..............................................................................30
Conclusion ......................................................................................................................32
Citations .........................................................................................................................34
About Bazaarvoice ..........................................................................................................35
6. The Conversation Index
Brands that lead, capture, and analyze customer conversations
Our analysis reveals the following key takeaways:
develop a deeper understanding of what they must do to meet
and exceed expectations and become truly customer-centric The way consumers talk about products changes
enterprises. In The Conversation Index, you’ll find the latest during economic downturns.
and most actionable customer insights, drawn from millions of Distinctions between offsite, onsite, and siloed channels are
customer conversations. crumbling, but some still exist. Onsite and offsite consumer
behavior is largely similar in some areas, but Facebook
The substance of these conversations is user-generated content
behavior is unique.
(UGC), an insight-rich form of social data that enables analysis at
every level of focus—from global consumer trends, to product- Those who control spending in a product category lead the
and customer-specific sentiment, and all degrees of granularity online conversation about that category.
in between.
Product quality and interactions with brand representatives
In serving over 220 billion pieces of UGC (and counting), are inseparable in the minds of many consumers—poor
Bazaarvoice has collected massive amounts of social data for customer service is the product sentiment killer.
hundreds of top brands across industries. From this pool of social Innovative product improvement ideas, as well as flaws, can
data, our Social Analytics team has extracted select findings to be found just as readily in positive reviews as in negative
be featured in this report. reviews by searching for pivot language.
The findings in this report are aggregated based on a 5 million Consumers seek facts and details, not opinions and
data point sample of UGC indexed in Q2-2011 (April–June), from subjective comparisons, when the price and complexity of
consumers in 210 countries around the world. products and services increase.
2
7.
8. Q1 2009: Highest number of reviews mention
Dow Jones 120 14K price when Dow and CCI at lowest points. 14% Price
Industrial references
Index 13K 12%
100
12K
Consumer 10%
80
Confidence 11K
Index 8%
60 10K
6%
9K
40
4%
8K
20 2%
7K
0 6K 0%
Q2 2008
Q3 2008
Q4 2008
Q2 2009
Q3 2009
Q4 2009
Q3 2007
Q4 2007
Q1 2008
Q1 2009
Q2 2010
Q3 2010
Q4 2010
Q1 2010
Q2 2011
Q1 2011
4 Price reference vs. Dow Jones Industrial and Consumer Confidence Index
9. Consumer language reflects economic uncertainty
One in ten reviews in Q2 mentioned “price.” To understand Mapping price references in reviews to the Dow Jones Industrial
how economic conditions affect price sensitivity in reviews, the Average reveals an even stronger negative correlation of -.68.
Bazaarvoice Social Analytics team compared references to price The pattern is nearly the same. Beginning September 2008,
in customer reviews to two key economic indices around and conversation around price skyrocketed as the Dow average
during the Great Recession, from July 2007 to August 2011. continued its fall. A second pattern began January 2010,
continuing until October 2010, with price mentions falling as the
When comparing price references to the Consumer Confidence Dow average improved.
Index, an overall correlation of -.66 was determined, meaning
reviewers mention price more often when the CCI is low. The During economic downturns, use UGC to promote value and ease
CCI sank to its lowest point in February 2009, the same month the hesitation price-sensitivity brings. Brands should highlight
price mentions hit a new high; 11.5% of all US reviews explicitly reviews and answers that praise a product’s high quality for its
mentioned price. price. Additionally, they should feature value-touting customer
content in online and offline advertising to reassure potential
customers that their money will be well spent.
5
10. All over the map: sentiment by region, country
Overall, customers in the southwest Pacific region showed the Even geographic proximity isn’t a reliable indicator of sentiment.
highest overall sentiment this quarter, followed by the North Moldova has the highest average rating of any country (4.54
Asia and Europe regions. out of 5 stars), while its neighbor Ukraine has the lowest (3.31).
While they share a border, the two countries are further apart in
Interestingly, sentiment by country doesn’t seem to correlate product sentiment than any two countries in the world,
with two intuitive explanations – money and happiness. according to our data.
We found essentially no correlation between a country’s
purchasing power and sentiment, and essentially no correlation
between sentiment and scores on Gallup’s Global Wellbeing
Survey, either.
Blog: Do happier people
leave higher product ratings?
http://bv-url.com/a5r9
6
11. 15 most
positive countries Average rating
Moldova 4.54
Georgia 4.52
15 most
Latvia 4.45 negative countries Average rating
Chile 4.41 Pakistan 4.00
Russia 4.38 Spain 4.00
Bulgaria 4.36 Nigeria 3.99
Australia 4.35 Romania 3.99
United Kingdom 4.34 United Arab Emirates 3.98
New Zealand 4.33 Argentina 3.98
Puerto Rico 4.32 Indonesia 3.97
Kenya 4.31 Peru 3.96
Germany 4.31 Algeria 3.95
Azerbaijan 4.31 Uruguay 3.94
Macedonia 4.30 India 3.89
Ireland 4.30 Morocco 3.80
Thailand 3.79
Iran 3.73
Ukraine 3.31
12. Europe
Average rating: 4.26
North America
Average rating: 4.23
North Africa
Average rating: 3.97
Mexico & Central America
Average rating: 4.21
South America
Average rating: 4.14
8 Global average ratings by region
13. North Asia
Average rating: 4.32
East Asia
Average rating: 4.09
South & Southeast Asia
Average rating: 3.95
West & Central Asia
Average rating: 4.11
Southwest Pacific
Average rating: 4.35
Sub-Saharan Africa
Average rating: 4.16
Ratings based on a 5-star scale 9
14. The Friday Facebook flood
Nearly a third (32%) of all reviews collected on Facebook in Q2 data, where 21% of reviews gathered this quarter were
were submitted on Fridays. The remaining reviews were collected submitted on a Tuesday, the most popular day of week for this
fairly evenly throughout the week. activity. In fact, Fridays accounted for only 11% of onsite review
submissions—only Saturdays (9%) saw lower activity rates.
However, a study of 50,000 posts from the top 14,000 pages
Facebook reviewers are also more on Facebook found that Friday posts had the highest weekday
interaction rates2, so it’s likely that when brands invite Facebook
positive, with a 12% higher average users to review their products, the calls to action posted on
sentiment for Facebook reviews Fridays are driving more review submissions than similar posts
throughout the week.
over reviews not collected on the
Facebook reviewers are also more positive, with a 12% higher
social network. average sentiment for Facebook reviews over reviews not
collected on the social network. With many of these reviews
coming from existing Facebook fans, this can be attributed to the
This Friday spike in Facebook reviews doesn’t match overall more positive disposition of Facebook fans toward the brands
usage patterns for Facebook pages, which see the most activity they like. In fact, 60% of Facebook fans say they’re more likely
on Wednesdays.1 Nor does it mirror onsite review submission to recommend the brand since becoming a fan.3
10
16. Women fuel
word of mouth
Women dominate social networks, as the majority on both
Facebook (58%) and Twitter (64%).4 This dominance is also seen
in onsite social use, with women producing 60% of all UGC in Q2.
Women were also more positive than men. The average rating for
female-written reviews was 4.43 stars out of 5, as compared to the
male average of 4.32 stars.
Overall, consumers age 35-44 contributed the most UGC across
Women produced Bazaarvoice clients this quarter. Collectively, consumers between
60% of all UGC in Q2. 25 and 54 years old were the biggest content drivers, contributing
70% of all UGC. The majority of social network users also fall in
this age range.4
17. 5.0
16% 16%
15%
4.5
11% 11%
9% 9%
Average 4.0
rating from
6%
women
Average 3% 3%
rating 3.5
from men
25-34 35-44 45-54 55-65 66+
F M F M F M F M F M
Reviews from women, Reviews from men,
percent of total, by age percent of total, by age
All reviews by gender and age 13
18. Controlling spending,
leading conversations
In consumer packaged goods (CPG), the vertical with the highest
average sentiment (4.68, 11% above the cross-industry average),
women contributed a huge majority of UGC in Q2 (84%), and
wrote 3% more content when doing so.
This is likely due to the fact that women typically control CPG
budgets for their households. In the US, women control 85% of
household spending and 93% of food purchases.5 It follows that
the women researching, considering, and purchasing CPG are
also talking about and recommending these products online.
Blog: Social media for the sexes: three
ways to capitalize on gender differences
http://bv-url.com/82sk
18
19. 5.0
25%
Average 4.5 20% 20%
rating
from men
Average 13%
rating from 4.0
women
6%
5%
4% 4%
3.5 3%
2%
25-34 35-44 45-54 55-65 66+
F M F M F M F M F M
Reviews from women, Reviews from men,
percent of total, by age percent of total, by age
Consumer packaged goods reviews by gender and age 15
20. 5.0
4.5 19%
17%
Average 16% 16%
rating from 15%
women
4.0 10%
Average
rating
from men
3.5
1% 1% <1% 4%
25-34 35-44 45-54 55-65 66+
F M F M F M F M F M
Reviews from women, Reviews from men,
percent of total, by age percent of total, by age
16 Consumer electronics reviews by gender and age
21. The CPG conversation peaked among 45-54 year olds,
with 35-44 and 55-65 trailing just behind. CPG saw higher
CPG review readers show
sentiment among older generations this quarter, keeping with
historical patterns. 6% higher average order
The consumer electronics vertical showed the opposite gender value than visitors who
makeup, with a large majority (77%) of word of mouth generated
by men, the plurality of which came from men age 45-54. don’t read reviews.
Again, we see that those who control the budgets tend to
control the conversation: men spent 53% more than women on
consumer electronics in 2010.6 To influence these spenders,
solicit and display UGC wherever purchase decisions are made.
In addition to product and category pages online, use customer
words and product ratings in advertising, on packaging, in
coupons, and via mobile to reach shoppers in stores.
21
22. Summer warms
traveler sentiment
The travel vertical saw a 4% increase in average rating in Q2
to 4.46 stars on a 5 star scale, while most other verticals
only fluctuated +/- 1%. We saw a similar seasonal sentiment
increase for the travel industry in Q2 2010, and the sentiment
spike is likely caused by the fact that summer is the peak
season for pleasure travel.
Overall, travel had the longest average review of any vertical
in Q2. Travel has always been a category that lends itself to
storytelling, so this is no surprise.
Blog: Use this word of mouth
metric to create customer fanatics
http://bv-url.com/yppj
22
23. All Consumer Mass
industries Automotive electronics CPG merchant Manufacturing Retail Travel
5.00 5.00 5.00 5.00 5.00 5.00 5.00 5.00
4.50 4.50 4.50 4.50 4.50 4.50 4.50 4.50
Average
rating
4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00
Q1 Q2
Word
count
71 94 101 59 58 77 66 110
Average rating by industry 19
24. 91% lower average rating when
tagged with customer service.
Percent of 18% 5.0
reviews with
customer 16% 4.5 Average
service 14% rating
mentions 4.0 based
12% on a 5
3.5 starscale
10%
3.0
8%
2.5 Average
6%
rating for
4% 2.0 reviews with
customer
2% 1.5 service
mentions
0% 1.0
ALL Clients
Retail
Manufacturing
Financial services
Apparel
Consumer electronics
Home improvement
Housewares
Mass merchant
Sporting goods
20 Customer service mentions by industry
25. Poor service destroys product sentiment
Across industries, 6% of reviews mentioned customer service a higher expectation of service from these businesses and are
in Q2. Overall, the average rating for these reviews was a therefore more likely to discuss service as part of the product.
staggering 91% lower than the general average rating. Many
consumers are clearly unwilling to evaluate a product’s quality Historically, February sees an increase in customer service
separately from their experiences with brand representatives. A mentions in reviews. This is likely due to the surge of holiday gift
great product won’t save brand word of mouth if the company purchases and returns, which accounted for nearly one quarter
doesn’t support customers with good service. of all retail returns in 2010.7
Though reviews are inherently product-specific, brands shouldn’t
overlook the massive opportunity to learn more about how
Financial services saw the highest customers view other elements of the brand like customer
service. Rather than waiting for and reacting to individual
portion of reviews that mention inquiries, proactively search for customer service issues in UGC.
customer service (17%). When appropriate, responding to service complaints in line with
a review shows other customers that service teams are truly
listening to their feedback, and can present an opportunity to
Financial services saw the highest portion of reviews that turn a negative experience into a positive one.
mention customer service (17%), but also showed the smallest
Additionally, aggregated customer feedback offers insights
gap in average rating between service reviews and reviews not
businesses can act on to improve their overall customer
mentioning service (62%). Service industries can expect a higher
experience. Analyze reviews mentioning service for key positive
rate of customer service mentions in reviews, as customers have
and negative themes, and distribute these trends to service
teams to reveal how to delight customers.
21
26. One in five
spread the love
One in five reviews in Q2 contained the word “love.” Ratings for
reviews containing “love” averaged 4.55 stars—10% higher than
the overall average.
The word “easy” was found in 7% of reviews overall in Q2, and
most often in the consumer electronics (19%), automotive (16%),
and financial services (16%) categories. Consumers in these more
complicated purchase categories are typically looking for ease
of use. One tenth of the reviews that contained “easy” also
contained both “quality” and “value.”
Analyzing high-rated and low-rated reviews for key themes
The word “easy” was reveals exactly what customers like and dislike about a product.
Brands should emphasize key words from positive reviews in
found in 7% of reviews product copy and marketing materials to highlight a product’s
best attributes.
overall in Q2.
27. “love” 4.55
“easy” 4.50
average
rating
7%
frequency
of word
average
22% 1%
rating
frequency “difficult” frequency
of word 3.62 avg. rating
of word
“price” 4.09
average 11%
rating frequency
of word
Word themes in correlation to ratings 23
29. Pivot language reveals
product opportunities
Low-rated reviews often contain ideas for improving products. But
when it comes to UGC, businesses shouldn’t rely on ratings alone to
uncover opportunities to innovate. While it’s not surprising that 86%
of reviews mentioning “never again” were one- or two-star, many
reviews contain “pivot language”—critical review points within an
otherwise positive review with a high rating.
For example, 75% of customers who “wish” a product was better
in some way still rated the product four or five stars. And 87% of
reviews that identified a product’s “only problem” also rated the
product four or five stars.
Even the best products can get better, so don’t focus solely on low-
rated products for opportunities to improve. Devote just as much
scrutiny to high-rated products for innovation suggestions. And while
it’s important to closely monitor negative reviews for potential issues,
some of the best insights for improvements come from a product’s
biggest fans. Use text analysis to locate pivot language in positive
reviews and uncover trends in suggestions. These insights empower
product teams to co-innovate better products with customers.
25
30. As price and complexity increase,
consumer questions get specific
The questions consumers asked about products in Q2 can Brands in these categories should closely monitor Q&A
be divided into two broad types—seeking questions and communities in near real time, and have representatives provide
discussion questions. fact-based answers to help shoppers find the right products
for their specific needs. Analyzing trends in seeking questions
“Seeking” questions ask for product-specific use cases and also gives insight into which product attributes and information
information, and are looking for facts rather than opinions. brands should emphasize in product copy.
Example seeking questions include:
“Discussion” questions are open-ended, or concern categories
“Does GPS navigation in this truck come standard?” or comparisons of products, and are often invitations to other
consumers to weigh in with more subjective answers and
“At what decibel do these speakers max out?”
recommendations. Example discussion questions include:
“Does this hotel offer free wireless internet?”
“What type of baby wipes do you recommend for infants?”
Seeking questions were more prevalent in verticals where
“Which of these two sweaters looks better on my body type?”
products are more expensive and complicated. Consumers risk
more if they have less knowledge about these products, so they “Which of these lamps is the best, and why?”
make sure to ask detailed questions before they purchase. In the
automotive, travel, and consumer electronics categories,
seeking questions comprised 81%, 79%, and 58% of
questions asked, respectively.
26
33. Discussion questions were seen most often in verticals where
product price and complexity were lower, allowing more leeway
to discuss subjective favorites and opinions. In the CPG, retail,
and general manufacturing verticals, discussion questions
comprised 64%, 58%, and 53% of questions asked, respectively.
For subjective discussion questions, answers from brand
representatives are less helpful and trusted than answers from
actual users. Brands in categories prone to discussion questions
should focus more heavily on fostering peer-to-peer discussion
communities amongst their customers. Follow up purchases and
review submissions with an email, inviting customers back to a
brand site to share their trusted, first-hand product knowledge
with shoppers. Site visitors who interact with
both reviews and customer
Q&A are 105% more likely to
purchase. They spend 11%
more than visitors who don’t
interact with UGC.
33
34. UGC served at 3 kilohertz, volume soars
Overall, UGC volume and submissions continue to soar across Answer volume continues to outpace question volume, keeping
industries. Q2 showed 34% year-over-year growth for review with the historical pattern.
volume. Overall average rating across all industries studied was
4.22 out of 5 stars for the quarter. Retailers saw the largest These increases in volume correlate with an increase in UGC
increase in volume this quarter, far outpacing other verticals. impressions. Two hundred and seventy-six million pieces of UGC
were served to online visitors each day in Q2—that’s over 3,000
Question and answer volumes likewise grew this quarter, with pieces every second, or 3 kilohertz.
119% growth in questions and 129% in answers year over year.
9M Running
8M total of
review
Monthly 7M count
review
volume 6M (Since
3/1/2010)
trend 5M
4M
3M
2M
1M
0
March 2010
April 2010
May 2010
June 2010
July 2010
August 2010
September 2010
October 2010
December 2010
January 2011
February 2011
March 2011
April 2011
May 2011
June 2011
November 2010
30 Monthly review growth across all industries
36. Conclusion
Social data is a wellspring of rich customer and product
intelligence. By analyzing customer conversations,
businesses uncover insights to develop a holistic picture of
their customers—who they are, what they want, and how
they talk about products and services.
But these insights are only as powerful as the actions they
enable companies to take and the choices they inform. Our
marketplace is increasingly driven to deliver exactly what
customers want, as soon as they want it. Only the companies
that can operationalize their social insights across the
business, and empower business decision makers in every
function with the insights to make customer-centric choices,
will fully capture the value of social data.
36
37. Contact us
This paper should raise just as many questions as it answers.
Contact us at bazaarvoice.com to see how we help brands
gain invaluable consumer and product insights by
enabling conversations.
US: (866) 522-9227
UK: +44 (0) 208.080.1100
France: +33 (0) 1.53.53.14.99
Germany: +49.89.24218508
Netherlands +31.20.301.2169
Australia/Asia-Pacific: +61.2.9362.2200
Sweden: +46.8.463.1083
37
38. Citations
1 “When Are Facebook Users Most Active? [STUDY].” Mashable.
http://mashable.com/2010/10/28/facebook-activity-study/
2 “Weekend Facebook Posts Shows High Interaction Rates.” Momentus Media.
http://momentusmedia.com/blog/?p=899
3 “Consumers Engaged Via Social Media Are More Likely To Buy, Recommend.” Chadwick Martin Bailey.
http://blog.cmbinfo.com/press-center-content/bid/46920/Consumers-Engaged-Via-Social-Media-Are-More-
Likely-To-Buy-Recommend
4 “Twitter, Facebook And LinkedIn: Age, Ethnicity And Gender Of The Major Social Networks [STUDY].” All Twitter.
http://www.mediabistro.com/alltwitter/pew-social-network-age-ethnicity-gender_b11681
5 “Social Media Marketing to Women: What Every Brand Needs to Know.” Pontiflex.
http://www.pontiflex.com/download/socialwomen.pdf
6 “Do Women Really Control 80% of Household Spending?” The Wall Street Journal.
http://blogs.wsj.com/numbersguy/do-women-really-control-80-of-household-spending-1054/tab/interactive/
7 “Returning Those Unwanted Gifts: The Life Cycle of a Gift Receipt.” Milo.
http://milo.com/blog/returning-those-unwanted-gifts-the-life-cycle-of-a-gift-receipt/?display=wide
38
39. About Bazaarvoice
Bazaarvoice’s Software as a Service (SaaS) solutions have powered more than 240 billion customer
conversations on brand web sites like Best Buy, Blue Shield of California, Costco, Dell, Macy’s, P&G, Panasonic,
QVC, and USAA. The company connects organizations to their influencers through a unique network that reaches
hundreds of millions of consumers around the globe, enabling authentic customer-powered marketing. Through
syndication, analytics, partnerships, and consulting, Bazaarvoice brings the voice of the customer to the center
of their clients’ business strategy, proving “social” can drive measured revenue growth and cost savings for
manufacturing, retail, travel, and financial services companies. Headquartered in Austin, the company has offices
in Amsterdam, London, Munich, Paris, Stockholm, and Sydney. For more information and access to client success
stories, visit www.bazaarvoice.com, read the blog at www.bazaarvoice.com/blog, and follow on Twitter at
www.twitter.com/bazaarvoice.
39