體育學報特別議題專刊 Social media in sports marketing 1

體育學報 DOI:10.3966/10247297201712500S001
中華民國體育學會
第 50 卷第 S 期，頁 1～18 (2017. 12.)

The applications of social media

in sports marketing
Yu Huang 1 and Jen-Hsien Hsu2
1
Department of Physical Education, National Tsing Hua University, Hsinchu City, Taiwan
2
Leicester Castle Business School, De Montfort University, UK

Abstract
Introduction: In the era of big data, sports consumer’s activities in social media become valuable
assets to sports marketers. In this paper, the authors review extant literature regarding how to
effectively use social media to promote sports as well as how to effectively analyze social media
data to support business decisions. Methods: The literature review method. Results: Our findings
suggest that sports marketers can use social media to achieve the following goals, such as
facilitating marketing communication campaigns, adding values to sports products and services,
creating a two-way communication between sports brands and consumers, supporting sports
sponsorship program, and forging brand communities. As to how to effectively analyze social media
data to support business decisions, extent literature suggests that sports marketers to undertake
traffic and engagement analysis on their social media sites as well as to conduct sentiment analysis
to probe customer’s opinions. These insights can support various aspects of business decisions, such
as marketing communication management, consumer’s voice probing, and sales predictions.
Conclusion: Social media are ubiquitous in the sports marketing and consumption practices. In the
era of big data, these “footprints” can now be effectively analyzed to generate insights to support
business decisions. Recommendations to both the sports marketing practices and research are also
addressed.
Keywords: social media, social media analytics, big data

Owing to such traffic, Facebook created $27.6

Introduction billion revenue from its advertising services and
was valued at $385 billion, making it the world's
Social media are now highly involved in sixth biggest company (The Telegraph, 2017).
business practices. For instance, Facebook has Such trend has also greatly influenced the
attracted 2.1 billion users around the world. landscape of sports marketing. From the
通訊作者：黃 煜 國立清華大學體育系 TEL: 03-5715131 轉 71500
300 新竹市光復路二段 101 號 E-mail: [email protected]
2 體育學報特別議題專刊

consumption side, there have been around 442.7 Secondly, advanced design, such as live
million sports fans who registered on sports streaming application, allow sports fans to watch
related Facebook pages in 2017, an increase of live program on smart phones. Many
21% compared to the figure in 2016. Other social behind-the-scene contents, such as interviewing
media platforms, such as Twitter, Instagram, and athletes in the locker room, are now transmitted
YouTube, also had a tremendous increase on in real time (Hochberg, 2015). All of such enrich
subscriptions or registrations (Sportcal, 2017). sports content as well as enhance entertaining
As to the impact upon sports marketing values of a sports brand. Thirdly, some
practice, social media are now widely employed corporations use social media to manage
in various sports sectors, such as professional customer or fan relationship management (Luarn,
sports leagues, national sports governing bodies, Chiu, Yang, & Kou, 2014). Lastly, social media
sports event organizers, sports sponsors, and elite help support sports sponsorship events, such as
athletes. Compared to 2014, 15 out of 16 Polish making sponsors viral among sports fans. For
football teams use utilized Instagram in 2016 to instance, Los Angeles Dodgers inserted sponsor's
promote their sports events (Siguencia, Herman, messages into the player statute within the
Marzano & Rodak, 2017). Elite athletes also stadium. Such an appealing scene made the
utilize social media to have direct interaction stadium became a popular geotagged location and
with their fans (Korzynski and Paniagua, 2016). also helped the sponsor reach more fans
Some of famous athletes have million followers (International Event Group, 2015). To sum up,
that actively participate in their social media social media greatly support sports marketing
pages, which further increases these athletes’ practices, such as for public relations, customer
popularity and commercial values. Sports satisfaction and sponsorship management.
organizations also exploit these advantages of In the era of big data where considerable
social media. For instance, the Wimbledon tennis advancements have been made in the aspects of
event organizer produces video contents and have data collection, data storage, and data analysis
them disseminated through various social media techniques, data analysts are now capable of
outlets, including snapchat, periscope, and processing high volume, velocity and variety data
Instagram (International Event Group, 2017a; to get insights for business decisions (Beyer &
International Event Group, 2017b). Due to its Laney, 2012; De Mauro, Greco, & Grimaldi, 2015;
inexpensive cost, Eagleman (2013) found that Qiu, Wu, Ding, Xu, & Feng, 2016). Social media
almost all national sports governing bodies data are one of the sources that are widely used
launch their own Facebook and Twitter pages to for data analysis owing to their characteristics
attract sports consumers. (Tirunillai & Tellis, 2012; Zhu & Zhang, 2010).
It is evident that social media create Firstly, they are now widely accessible to
tremendous business values to sports marketers. marketers with a low cost. More importantly,
Firstly, social media are now a very important social media data are spontaneous, passionate,
promotional tool for non-mainstream sports and temporally disaggregate (days, hours,
organizations to improve the publicity as well as minutes), which faithfully reflect consumer’s
to provide interactive forum (Vann, 2014). attachment, emotions, and opinions towards a
 Social media in sports marketing 3

brand (Tirunillai & Tellis, 2014). For instance, then to surf the friends lists (Boyd & Ellison,
many sports celebrity attract millions of global fans 2008). With the development of web technology,
to join their social media sites. Or fans do not there are now various types of social media
hesitate to express their voice on social media in platforms that provide the aforementioned
some international sports events. These interactions web-based services to consumers. Kaplan and
indeed create a massive dataset. As such, how to Haenlein (2010) utilized the level of
analyze social media data is now a crucial task for self-presentation (low vs. high) and that of media
sports marketers and there is a need to investigate richness (low vs. medium, vs. high) to categorize
the agenda of social media analytics so as to social media into 6 types, including micro blogs
provide implications to both sports marketing (high self-presentation and low media richness),
researchers and practitioners. social network sites (high self-presentation and
In summary, social media widely influence medium media richness, e.g., Facebook), virtual
the landscape of sports marketing - not only do social worlds (high self-presentation and high
consumers highly rely on them in the consumption media richness), collaborative projects (low
process but sports marketers ubiquitously make self-presentation and low media richness, e.g.,
use of social media for business purposes. With the Wikipedia), content communities (low
advent of data analytics, social media data now self-presentation and medium media richness, e.g.,
offer great implications to sports marketers for YouTube), and virtual game worlds (low
marketing decisions. Owing to a dearth of attention self-presentation and high media richness). In short,
in the field of sports marketing, in this paper, the the term social media encompass six types of
authors attempt to investigate how the social media web-based services. To date, there are around 12
are now employed in different sports settings as billion active social media users (Statista, 2017).
well as to review extant literate on social media Such trend also influences the landscape of
data analytics. sports consumption. In 2017, there are around
747 million users who utilize social media to

Social media and sports access sports related contents. For example,
Facebook has attracted the most amount of users
consumption (442.7 million), following by Twitter (174.2
million), Instagram (95.2 million), and YouTube
Boyd and Ellison (2008, p.211) defined
(34.9 million) (Sportcal, 2017). Compared to the
social media as web-based services that allow
figures in 2016, there was a 26% increase in
individuals to construct a public or semi-public
terms of sports social media users. Among which,
profile within a bounded system, to articulate a
Facebook has attracted the most amount of new
list of other users with whom they share a
users (77.3 million) whereas YouTube has the
connection, and to view and traverse their list of
highest increase rate (72%). All of such indicate
connections and those made by others within the
how social media are entrenched in the sports
system. Based upon this definition, the first
consumption process at a global level.
social media platform was introduced in 1997
termed “SixDegrees.com” which allowed users to
create their own profiles, to list their friends and
4 體育學報特別議題專刊

The application and analysis foregoing goals through social media (Eagleman,
2013).
of social media in sport
Event organizers
As a result of overwhelmingly phenomena of FIFA World Cup is viewed as the most
usage of social media consumption, sport popular sports event in the world in terms of
managers in all sectors of sports setting could not viewership. Such popularity is fully shown on
but utilize new media platform. To understand Tweeter pages. For example, the 2014 World Cup
how this advanced technology of communication event received approximately 672 million tweets
was embraced in the sport sector, this following (Billings, Burch, & Zimmerman, 2015). They
section presents the application of the social attempted to analyze complementary content in
media in various segments of sport industry as the World Cup tweet from the perspective of
well as how they can be benefited from the nationalistic. The commentary was divided into
advanced analysis of mass data. three dimensions, including the attributions of
National governing body success, the attributions of failure and the

According to the study of usage of social characterizations of personality/physicality. For

media among national sport governing bodies example, attributes of success/failure may include

(hereafter, NGSBs) in U.S., the results indicated playing styles, team poises, playing approaches,

that all the NGSBs have their own Facebook experiences, intelligence and fortune/consonance.

pages and 98.4% of which manage their own Characterizations of personality/personality/

Twitter pages. Slightly more than half (51.6%) of physicality were categorized into extroversion,

NSGBs took Youtube. Other social media introversion, emotion, appearance, size/parts of

platforms, such as LinkedIn, Pinterest, body and other. They found that the audiences

FourSquare, and Google+, were also widely were more likely to be highly identified football

utilized. There are two reasons that explain why fans. These fans expressed much how they enjoy

these governing bodies widely used social media: the skills of players rather than fervent patriotism.

for sports fan relationship management and Audiences made more comments about opponents

marketing communication purposes. For the out of field of the 12 classification distinctions.

former, the organization can actively disseminate This may indicate that fans expressed various

news related to the organization itself, guide fans emotions in the social media environment.

how to enjoy the sport, bridge the link between Wimbledon, one of the most prestigious

the organization and fans, as well as set forums to tennis events in the world, attempted to adopt

create interaction among fans. As to the latter, new media while maintaining traditional values

NSGBs can use social media to increase publicity, as well as new communication models. This event

both for the organization itself and the sport that organizer not only create as many contents as

it advocates. Consequentially, it is likely for they can but also build multiple platforms. They

NSGBs to improve their image, to maintain their generated three-series video contents. The first

relationships with the public and press. More type is pre-planned. Some well-known tennis

importantly, it is also cost effective to achieve the players will be involved in the production of
these short videos prior to the tournament. These
 Social media in sports marketing 5

videos will be launched for the purpose of Professional sports team

positioning. The second type is preemptive. It is evident that professional sports leagues
Certain moments are very important for this in various continents make use of the social
tournament, including the open match with media boom as well. Siguencia, Herman,
defending championship at the center court and Marzano, and Rodak (2017) showed that only
five championship match at the end of the week. nine of 16 football teams in Polish Top League
Also, how to connect traditional holidays utilized the Instagram in 2014. The number has
throughout the year with the tennis event should been increase to 15 in 2016. It also indicated that
be employed. The third type is reactive. Some number of photo posted thru Instagram was
celebrities come into the game and interact with skyrocketed in certain seasons as well. Siguencia
any given event person. This scene should be et al. (2017) concluded that these teams only
very valuable content. This means that the event posted a lot of photos and short videos for
organizer should be ready for such things. In promotion purposes but there is a need to put
terms of multiple platforms, the tennis event more effort to create interactions with their fans..
organizer utilizes Snapchat, Periscope, and Social media seem to be a valuable
Instagram. Snapchat is used for creating a short marketing tool for non-mainstream professional
videos with some special guests/celebrities on sports leagues as most of the mainstream media
site. Periscope is aimed to bring tennis do not show much interest to cover their events.
heavyweights a behind-the-scene tour so that fans Therefore, these secondary professional sports
would be surprise the role play of these tennis leagues have to take this opportunity to promote
heavyweights. Instagram was used to cover themselves through the use of social media. For
unique aspects of the tennis amenity as well as example, despite that the Australia netball
infrastructure which enrich the experience of the Championship has a huge number of female
heritage event (International Event Group, 2017a; participants and spectators, it does not have much
International Event Group, 2017b). exposure from the mainstream media (Vann,
Running-mania is widespread all over the 2014). As such, Vann (2014) investigated how the
world. More and more running race was created for Australia netball federation utilize Twitter helps
different markets. The Fun run aims to attract new to promote their events. This study determined the
runners to enjoy the race within a very entertaining performance indicator to investigate the efforts by
atmosphere. The Color run was introduced to these women’s teams, including the number of
Taiwan in 2013. Its fan page accumulated around tweets in the dataset, the number of unique users
130,000 followers within 6 month’s time. Its contributing to the data set, the percentage of
content, including messages, photos and videos, did original tweets in the data set (i.e., tweets that are
attract positive responses. Viewers were willing to neither @replies nor retweets), the percentage of
share these creative contents. Live streaming genuine @replies in the data set (i.e., @replies
services enabled the event organizer to transmit that are not retweets), the percentage of retweets
delightful and festival-style event simultaneously in the data set, the percentage of tweets in the data
(Chiu & Hsiao, 2014). set that contain URLs. Vann also examined
interactive format from the team’s as well as the
6 體育學報特別議題專刊

fan’s perspectives. First, teams can initiate direct influence than in-game factors. For instance, the
conversations (@replies), the promotion of chat interactions were more intensive when two
existing message (tweets with URLS), tweets teams had won a number of their recent games.
contributing original content (original tweets), or Or the closer the team was to the top of the
the reproduction of tweets posted by someone else rankings (e.g., first- or second-ranked team), the
(retweets). Original Tweets, @replies, and more interactive the viewers were. Also, a
Retweets Original tweets were the most frequently number of interesting plays associated with the
posted type of tweet by all teams, followed by outcome of the game would initiate more chat
retweets and @replies interaction among the in-game factors. This study
Most sports teams would post new also mentioned that chat format would also be
information through Twitter pages, ensuring it is varied based upon people with different level of
an outlet where fans can gain information that is involvement. For instance, the pre-game factors
not otherwise available and enhance their were more influential for loyal fans whereas
transmedia sports experience. Second, hashtags in-game factors were influential for the causal
have been employed to make a connection among games. Such a finding initiated further research
fans. Once sport-event related hashtags are issue to analyze different types of chat
initiated, interaction among fans would begin and interactions in various online sports-viewing
lead to live experience. Vann (2014) pointed out services.
fans tweeted more frequently during breaks, such Sport event delegations
as time outs and between quarters. These breaks
Taipei Universiade in 2017 was the biggest
allow people to switch their attention from the
multi-sports event that this nation has ever been
field to social media. Undoubtedly, social media
organized in Taiwan. More than 140 delegations
play a complimentary role with traditional media
attended this event. To promote the delegations as
and help team disseminate information to a
well as Universiade in its own nations, some
further extent.
embassy offices in Taipei used its official
Ko, Yeo, Lee, and Jang (2016) addressed
Facebook to update team-related events. American
that sports fans tend to interact with others when
Institute in Taipei (AIT) and Netherlands Trade
watching sports events. Social media seem to
and Investment Office (NTIO) were very actively
play a facilitating role to make communication
promoting its delegations. The former disseminated
easier among fans. In general, chat messages
51 news articles from the end of July and Early
consist of a game ID, user ID, chat message text,
September whereas the latter issued 61 news
favorite team (i.e., the team selected when the
articles from the beginning of the August to the
user posts a chat message), and when the message
early September. Through analyzing contents from
is posted. Indeed, these contents form the base
AIT and NTIO’s Facebook pages, it found that the
for further social media analytics. They studied
majority of information prior to the opening
the chat interaction in live streaming professional
ceremony was related to the introduction of the
baseball games in Korea and identified the factors
individual sports team whereas many updates were
that affected chat interaction while viewing
game results after opening ceremony. Indeed, the
games. First, pre-game factors had a stronger
introduction of the teams received the attention
 Social media in sports marketing 7

from local media which created media coverage. to reach 5 million followers within 3 days
This means that Facebook had a positive effect to (Dewhurst, 2015).
boost media exposure. Furthermore, AIT launched To fully take advantage of social media,
a delegation quiz in its Facebook page which Korzynski & Paniagua (2016) indicated that social
received around 350 comments, the most popular media enabled athletes to play an influential role
one among all their posts. Facebook also reveal the in three aspects: Informing, interacting and
information related to community visit by USA inspiring. Informing means that athlete should be
delegations. This may demonstrate its good will to willing to share personal information or off the
the host city. Also, hashtag for both Facebook was field events to keep interest level of followers.
used. To enrich the content, both offices also Interacting addressed that athletes are allowed to
disseminated relevant photo and video. To sum up, interact with fans directly and then increase
Facebook really played a helpful role to promote its visibility or popularity in which may lead to
national delegation in the host city region as well as attract prospective sponsors. In addition, inspiring
improve the visibility of these embassy offices. stated that athletes demonstrated their positive

Athletes characters on social media and motivated or

encouraged followers to share. Korzynski &
Social media can be an effective communication
Paniagua concluded that athletes should manage
tool for athletes, including retired athletes. The
fan relations though the use of social media and
popularity of the athletes can be measured
lead to improve its market value in the long run.
through social media metrics, such as the of
Furthermore, Geurin (2016) addressed that
followers and video viewers, the amount of share
social media is increasingly common for elite
and comments. Neymar, a Brazilian professional
athletes and created several benefits for them as
soccer player who is valued at 80 million, has 67
well. In order for athletes to utilized social media
million followers on Twitter and Facebook,
practice effectively, Geurin formulated a
attributed partly to his high online efforts. He
seven-step strategic guide. This study also
participates regularly in online discussion
mentioned that use of social media enabled
(approximately 40,000 tweets on his Twitter profile)
athletes to build a personal brand, search for
and posts frequently (approximately 2,600 updates
prospect sponsors, interact with stakeholders and
on Instagram) (Korzynski & Paniagua, 2016).
deepen fan engagement.
Number one female badminton player, Tai,
decided to play for national team, rather than for
University sports
World Championship, and then obtained the gold A study by Stoldt (2012) about perceptions
medal in Taipei Universiade. Followers of her fan of athletic administrators upon influence of social
page were doubled from 150,000 up to closely media on U. S. college sport stated that social
300,000 with one month (Sports Center, 2017). In media played an important and inexpensive
addition, David Beckham, a British soccer player, communication tool with external publics and
used to be the most popular individual on help colleges developed long-term relationships
Instagram ever. This retired English captain is the with key publics. In addition, colleges were
fastest to earn one million followers in his social required to response in a rapid manner.
media debut. Also, his account is the fastest ever
8 體育學報特別議題專刊

To further understand how college sports information, product information, and information
embraced the social media, Wallace, Wilson, and related to High school Basketball League. Their
Miloch (20110) examined the NCAA study showed that in general each post would
Organizational sport page and Big 12 Conference receive 1,357 times of ”like”, 15 ”comments” and
Athlete Department pages. Their study presented 58 times of ”share.” However, followers may
several important findings. First, college sports behave somewhat differently towards the type of
used Facebook features to promote linked sources the message. The product information received the
to Web sites outside of Facebook. Links allow most followers. Tournament message seemed to
users to communicate an outside Web link to enjoy the most comment. Followers would share
consumers while they are navigating Facebook team information the most.
content. Other information in the Facebook was Sports drink is one of the most active
categorized as status update, picture, video and sponsorship in sport. Gatorade, Pepsi’s sports
notes. Secondly, Through posting plenty drink brand, created a new product line based
information related to star players, performance upon the needs of various customer segments
of the team, the athletes department attempted to after recent recession. They attempted to meet the
present its product as well as experience. Also, needs of three major segments, including the teen,
The NCAA pages emphasized upon commitment senior and fitness athletes markets through
and organization attributes in which echoed the introducing the G series, G Series Pro and G
overall brand image of student-athlete as well as Series respectively. To promote their new product
organizational mission of the NCAA brand. line, Gatorade initiated a series of white and
Thirdly, college sport departments were able to black videos that featured with celebrities, such
build lasting relationships with fans and provided as Kevin Garnett, to attract consumers of the
fans a real-time interactive environment with same lifestyle and values. Also, an ad campaign
introduction of various sporting content. was launched to educate customers to appreciate

Sponsor the ingredients of the new product. For young

athletes, this company created a social media
Nike has been a long-term sponsor for
campaign to establish a mission control with six
various properties many years and built very
screens. This allowed the corporation to deepen
strong product line in some sports, such as
consumer relations and monitor feedback of
basketball. Nike launched fanpage of Nike
customers. Gatorade Facebook receive more than
basketball Taiwan in 2010. The number of
a million “likes”, and has obtained 500, 000 new
followers reached 100,000 in 2012. The Top
followers compared to just 200,000 in the
One-Hundred fanpage report mentioned Nike was
previous year. (Datamonitor, 2011).
the only sports brand in the list. This report
Also, Do, Ko, & Woodside (2015) examined
revealed that Nike Basketball Taiwan ranked 5th in
the impact of social media in sports sponsorship
the average engagement rate. (Lee, 2013). Chang
of brand upon brand consumers’ congruity and
and Chen (2014) analyzed its fan page and found
relationship quality. The results showed that
that its content could be categorized into five
social media activities of brands’ sports
areas, including high school basketball players,
sponsorship improve its brand image. Their study
high school basketball teams, tournament
 Social media in sports marketing 9

pointed out that age as well as gender act as a As to assessing the traffic level of a social
moderating variable in the relationship between media site, marketers use the page view and
brand’s sports sponsorship in social-media unique visitor metric (Web Analytics Association,
activities and the brand-related assets. 2007). More precisely, the former is defined as
These aforementioned examples not only the amount of visits that a social media site has
presented the widely adoption of social media generated in a period of time, which also signals
network, but also demonstrated its multiple the amount of “impressions” that a social media
functionality as well as some practical site has generated in the media planning practices.
achievements. It was concluded that utilization of The unique visitor metric is defined as the
social media can play several functions, including number of visitors, which captures the capability
public relations, sales promotion, sponsorship of a social media site in “reaching” its potential
solicitation, brand image enhancement, and consumers (Sissors & Baron, 2010; Web
value-added entertainment, and new consumer Analytics Association, 2007). The two metrics
acquisition. help sports marketers to estimate the effectiveness
of their past marketing communication campaigns.

Social media analytics and Furthermore, marketers can also capture how
effective their social media sites are in attracting
business decisions potential consumers and in supporting their future
marketing communication campaigns.
With the advance of the information science
As to assessing how effective a social media
and the data analysis techniques, marketers can
site is in engaging consumers, the following
now capture consumer’s activities on social media
metrics, visit duration, bounce rate, viral rate (for
sites and make a big dataset. Then, they can get
Twitter pages), and engagement rate (for
insights from these data to support various business
Facebook Pages), are widely applied by
decisions, such as marketing communication
marketers (Clifton, 2012; Sterne, 2010; Web
campaign management, consumer’s voices probing,
Analytics Association, 2007). With the aid of
and sales predictions. In this section, the authors
information technology, marketers can track how
introduce the key metrics and analytical techniques
long do consumers stay on their social media
that help find insights for these business decisions.
sites on average, termed visit duration (Web
Marketing communication campaign Analytics Association, 2007). The longer the time
management
that consumers stays on a social media site, the
Consumer’s activities on social media sites more likely such site is effective in engaging
provide great insights for marketers for managing consumers. Yet, there are other facets about
marketing communication campaigns. More measuring customer engagement in social media.
precisely, marketers can examine the traffic and For instance, marketers also use the bounce rate
engagement level of their social media sites. metric to identify the percentage of total visits
Below address the key metrics that marketers use which are single page view based (Clifton, 2012;
when capturing these insights. Web Analytics Association, 2007). As such, a low
bounce rate indicate that a social media site has
10 體育學報特別議題專刊

attracted the right type of audiences and provided classification criteria or clustering characteristics
the right type of contents so that these audiences and then quantify these “labels”. (Aggarwal &
were engaged. Specific to Twitter and Facebook Zhai, 2012; Allahyari et al., 2017; Nisbet, Elder,
pages, marketers also calculate the the viral and & Miner, 2009). Below address the data
engagement rate to probe the level of engagement. classification and clustering techniques.
More precisely, the viral rate is calculated by As to the data classification techniques,
dividing the aggregation of the numbers of there are three major approaches, the machine
retweets by the number of followers of a twitter learning, the lexicon-based and the hybrid
page. The engagement rate is calculated by approach, which are widely used by data
dividing the sum of numbers of reactions, scientists (Beigi, Hu, Maciejewski, & Liu, 2016;
comments, and shares with the total reach of the D'Andrea, Ferri, Grifoni, & Guzzo, 2015; Hotho,
whole Facebook site/ an event page (Clifton, Nűrnberger, & Paass, 2005). The machine
2012; Sterne, 2010). learning approach involves setting a “pattern”
In summary, the aforementioned metrics help (e.g., computer learning algorithms) to predict the
capture the level of traffics and engagement that a polarity of the texts. Within such approach, data
social media site has generated. Sports marketers scientists have developed a number of methods to
can thus these insights to monitor the effectiveness predict text polarity, such as the Bayesian
of their past marketing communication performance. Networks (using probability classifiers) and the
Furthermore, they can estimate how effective their Support Vector Machine method (using linear
social media sites are in creating impressions classifiers). As to the lexicon-based approach,
(e.g., attracting active potential consumers for data analysts utilize a predefined list of words
their future campaigns) which also helps to plan (e.g., through dictionaries or corpus) as a basis to
their future marketing communication campaigns count the number of positive/ negative words in
(Clifton, 2012; Sterne, 2010; Wilson, 2004). the text so as to classify the texts. To date, both

Consumer’s voices probing the machine learning and lexicon based

approaches are tested to have satisfactory
Analyzing consumer’s comments in social
performance in text classification (Beigi et al.,
media also helps capture insights to support
2016). There are also a number of software
business decisions. Such analytical work is
packages developed based upon these methods to
termed sentiment analysis that denotes using
support sentiment analysis, such as SenticNet
programming language to classify, cluster,
(machine learning based) and Senti WordNet
quantify, and visualize the themes behind
(lexicon based) (D’Andrea et al., 2015). In the
consumer’s comments in social media (Liu, 2012).
recent development, data scientists also propose
Conducting sentiment analysis generally involves
the hybrid approach, integrating lexicon while
the following processes. Firstly, data analysts
setting of machine learning pattern, to improve
need to filter out non opinion-led statements -
the effectiveness and efficiency of the sentiment
where consumer’s voices reside. Then, they need
analysis process (Behera, Roy, & Dash, 2016;
to use programming language to classify or
Lalji & Deshmukh, 2016).
cluster the data. Lastly, data analysts label the
social media contents based upon the
 Social media in sports marketing 11

Once the texts are classified, data analysts illustrates what brand was seemed as having the
can quantify the amount of emotions in social best performance compared to its competitors by
media (e.g., the number of positive or negative consumers.
comments) in a given time period and across time In short, analyzing consumer’s comments in
periods. Such information can help brand social media sites also provides rich information
managers to monitor consumer’s voice towards for business decisions. As such, it is suggested
their brand performances. Further analyses can that there is a need for sports marketers to
also be undertaken to visualize the sentiments, systematically examine consumer’s comments
such as making percetual maps to illustrate the from their social media sites. They can apply the
strength and plearity of consumer's voices across data classification techniques to capture how their
different brands (Beigi et al., 2016; Misopoulos products or services are perceived in consumer’s
& Mitic, 2014). minds (e.g., positive or negative as well as the
Data clustering process is different from the exact brand associations). Furthermore, they can
aforementioned data classification process. More also apply the data clustering techniques to find
precisely, data clustering is to group contents the latent needs that consumers want from their
based upon similar characteristics whereas data product or services (e.g., the kind of atmosphere
classification is to use a pre-defined rule to or experiences that sports consumers are looking
categorize social media contents. As such, for when attending a match).
clustering social media data helps marketers to Sales predictions
find the latent themes within consumer’s remarks.
Consumer’s comments on social media sites,
Regarding the data clustering techniques,
once quantified through sentiment analysis, are
there have been a number of algorithms that data
found as a strong predictor to estimate a brand’s
scientists use to cluster social media contents,
future sales figures. In the marketing academic
such as the K-means clustering and the Latent
field, such agenda is termed the electronic
Dirichlet Allocation (LDA) techniques (Aggarwal
word-of-mouth (eWOM) elasticity which has
& Zhai, 2012; Allahyari et al., 2017; Nisbet et al.,
received a considerable amount of attentions
2009) and have been some innovative works in
(Bollen, Mao, & Zeng, 2011; Oliveira, Cortez, &
the academic field that demonstrate how to
Areal, 2016; Tirunillai & Tellis, 2012; Zhu &
cluster social media contents to generate insights.
Zhang, 2010). The analytical work generally
For instance, Tirunillai and Tellis (2014) analyzed
requires the following steps. Firstly, there is a
consumer reviews from multiple social media
need for data analysts to integrate various type of
platforms (e.g., cnet.com and zdnet.com) between
sources, such as a brand’s social media data, sales
2006- 2010 to probe the key factors that
figures, and marketing budgets, into a giant
constitute product satisfaction. With the use of
dataset. By doing so, data analysts can obtain the
the Latent Dirichlet allocation framework, they
key parameters for the modelling purposes (e.g.,
managed to find the latent themes regarding what
using econometrics techniques) where possible
consumers consider when making a purchase (e.g.,
exogenous variables can be controlled to increase
mobile phones and computer). Furthermore, they
research validity.
use these insights to create a perceptual map that
12 體育學報特別議題專刊

In extant marketing literature, researchers are related merchandise, sports products) is sensitive
currently investigating what type of eWOM metric to consumer’s comments from social media sites?
(e.g., on a volume or on a valence basis) best Secondly, what type of metric (on a volume or a
predicts a brand’s sales figures as well as in what valence basis) is relative more predictive to what
marketing context that metric has an higher types of sales figures is also needs to be
predictive power. The initial conclusions are as inrestigated. Lastly, there is a need for future
follows. Firstly, researchers verify that sales researchers to consider other types of endogenous
figures can be predicted either on a volume (the variables in the modelling process owing to the
amount of posts in social media sites) or on a unique feature of sports industry. For instance,
valence (the polarity revealed by consumer’s team identification and team performance could
comments) basis. Yet, there is still a discrepancy possible be two important exogenous variables
regarding which one is relative more predictive to that sports marketing researchers need to consider
sales (Rosario, Sotgiu, Valck, & Bijmolt, 2016; You, in the modelling process (Gray & Wert-Gray,
Vadakkepatt, & Joshi, 2015). Secondly, variance 2012; Wann & Branscombe, 1993). Their impacts
among consumer’s comments (e.g., consumers rate have not been systematically examined in the
a product in an inconsistent way) is found can extant marketing literature owing to the
negatively influence sales figures, in particular to uniqueness of the sports marketing industry.
those brands who are not well known to consumers.
Thirdly, product types moderate the effect of Conclusion
consumer’s social media comments on sales. More
precisely, the positive impact of eWOM on sales is The convergence of advanced information
stronger in service, hedonic and high financial technology and communication has created
risked based product categories compared to the tremendous impacts upon sports marketing and
counterparts (Rosario et al., 2016). sports consumption. Such agenda has also
These initial conclusions provide implications received a considerable amount of attentions from
both to the sports marketing academic and the research community, as such, there have been
practitioners. On the one hand, they indicate the 123 research articles centering on the agenda of
importance of social media in sports marketing social media form 29 journals in the timeframe
management as sports products or consumptions between 2008 (earliest found) and 2014 (Abeza,
by and large are hedonic based (Shank & O’Reilly, Séguin, & Nzindukiyimana, 2015).
Lyberger, 2014). As such, a successful social Extent literature suggests that social media
media sites management (e.g., having a huge sites are highly involved in contemporary sports
number of sports consumers/fans who are highly consumption practices. As to the former, there
active and generating positive messages) is likely have been around 747 million sports consumers/
to boost a sports brand’s sales figures. On the fans who are engaged in sports based social
other, they also shed a new light on sports media sites (Sportcal, 2017). Their “footprints”
marketing research. Firstly, due to a lack of can be found in various types of platforms, such
empirical evidence, it is to be investigated that as micro blog based sites, social networking
what types of sales (e.g., game attendance, based sites (e.g., Facebook), collaborative based
 Social media in sports marketing 13

sies (e.g., sports forum), and content communities The advent of analytics techniques makes
based sites (e.g., YouTube) (Kaplan and Haenlein, social media further important to sports
2010). In the era of big data, sports marketers can marketing practices. There have been a host of
now store and analysis a gigantic volume of techniques to help sports marketers to transform
consumer’s data from social media (Beyer & social media data in supporting various business
Laney, 2012; De Mauro et al., 2015; Qiu et al., decisions. More specifically, sports marketers can
2016). These data are of great values to sports analyze the traffic and engagement levels of their
marketers as they reflect consumer’s spontaneous social media sites so as to monitor and plan their
emotions, and opinions towards a sports brand in existing and future marketing communication
great details (Tirunillai & Tellis, 2014). campaigns. In addition, they can conduct
As to how social media impact sports sentiment analysis to probe how consumers feel
marketing practices, their roles in supporting about their products and services (e.g., positive or
promotional activities for sports sections are well negative, the classification process) as well as to
recognized. More precisely, sport organizations, discover consumer’s latent needs (e.g., what
including national governing bodies and some makes consumers engaged during the game, the
non-mainstream professional sports leagues, view clustering process). Lastly, the quantified results
social media as a very important publicity tool to of consumer’s comments in social media, such as
serve as a bi-lateral communication tool with the amount of posts and the polarity of the
their stakeholders. For instance, owing to a high comments, also supports predicting future sales
level of transmission capacity, live broadcasting figures (e.g., Bollen et al., 2011; Oliveira et al.,
on social media such as Youtube and Facebook 2016; Tirunillai & Tellis, 2012; Zhu & Zhang,
live streaming, are now a useful tool to attract the 2010). Recent research reveals that the amount of
attentions from consumers. In addition, social posts made by consumers as well as the sentiment
media are a very useful tool for sports resides in consumer’s comment positively predict
sponsorship - professional sports teams utilize a brand’s sales figures. Such predictive power is
social media as part of their sponsorship particular strong in service, hedonic and high
campaigns. Lastly, in light of real-time financial risked based industries - where sports
interaction, it has found that many sports teams commodities and services operate (Rosario et al.,
use social media to encourage sports fans to 2016; Shank & Lyberger, 2014; You et al., 2015).
express their voices, such as through Twitter and This study also proposes several
Facebook with the uses of hashtags. This design implications for sports marketing researchers and
would allow sports organizations to collect and practitioners. As to the former, firstly, despite
analyze the feedback from consumers, where that sports organizations in Taiwan have already
these insights can be further applied to for deployed social media in their practices, there is
customer relationship management and product a need of further research to investigate the
development. In addition, hashtags, a unique objectives as well as motivations of using social
function in social media, helps setting keywords media in professional teams sports as well as
to allow consumers to invite others to join national sports governing bodies. Secondly, as
discussion. social media are now an important tool for
14 體育學報特別議題專刊

sponsorship inventory, the authors also argue [Chiu, S. T., & Hsiao, C. H. (2014). Viral
marketing campaign on the Facebook fan
there is a need for further research to investigate
page of the sport event. Quarterly of Chinese
how they can facilitate the sponsor’s promotional Physical Education, 28(1), 11-20.]
campaigns as well as measure the impacts that
these campaigns generate. Thirdly, to understand 張家瑋、陳光宗 (2014)。HBL 三月瘋： Nike
the effectiveness of social media is an essential basketball Taiwan 社群內容效益分析。成大
體育學刊， 46(2)，36-51。
issue while social media practice is prevalent.
[Chang, C. W., & Chen, K. T. (2014). HBL march
This means that more research efforts should be
madness: Social content performance of
spent is this realm. Fourthly, the authors also Nike basketball Taiwan. Journal of NCKU
suggest sports marketing researchers to further Physical Education Research, 46(2), 36-51.]
capture the impact of endogenous variables (e.g.,
team identification and team performance) in the 體育中心 (2017)。戴資穎臉書粉絲狂奔 超越這
位臺灣運動巨星。取自 http://sports.ltn.com.
way they use social media data to predict a tw/news/breakingnews/2192049
brand’s sales performance sales, which is
[Sports Center (2017). Her overwhelming
currently understudied in the contemporary popularity brought her Facebook top in fan
literature. number. Retrieve from http://sports.ltn.
com.tw/news/ breakingnews/2192049.]
As to the implications to sports marketing
practitioners, the suggestions are as follows. First,
鑾斌、邱于平、楊芢潔、郭致妘 (2014)。全球
the effectiveness of a social media site is based 品牌企業之社群媒體應用分析， 電子商務
upon its contents. As such, sports marketers need 研究， 12(2)，121-142。
to offer customized contents, with the use of [Luarn, P., Chiu, Y. P., Yang, J. C., & Kou, C. Y.
hashtags, to sports consumers so as to increase (2014). The analysis of social media
application on global brands enterprise.
the level of traffics and engagement. Secondly,
Electronic Commercial Studies, 12(2),
sports marketers need to develop a monitor 121-142.]
mechanism in responding consumer’s voices in
social media which is essential to create a high Abeza, G., O'Reilly, N., Sėguin, B.,
level of customer satisfaction and fan relations. Nzindukiyimana, O., (2015). Social media
scholarship in sport management research: A
critical review. Journal of Sport
References Management, 29, 601-618.
Aggarwal, C. C., & Zhai, C. X. (2012). Mining
李育璇 (2013)。2012 年企業粉絲團 100 強 。取 Text Data: New York, NY: Springer.
自 https://www.bnext.com.tw/article/25959/ Allahyari, M., Pouriyeh, S., Asse, M., Safaei, S.,
BN-ARTICLE-25959 Trippe, E. D., Gutierrez, J. B., & Kochut, K.
[Lee, Y. H., (2013). Top 100 fan pages in (2017). A Brief Survey of Text Mining:
corporations report. Retrieved from Classification, Clustering and Extraction
https://www.bnext.com.tw/article/ Techniques. CoRR, abs/1707.02919
25959/BN-ARTICLE-25959] Behera, R. N., Roy, M., & Dash, S. (2016).
Ensemble based hybrid machine learning
邱仕騰、蕭嘉惠 (2014)。病毒行銷於 Facebook approach for sentiment classification- A
運動賽事粉絲專業之應用。 中華體育季 Review. International Journal of Computer
刊， 28(1)，11-20。 Applications, 146(6), 31-36.
 Social media in sports marketing 15

Beigi, G., Hu, X., Maciejewski, R., & Liu, H. Do, H., Ko, E., Woodside, A. G., (2015). Tiger
(2016). An overview of sentiment analysis in Woods, Nike, and I are (not) best friends:
social media and its applications in disaster how brand's sports sponsorship in
relief. In W. Pedrycz & S. M. Chen (Eds.), social-media impacts brand consumer's
Sentiment Analysis and Ontology congruity and and relationship quality.
Engineering: An Environment of Journal of Advertising Research, 34(4),
Computational Intelligence (pp. 313-340). 658-677.
New York, NY: Springer
Eagleman, A., N. (2013). Acceptance,
Beyer, M. A., & Laney, D. (2012). The motivations, and usage of social media as a
Importance of “Big Data”: A definition. marketing communications tool amongst
Gartner, G00235055. employees of sport national governing
bodies. Sport Management Review, 16,
Billings, A. C., Burch, L. M., & Zimmerman, M.
488-497.
H. (2015). Fragments of us, fragments of
them: social media, nationality, and US Geurin, A. N., (2016). Developing a social media
perceptions of the 2014 FIFA World Cup. strategy guide for elite athletes. Sport &
Soccer & Society, 16(5/6), 726-744. Entertainment Review, 2, 70-76.
Bollen, J., Mao, H., & Zeng, X. (2011). Twitter Gray, G. T., & Wert-Gray, S. (2012). Customer
mood predicts the stock market. Journal of retention in sports organization marketing:
Computational Science, 2(1), 1-8. doi: Examining the impact of team identification
https://doi.org/10.1016/j.jocs.2010.12.007 and satisfaction with team performance.
International Journal of Consumer Studies,
Boyd, D. M., & Ellison, N. B. (2008). Social
36(3), 275-281. doi: 10.1111/j.1470-6431.
Network Sites: Definition, history, and
2011.00999.x
scholarship. Journal of Computer-Mediated
Communication, 13, 210-230. Hochberg, M., (2015). Upperiscope. Sport
Business International, 212, 18-19.
Clifton, B. (2012). Advanced Web Metrics with
Google Analytics: Idianapolis, Indiana: John Hotho, A., Nűrnberger, A., & Paass, G. (2005).
Wiley & Sons. ABrief Survey of Text Mining. Journal for
Computational Linguistics and Language
D'Andrea, A., Ferri, F., Grifoni, P., & Guzzo, T.
Technology, 20, 19-62.
(2015). Approaches, tools and applications
for sentiment analysis implementation. International Event Group. (2015). How the Los
International Journal of Computer Angeles Dodgers use Instagram to extend
Applications, 125, 26-33. doi: 10.5120/ sponsor reach. Retrieve from
ijca2015905866 http://www.sponsorship.com/iegsr/2015/02/
23/How-The-Los-Angeles-Dodgers-Use-
Datamonitor (2011). Gatorade case study: Using
Instagram-To-Exten.aspx
consumer segmentatioin & social media to
drive market growth. UK: Author. International Event Group. (2017a). Inside
Wimbledon’s social media strategy. Retrieve
Dewhurst, A. (2015). It’s official: David
from http://www.sponsorship.com/iegsr/
Beckham is the most popular person on
2017/04/10/Inside-Wimbledon-s-Social-
Instagram EVER. Grazia. Retrieved from
Media- Strategy.aspx
http://www.graziadaily.co.uk/2015/05
International Event Group. (2017b). Three
De Mauro, A., Greco, M., & Grimaldi, M. (2015).
examples of how Wimbledon tailors content
What is big data? A consensual definition
for snapchat, periscope and Instagram.
and a review of key research topics, AIP
Retrieve from http://www.sponsorship.com/
Conference Proceedings 1644, 97.
IEGSR/2017/04/10/Inside-Wimbledon-s-
doi.org/10.1063/1.4907823
Social-Media-Strategy/Three-Examples-Of-
How- Wimbledon-Tailors-Content-Fo.aspx
16 體育學報特別議題專刊

Kaplan, A. M., & Haenlein, M. (2010). Users of Rosario, A. B., Sotgiu, F., Valck, K. D., &
the world, unite! The challenges and Bijmolt, T. H. A. (2016). The effect of
opportunities of Social Media. Business electronic word of mouth on sales: A
horizons, 53(1), 59-68. doi: doi.org/ meta-analytic review of platform, product,
10.1016/j.bushor.2009.09.003 and metric factors. Journal of Marketing
Research, 53(3), 297-318. doi: 10.1509/jmr.
Korzynski, P,. & Paniagua, J. (2016). Score a
14.0380
tweet and post a goal: Social media recipe
for sports stars. Business Horizons, 59, Shank, M. D., & Lyberger, M. R. (2014). Sports
185-192. marketing: A strategic perspective (5 ed.).
New York, NY: Routledge.
Ko, M., Yeo, J., Lee, J., L, U., & Jang, Y. J.
(2016). What makes sports fans interactive? Siguencia, L., O., Herman, D., Marzano, G., &
Identifying factors affecting chat interaction Rodak, P. (2017). The role of social media
in online sports viewing. PLoS ONE, 11(2), in sports communication management: An
e0148377. doi: 10.1371/journal.pone.0148377 analysis of polish top league team’s strategy.
Procedia computer science, 14, 73-80.
Lalji, T. K., & Deshmukh, S. N. (2016). Twitter
sentiment analysis using hybrid approach. Sissors, J. Z., & Baron, R. B. (2010). Advertising
International Research Journal of Engineering media planning (7 ed.) New York, NY:
and Technology, 3(6), 2887-2890. McGraw Hill.
Lee, R. S. T., Liu, J. N. K., Yeung, K. S. Y., Sin, Sportcal. (2017). Sportcal's social media index.
A. H. L., & Shum, D. T. F. (2009).
Statista. (2017). Most famous social network sites
Agent-Based Web Content Engagement Time
2017, by active users. Retrieved from
(WCET) Analyzer on e-Publication System.
https://www.statista.com/statistics/272014/
2009 Ninth International Conference on
global-social-networks-ranked-by-number-of
Intelligent Systems Design and Applications,
-users/
Pisa. doi: 10.1109/ISDA.2009.189
Sterne, J. (2010). Social media metrics.
Liu, B. (2012). Sentiment Analysis and Opinion
Indianapolis, Indiana: John Wiley & Sons.
Mining: San Rafael, CA: Morgan &
Claypool Publishers Stoldt, G. C. (2012). The impact of social media
on college athletics communications
Misopoulos, F., & Mitic, M. (2014). Uncovering
Retrieved from http://www.cosida.com/
customer service experiences with Twitter:
media/documents/2012/7/Social_media_
The case of airline industry. Management
impact_2012_Stoldt_study.pdf.
Decision, 52(4), 705-723. doi: 10.1108/
MD-03-2012-0235 The Telegraph. (2017). How Facebook dominates
the internet - in five numbers. Retrieved
Nisbet, R., Elder, J., & Miner, G. (2009).
from http://www.telegraph.
Handbook of Statistical Analysis and Data
co.uk/technology/2017/02/02/facebook-
Mining Applications. San Diego, CA:
dominates-internet-five-numbers/
Academic Press.
Tirunillai, S., & Tellis, G. J. (2012). Does chatter
Qiu, J., Wu, Q., Ding, G., Xu, Y., & Feng, S.
really matter? Dynamics of user-generated
(2016). A survey of machine learning for big
content and stock performance. Marketing
data processing. EURASIP Journal on
Science, 31(2), 198-215. doi: 10.1287/mksc.
Advances in Signal Processing, 67.
1110.0682
doi: 10. 1186/s13634-016-0355-x
Tirunillai, S., & Tellis, G. J. (2014). Mining
Oliveira, N., Cortez, P., & Areal, N. (2016).
marketing meaning from online chatter:
Stock market sentiment lexicon acquisition
Strategic brand analysis of big data using
using microblogging data and statistical
Latent Dirichlet Allocation. Journal of
measures. Decision Support Systems,
Marketing Research, 51(4), 463-479.
85(Supplement C), 62-73.
doi: 10.1509/jmr.12.0106
doi: doi.org/10.1016/j.dss.2016.02.013
 Social media in sports marketing 17

Vann, P. (2014). Changing the game: The role of Wilson, R. D. (2004). Using web traffic analysis
social media in overcoming old media’s for customer acquisition and retention
attention deficit toward women’s sport. programs in marketing. Services Marketing
Journal of Broadcasting and Electronic Quarterly, 26(2), 1-22.
media, 58(3), 438-455.
You, Y., Vadakkepatt, G. G., & Joshi, A. M.
Wallace, L., Wilson, J., & Miloch, K. (2011). (2015). A meta-analysis of electronic word
Sporting facebook: A content analysis of of mouth elasticity. Journal of Marketing,
NCAA Organizational sport page and Big 12 79(2), 19-39. doi: 10.1509/jm. 14.0169
Conference Athlete Department pages.
Zhu, F., & Zhang, X. (2010). Impact of online
International Journal of Sport
consumer reviews on sales: The moderating
Communication, 4, 422-444.
role of product and consumer characteristics.
Wann, D. L., & Branscombe, N. R. (1993). Sports Journal of Marketing, 74(2), 133-148.
fans: measuring degree of identification with doi: 10.1509/jmkg.74.2.133
their team. International Journal of Sport
Psychology, 24(1), 1-17.
Web Analytics Association. (2007). Web analytics
definitions - Version 4.0. Retrieved from 投稿日期：106 年 11 月
https://www.digital analytics association. 通過日期：106 年 12 月
org/Files/pdf_standards/webanalyticsdefiniti
onsvol1.pdf
18 體育學報特別議題專刊
Physical Education Journal
Society of Physical Education, Taipei, Taiwan
Vol. 50, No. S, pp. 1～18 (2017. 12.)

社群媒體在運動行銷應用之探討
黃 煜 1、徐任賢 2
1
臺灣 新竹市 國立清華大學體育系
2
英國 德蒙特福特大學萊斯特商學院

摘 要
緒論：近年社群媒體的無所不在讓多企業感受到其魅力，運動行銷人員也認同其價
值並應用在許多層面，同時，大數據時代的來臨也提供社群媒體內容分析的技術已進一
步了解使用者的社群媒體的消費行為，針對此，本文試圖分析社群媒體在運動產業較常
見的應用方式以及探討社群媒體可應用於運動商業領域的社群媒體分析技術。方法：本
文主要是透過相關文獻回顧及探討。結果：顯示社群媒體可以提供多元的功能，首先，
社群媒體是促銷媒介可擴大宣傳成效。第二，提供即時傳播服務以提升內容娛樂價值。
第三，雙向互動可彙整消費者建議。第四，社群媒體可納入贊助權益及活化規劃元素之
一。另外，社群媒體有效地群聚消費者。針對社群媒體內容分析技術部分，本文建議運
動行銷經理人針對其所經營之社群網站進行流量 (traffic analysis)、參與度 (engagement
analysis) 以及情感分析 (sentiment analysis)。分析結果將有助於管理行銷溝通成效、了解
消費者的意見氣候以及預測未來銷售。結論：基於社群媒體使用族群龐大，許多運動組
織也紛紛提出行銷相關應用，而消費者在使用社群媒體時產生了巨量資料，這些巨量資
料形成的大數據分析的基礎，另藉由現今資訊分析的技術可以探討社群媒體所產生的資
料並作為了解消費者使用行為及預測分析的依據。最後，本文也提出學術界及實務界之
建議事項。

關鍵詞：社群媒體、社群媒體資料分析、大數據