What follows is an example of the types of analyses that can be performed with survey data. The example is provided primarily to aid my class at Emory University. However, I decided to share it with the broader world because the topic, the structure of influence, and the output, a classification scheme for sports influencers, is relevant to sports and fandom-oriented businesses. The analysis uses data from the Atlanta Sports Survey. I’m doing this one real quick, so please ignore any typos.
Social influence and influencer marketing have become essential concepts for marketers. Marketing, especially marketing of culturally important products, has always been about influence. What has changed is that social media and information technology now make it increasingly possible to observe and measure influence.
Also streaming on Apple Podcasts, Spotify, and Stitcher.
In some respects, Fandom and Social Influence are closely linked. Fandom is about passion and involvement with a cultural entity. Social influence is about how a cultural entity affects people. Influence and fandom are the two sides of the same coin. Organizations that want to build extreme passion (i.e. fandom) need to influence consumers into becoming committed members of a brand community.
Being a fan of the Dallas Cowboys, Star Wars, Dua Lipa, or Donald Trump (or Barack Obama) can be an important part of an individual’s identity. Being a fan of something is more than just appreciating and athletic performance or enjoying the action in a film. Fandom is also about being part of a community. A Georgia Bulldog fan can make a friend by barking at a stranger wearing a UGA hat. The communities may vary in intensity but they all have one thing in common – membership is voluntary, and people want to belong.
How do you motivate people to want to belong to something? One short answer is influence. Have a trusted advocate for the organization. Have a spokesman that people want to listen to too. Have an attractive promoter that people want to emulate. Have influential people and organizations as supporters for your fan community.
This is especially true for marketers in the categories driven by cultural and societal trends. Cultural products (sports, entertainment, fashion, and even politics) are very often linked to consumers’ social identities.
This post investigates the structure of influence for different categories of sports fans. The investigation uses data from the Atlanta Sports survey (a survey of sports fandom across the Atlanta market). This survey includes questions about preferred sources of information, fandom, and demographics.
The plan from here is to start by first examining the types of people and platforms fans want to access to keep up with sports. Second, we will consider a couple of ways of segmenting fans. The third step (the big one) is to examine patterns of preferences for influencers to come up with categories or clusters of influencers. The fourth step is to link fan segments to preferences for influencer types.
Influencer / Information Preferences
Our focus is on where sports fans want to get sports content and commentary. Wanting information is a crucial property of being an influencer. As a starting point Figures 1 and 2 show the percentage of respondents interested in different platforms (Figure 1) and people (Figure 2). These figures show a small subset of the media and people rated by respondents.
Figure 1 shows the percentage of survey respondents interested in getting sports information from ESPN, local sports radio, Facebook, Twitter, and Instagram. The immediate takeaway is that ESPN continues to be the dominant source of sports content. Sports radio is a more important source than social platforms. The social media platforms are each used by about 40% of the surveyed sports fans. We provided this comparison between traditional and social media to highlight the continuing relevance of the established channels. ESPN and sports radio are still relevant for sports fans. The crucial question is, of course, how this might vary by age.
Figure 1
Figure 2 shows a similar analysis for a selected group of personalities (influencers). The group includes Dave Portnoy, Steven A Smith, Nate Silver, Megan Rapinoe, and Shaquille O’Neal. Of this group, Shaquille O’Neal has by far the greatest appeal at 40%. Steve A Smith follows at 23%. The other three are cited by 10% or less. As before, we are showing a small number of fairly different types of personalities to illustrate preferences across exemplars rather than just providing a list of the most popular celebrities.
Figure 2
The celebrities we used vary on multiple dimensions. Rapinoe is the only woman, the only soccer player, and she is very politically active. Steven A Smith is an African American and a featured performer on ESPN. Nate Silver is a statistician who analyzes sports and politics. Shaquille O’Neal is an all-time great athlete, a veritable giant, and one of America's most popular product endorsers. Dave Portnoy is a media entrepreneur who runs a relatively new and outspoken sports and culture website. He also does pizza reviews. The essential point is that there are lots of dimensions on which we can compare influencers. The even more essential point is that we ideally want to know the dimensions that fans consider.
These figures show raw popularity across the broader sports fan population. The results are roughly as expected as the more established platforms and personalities have higher interest levels. However, there is often an interest in attracting a specific segment of fans rather than the mass market.
Sports Fan Segments
Consumers and fans can be segmented in a wide variety of ways. The classic approach to segmentation is to use demographics such as gender or age. Alternatively, segmentation can be based on behaviors or preferences. We investigate two segmentation schemes: Gender and Sport Interests.
Table 1 shows interest in various information sources based on gender. Specifically, Table 1 shows the percentage of each gender that expresses an interest in information from ESPN, Twitter, Steven A. Smith, Shaquille O’Neal, Joe Rogan, and Paul Finebaum. This list of “Influencers” is a subset of the questions about information sources and preferences.
Table 1
ESPN is the most cited information source at 83.7% for men and 77.3% for women. Male sports fans also have more interest in Twitter. In terms of personalities, male fans are far more interested in Paul Finebaum and Joe Rogan, while female respondents are more interested in Steven A Smith. Shaquille O’Neal has the same appeal to men and women. This set of channels and influencers is small, but it illustrates the differences and similarities in information preferences across the sexes. It highlights the importance of considering segment differences when thinking about influencers.
Table 2 repeats the analysis but uses fandom for teams in MLB, the NBA, the NCAA, the NFL, MLS, and the WNBA as the basis for segmentation. The table shows significant variation across types of fans. Some of the results are consistent with influencer expertise as NBA fans most prefer Shaquille O’Neal, and Paul Finebaum has substantial strength with NCAA fans.
Other results begin to reveal less obvious insights. ESPN and Steven A Smith do exceptionally well with NBA and WNBA fans. This makes sense, given the attention that ESPN devotes to the NBA. However, NFL fans and MLS fans have similar interests in ESPN as an information source despite the vast differences in football and soccer coverage.
Table 2
The universality of influencer appeal varies considerably. The controversial Joe Rogan scores the best with MLB and NCAA fans but has little appeal to WNBA fans. Shaquille O’Neal has powerful appeal across all fan types. Paul Finebaum’s appeal to baseball fans suggests a correlation between baseball and college football fandom.
This type of data provides insight into how to identify appropriate influencers for different fan segments. If the target audience is female fans, Paul Finebaum would be a poor choice. O’Neal is an ideal choice for the mass market as his appeal is balanced and high across all groups. At the sport fandom level, Steven A Smith has consistent appeal across men’s and women’s basketball, while Joe Rogan has about 5 times the appeal for NBA fans than for WNBA fans.
A challenge with this type of analysis is that it is cumbersome and difficult to extend beyond specific data points. With many personalities or fan segments, the list of positive and negative relationships can quickly become excessive. It is also difficult to generalize the findings. For example, while Shaquille O’Neal does very well with basketball audiences, we may struggle to explain why he does well? It might be because these audiences love former players, NBA studio show hosts, or funny giants. We don’t know how to extend the findings to other possible influencers.
Influence Clusters
The next stage of the analysis involves creating clusters of influencers. These clusters are based on correlations in individual’s preferences for information channels (ESPN, sports radio, podcasts, etc.) and celebrities (Steven A Smith, Colin Cowherd, Joe Rogan, Michael Wilbon, etc.). We use some statistical magic to determine groups of influencers that tend to have similar patterns of being liked (or disliked) across survey respondents. However, clusters can also be formed by carefully looking at correlation patterns.
The results from this analysis provide insights into how sports fans think about the universe of influencers. Humans use categories to help process information, and celebrity influencers and platforms are no exception. We identify 5 clusters of influencers and describe them below.
1) Legacy Sources:
The “Legacy Sources” cluster features established channels and prominent media figures. ESPN is the preferred platform for this group, and they are interested in celebrities like Charles Barkley, Shaquille O’Neal, Jemele Hill, Mike Greenberg, and Skip Bayless. The celebrities are individuals that have been featured for decades in prominent media outlets. Members of this segment are uninterested or negatively inclined towards podcasts, Reddit, and figures like Dave Portnoy and Joe Rogan. I almost named this one mainstream sources, but the mainstream is a moving target.
2) Bro Culture:
The “Bro Culture” cluster of influence features newer voices and content often at odds with mainstream narratives. The preferred media sources include Barstool Sports and Bleacher Report. The individual commentators include Dave Portnoy, Joe Rogan, and Clay Travis. This group does not like Jemele Hill or Megan Rapinoe. The label “Bro Culture” may capture the tone of this group, but this could have also been labeled the “Counter Culture” segment.
3) Old School Media:
The Old School Media cluster is trapped in the 1990s. This group is interested in information from sports radio, the local newspaper, network television, and local sports news. They also have an interest in long-time sports personalities like Charles Barkley and Colin Cowherd, and a slight dislike for controversial current figures like Colin Kaepernick, Joe Rogan, and Megan Rapinoe. This group yearns for a simpler, less controversial, and less celebrity-driven world of sports.
4) Modern News:
The Modern News cluster utilizes modern channels and is interested in topical figures. The preferred information channels include Twitter, Reddit, and podcasts. The most influential personalities include Bill Simmons, Nate Silver, and Megan Rapinoe. This cluster disdain the local news, Steven A. Smith, and Joe Rogan.
5) Social Media & Justice:
The Social Media & Justice cluster is the closest thing to a “woke” cluster. This group prefers fun (Instagram and Facebook) rather than combative social media (Twitter). They are also interested in personalities like Colin Kaepernick, Megan Rapinoe, and Jemele Hill. This group's dislikes are especially instructive. Sources like talk radio and network TV are avoided. The group also dislikes a wide array of personalities ranging from Nate Silver to Mike Greenberg.
Figure 3 shows the relative preference for the 5 Influence Types across the population of respondents. The most popular influence cluster is the “Old School Media” category. Social Media & Justice is the second-ranking cluster, and “Bro Culture” is the third. This structure highlights the challenges facing sports as the largest cluster is popular among an aging cohort of fans. In contrast, the second- and third-ranked clusters are the most politically charged groups of influencers.
Figure 3
Mapping Fans to Influencer Clusters
Next, we map the influence clusters to fan segments. We first examine how influencer preferences vary with age. Table 3 shows the cluster preferences for the 44 and under age group and the 45 and over age group.
The age-based segmentation reveals significant generational differences. The 44 and under group is more political and less interested in established personalities and channels. The younger fans are much more inclined to use social media. Teams face the challenge of appealing to younger fans with opposed preferences in influencer types. Young fans are influenced by polarizing figures like Joe Rogan, Dave Portnoy, Megan Rapinoe, and Jemele Hill.
Table 3
There are also significant differences across segments defined by sports fandom. Figures 4 and 5 show the distribution of cluster preferences for fans of the WNBA and college football. The WNBA figure reveals a strong preference for legacy sources of information like ESPN and a strong interest in social justice-oriented influencers. In contrast, these fans have little interest in “Bro Culture” type influencers. The NCAA football figure reveals much less interest in Legacy Sources and a much stronger interest in “Bro Culture”. However, NCAA football fans have an even stronger interest in Social Media & Justice influencers.
Figure 4
Figure 5
Breakdowns for the other sports also reveal substantial differences in influencer preferences. For example, Baseball has the highest percentage of “Bro Culture” and “Old School Media”. The NFL’s largest segment is the “Modern News” segment. The NBA’s largest segment is the “Legacy Sources” cluster.
The analysis could continue as we could look at a combination of fandom and age. Or start to bring in other segmentation schemes.
Last Word
The growth of social platforms has transformed marketing by highlighting the vast appeal or influence of certain personalities. The goal today was to lay out a path to analyzing how sports fans are influenced. The critical part is the clusters of influencer types. The mapping of these influencer clusters to fan segments is meant to illustrate the possibilities rather than provide a comprehensive treatment of influence structure in sports.
Much more is possible. We could segment the fans into myriad segments based on demographics and behaviors – age, income, game attendance, social media followings, etc. We could then map these segments to the influencer clusters. Alternatively, we could profile the different clusters by finding the types of people that prefer each type of influencer.