Superboosting the athlete social media brand: events as an opportunity for follower growth

ABSTRACT

Research Question

An athletes’ social media following is a proxy of their popularity and a key metric for brand monetization. Yet, how a following can be grown strategically remains unclear. This research investigates the effects of newly formed brand networks on athlete follower growth during a non-league event with representative teams. We used the sport brand ecosystem framework and examined athlete-related, event-related, and brand-networking-related factors as determinants of follower growth on Instagram.

Research Methods

We collected longitudinal behavioral data, namely social media following and tagging behavior of athletes in the context of Laver Cup, an elite men’s team tennis event. A sociogram was used to visualize brand networking of athletes and the event. The hypotheses were tested using a multiple linear regression with a wild-cluster bootstrap-SE.

Results and Findings

Results indicated that the pre-existing size of an athlete’s following and brand networking with athletes’ and the event’s brands through the user tagging function predicted follower growth. This highlights the impact of exposure on social media during an event and the value of brand networking as a brand-building strategy for athletes.

Implications

The findings contribute knowledge on athletes’ vertical and horizontal brand relationships. The study uncovers coopetitive relationships between athlete brands and shows that new brand networks, visible through social media user tagging, spur athlete brand growth. To practitioners, this demonstrates that events enable athletes to strengthen their social media brands, which can be amplified through athletes’ large pre-existing social media following and strategic collaborations with other athletes.

KEYWORDS:

• Athlete brand
• brand architecture
• brand relationships
• social media
• digital marketing

Acknowledgment

The authors are grateful for the support from the Sport Industry Research Center at Temple University, Philadelphia, PA and collaboration with the Laver Cup. All data collected for this research were publicly available and no proprietary data were used. The authors are also grateful to two anonymous reviewers for their helpful feedback .

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 VIFs for all explanatory variables were under 1.809 across models. VIFs for controls TEAM and RANKING were 1.297 and 1.403 when introduced separately, and 4.090 and 4.426 when introduced together.

2 After conducting the multiple linear regression with the wild-cluster bootstrap-SE, we compared the results with CRSE estimation. In terms of significance/insignificance of estimates, results were similar although there were slight differences in estimates of standard errors and regression coefficients. We base our discussion on the results of the wild-cluster bootstrap-SE procedure, which is considered more robust in such a setting (Cameron et al., 2008).