An examination of social media operations and behaviors in the absence of the core sport product
by Michael L. Naraine, Brock University
While sport management professionals have embraced social media as a vehicle to stimulate engagement amongst fans, it is unknown what that engagement consists of when there is turbulence in the sport industry that causes a complete shutdown of the core product. Scraping social media data produced by all 30 National Basketball Association teams during the core product pause caused by the COVID-19 pandemic, the findings revealed teams delivered their social media posts in suboptimal periods, contrary to when fan engagement was at its highest. Specifically, engagement was at its highest during the traditional hours of a workday and in the middle of the work week, contrary to the belief social media enhances the sport product in the nighttime, weekend periods. The findings of this study suggest social media operations refine their strategy to capitalize on when fans seek to engage.
On March 11th, 2020, Utah Jazz center, Rudy Gobert, tested positive for the novel coronavirus (COVID-19), an event which arguably spurred a cataclysmic wave across professional and amateur sport in North America (Golliver, 2020).
Prior to the positive test, there were initial activities erring on the side of caution, such as prohibiting spectators at the collegiate sport level, informing patrons of the precautions and risks associated with attending professional ice hockey, and even outright cancelling major tennis tournaments (Sheinin, 2020).
However, the confirmation of a professional athlete in one of the major professional sports in North America contracting the virus was enough to push “pause” on all activities, allowing sport managers, public health professionals, and local politicians time to assess the situation.
While COVID-19 created this unique circumstance, there remained an opportunity for fans to connect with sport teams and leagues during the pause using social media.
Specifically, social media enabled fans to share (sport) content with their personal networks (e.g., Wakefield & Bennett, 2018), dialogue with other sport stakeholders (e.g., Yan et al., 2019), and like content emanating from the teams they follow (e.g., Naraine, 2019). Collectively, these social media engagement characteristics – sharing, commenting, and liking – were not only possible while the sport product is “on,” but they were also possible during the turbulent period caused by COVID-19.
As Mastromartino et al. (2020) explained, although the global health pandemic abruptly halted live sport, there was scope for sport marketing practitioners to pivot towards adjusted engagement strategies via social media and strengthen the connection with fans.
However, it is unclear how social media content is disseminated by sport teams during such periods of turbulence and, along this vein, what types of engagement figures are derived from fans to said activity.
In other words, when there is “turbulengagement,” the portmanteau of turbulence and engagement, in sport, it is unknown how sport entities operationalize their social media and, subsequently, how fans respond.
Accordingly, the purpose of this study was to uncover how social media is operationalized in a period of turbulence with no core sport product, and how fans engage thereafter.
Following the narrative of the Gobert positive test at the beginning of the COVID-19 pandemic in North America, this study examined the social media accounts of all 30 National Basketball Association (NBA) teams in the aggregate, as well as the corresponding social media engagement metrics (i.e., comments, likes, and shares) for contrast, ceasing its analysis on July 8, 2020 when teams resumed activities at the Disney World “bubble”.
While the COVID-19 situation can be viewed as an anomaly for sport (Byers et al., 2021), it is critical to understand social media activities during this period (Mastromartino et al., 2020).
For one, the pandemic continues to impact sport operations and there remains the potential for no sport and its entertainment alternatives (e.g., concerts, movies) as further health restrictions are imposed. Secondly, sport marketing practitioners should not rely on the sport product always being there, as labor stoppages, terrorist attacks, or other major events (e.g., war, natural disasters), in addition to offseason period lulls, could impact whether sport is able to proceed on a large-scale.
Therefore, looking at turbulengagement can help sport marketers reflect on their operationalization of social, and develop enhanced approaches to online fan engagement (Naraine & Karg, 2019).
One of the opportunities for sport marketing practitioners is simply using social media to connect with fans. Social media purports an egalitarian structure which is fundamentally different than the traditional communication paradigm (Peters et al., 2013).
Prior to 2010, the many sport organizations would typically communicate to consumers through mediated points such as through written press releases, radio interviews, and televised news conferences (Pedersen et al., 2021).
This rigid, one-way communication system did not allow for fans to communicate directly with leagues, teams, or athletes. Conversely, social media has broken down that mediated barrier, allowing consumers to engage directly and providing insight into levels of fan interactivity.
Yet, despite this development, there still exists significant challenges about interactivity on social media, especially as it relates to fan engagement, unbeknown to sport practitionersAchen et al., 2018
For instance, connecting with millions of followers is a difficult proposition when considering the entire group as a monolith with no discernable differences in demographic and behavioral attributes.
This highlights the contribution of Naraine’s (2019) work which segmented followers of professional sport teams’ social media. There, the social media followers of several professional sports teams were classified into upwards of 20 groups, including (but not limited to) college students, video game enthusiasts, and mothers.
In another example, Naraine and his colleagues (2019) mapped the temporal activity of professional sport teams’ fans on social media, emphasizing the desire for fans to connect with teams beyond time periods when games are live and in action.
Specifically, they determined that fan engagement was at its peak during the morning and afternoon commutes, periods when consumers seek gratification to overcome these lulls in their day.
This is especially important for teams who seek to learn more about turbulengagement, and whom are unclear as to what decisions would yield optimal results (Mastromartino et al., 2020).
To unpack turbulengagement in sport, social media data from all 30 NBA teams were extracted using MVPindex, an online analytics platform. All social media posts on Facebook, Instagram, Twitter, and YouTube (FITY) were collected for each team from March 11, 2020 to July 8, 2020 inclusive, the period in which the core sport product was paused and turbulence existed in the sport environment through COVID-19.
Once identified on MVPindex, the data that were delineated to the NBA teams and in the given timeframe were downloaded into individual CSV files, and then parsed as one overarching database for an aggregate analysis. Specifically, the dataset contained descriptive (e.g., platform) and temporal (e.g., day, month) elements, adhering to the social media analysis principles outlined by Brooker et al. (2016).
For the latter, specific time of day was captured in UTC and then converted to Eastern Standard Time (the most frequent time zone in the NBA). Furthermore, the dataset also contained the engagement variable, the metric composed of likes, shares, and comments on social media posts, which were also analyzed in the aggregate.
This procedure of scraping social media data has been embraced in sport management as a legitimate form of data collection (Watanabe et al., 2021), the aggregated analysis of which can yield important insights for practitioners (Naraine et al., 2019).
In total, there were 62,434 FITY posts by NBA teams during the 120-day period. On average, teams posted 17.5 times per day, with the Golden State Warriors at 8,107 total posts (or 67.6 posts/day) at the maximum, and the Minnesota Timberwolves at 822 total posts (or 6.85 posts/day) at the minimum.
From a temporal perspective, FITY posts were delineated into three groups: morning (i.e., 12:00 AM to 7:59 AM), workday (i.e., 8:00 AM to 3:59 PM), and night (i.e., 4:00 PM to 11:59 AM).
Curiously, the night period was the time frame in which 28 of the 30 NBA teams were most active with their social media posting; the Minnesota Timberwolves and Cleveland Cavaliers were most active in the workday. On a percentage basis, the night period accounted for 65% of all posts, with the afternoon (23%) and morning (12%) following.
When the temporal data is examined by days and months, the findings suggest these teams were posting in a parabolic fashion, with operations slow to begin, peaking in the midst of the pause, and slowing back down before the resumption of play.
As Figure 1 depicts, posts were scarce on Saturdays (10%) and Sundays (7%), the traditional window for professional sport. As the weekday progressed, teams began to post more frequently – Mondays accounted for 12% and Tuesdays were 14% of all FITY posts – and culminated on Wednesdays (23%).
In this parabolic fashion, Thursdays and Fridays saw a decline from the Wednesday highpoint, with 17% and 16% of all posts respectively. Month of posting saw a similar trend, too (see Figure 2).
March, the month in which this pause began, only accounted for 11% share of the posts, while June and July, the two months concluding this turbulent period with no sport product, were at 14% and 5% respectively.
Interestingly, it was the middle of the period, April, followed by May, when teams were most active, with 32% and 38% of all team posts on the FITY platforms coming during those two months.
When and where teams posted on social media during this period is relevant insofar as it frames the engagement data that were produced. There was an average of 5,391 engagements (i.e., likes, comments, and shares) per FITY post, with Instagram being the most engaging platform with an average of 32,940 engagements per post.
The remaining platforms saw averages of 2,274.5 for Facebook, 570.6 for Twitter, and 275.4 for YouTube. While teams were more inclined to post in the night time period, fan engagement averages were higher for the morning (n = 5,302.6) and workday (n = 6,125.4).
Concurrently, day of the week and month of the year trends differed for user engagement relative to teams’ behaviors. The most engaging days of the week were the weekend days; Sunday (n = 8,234.6) and Saturday (n = 6,402) led the way, with Tuesday (n = 6,005.6), Thursday (n = 5,663), Monday (n = 5,294.9), Friday (n = 4,797.1), and Wednesday (n = 3,908.1) following thereafter.
Month of the year also deviated from team behavior, with engagements at their peak in March, the beginning of the pause, falling throughout, and then rising again with anticipation for the resumption of play.
Discussion and Recommendations
The findings confirm that turbulengagement is, in fact, real. That is to say, when the sport product is no longer present or “paused”, fan engagement behaviors still remain. That might seem logical and obvious, but having this empirical data is an important base from which to draw further conclusions given there are alternatives to social media engagement (e.g., video games, analog exercise, conversing with household members).
Specifically, as Figures 3, 4, and 5 highlight, teams in this period of turbulence operate their social media operations diametric to that of digital consumer behaviors. This is quite a startling result that practitioners and those within the sport management academe need to consider further.
While that may seem obvious for practitioners and academics alike, the findings here would suggest that teams are blindly posting on major platforms (i.e., FITY) without regard for when their publics, notably fans, seek to engage vis-à-vis when engagement occurs most often.
Whether it is the time of the day, day of the week, or month of the year, sports teams are operating with the hopes for an engaging result, but incongruent with when fans and other digital followers want to offer engagement. Delving further into this result, perhaps the most erroneous practice is the time of day distinction.
The core sport product at the professional sport level is typically delivered in the night time period and, concurrently, much of the promotional and engagement activities are concentrated around this timepoint, too.Dees et al., 2022
However, when there is turbulence causing the core sport product to pause or stop (e.g., labor stoppage), this strategy should be questioned. As the findings highlighted, realized engagement occurs more frequently in the morning and workday periods, which should stimulate professional sports teams to refine their actions in trying times, as well as non-trying times.
Certainly, the argument can be made that with COVID-19, there has been an increase in the number of sports fans (and individuals in society at large) who are working from home, schooling from home, and, generally, sheltering in place, who are more inclined to be on their smartphones, tablets, and laptops and engage via social media (Mastromartino et al., 2020). However, this should not be the strategy employed solely for turbulengagement principle.
As Naraine et al. (2019) explained, fans in non-turbulent, “normal” periods (i.e., where the core sport product is active) seek to engage with professional sports teams in the workday period, particularly during commute times and other breaks in the day (e.g., lunch hour).
In this respect, the findings here should not be disregarded as a one-off that practitioners can chalk up to COVID-19, a major anomaly in the delivery of professional sport.
Indeed, the true value of unearthing this turbulengagement through NBA teams is that whether the sport product is “on” or “off”, fans want engagement throughout the day, not just when the product is being delivered live.
Furthermore, practitioners should also be cognizant not to try and fit their social media approach into a “one-size-fits-all” given the demographic and behavioral characteristics of its followers.Naraine, 2019
Given there are different consumer segments following their communication, a dynamic, fluid social strategy should be employed to support the sport product and facilitate an ongoing fan-to-fan connection in the network Yan et al. (2019).
Although practitioners may be weary of this study’s findings, they should not be dismissed as a simple anomaly. Nearly two years after Rudy Gobert’s positive test, the world is still experiencing the COVID-19 pandemic.
While most sport operations in North America have resumed play, there is a very real possibility of a reversion back to the strict restrictions imposed at the beginning of the pandemic as variants of concern (e.g., Omicrom) spread across the Western Hemisphere.
In this pessimistic view, there exists the very real possibility of professional sport being paused, again, alongside alternative entertainment options (e.g., concert- and movie-going).
Consequently, the findings here provide substantial impact for practitioners to refine their social strategy in this fluid environment. With a more optimistic view, even with an increased return-to-work and return-to-play dynamic, practitioners should not dismiss the potential for increased digital behaviors (Mastromartino et al., 2020).
The COVID-19 pandemic has paved a path towards an increased digital presence for consumers through increasing the work-from-home lifestyle. In that spirit, the findings of this study still hold for periods where the core sport product is absent whether due to a major catastrophe or labour stoppage.
Practitioners should harness this study to data-inform their social media decision making as opposed to negating these findings simply on the basis of its one-off context; the COVID-19 environment continues to wreak havoc on professional sport and, although practitioners, researchers, and consumers may suffer from COVID-19 fatigue, that fatigue should not be used as an excuse to invalidate data points which can help inform strategy moving forward.
While the findings here are novel, there remain important limitations and avenues for future research. For one, this study was limited to one turbulengagement period, the NBA pause due to COVID-19 in 2020; the dates of the analysis (i.e., March 11, 2020 to July 8, 2020) were chosen to represent the ceasing and resumption of core product activities in Orlando, Florida, though it was possible some team activities may have resumed slightly before that date altering the potential communication sent out via social.
Second, data were scraped via the FITY platforms, aligned to the Eastern Standard Time zone, and purposefully excluded other platforms such as Snapchat, Reddit, and Sina Weibo.
Third, only NBA team activity and the engagements to that activity were captured and amalgamated, and the study did not distinguish between specific engagement metrics, nor did it ascertain fan interests, demographics, and behaviors (aside from their engagements with team posts). Related, no share metrics were obtained from YouTube and Instagram given the privacy limitations of those specific platforms, which may have impacted the total number of possible engagements.
A final limitation exists in the type of social media analysis undertaken; while temporal analysis is valuable, there was no complex statistical modeling or conducted in this study.
Furthermore, the specificities and themes of team content produced throughout the morning, afternoon, and evening were not discussed (especially as the analysis was performed at a league level aggregate).
This produced an inability to determine what types of content receive the highest amount of engagement or opine the reasoning behind why some teams posted more than others.
In light of these limitations, there are opportunities for future research to examine this turbulent period with other sports teams and leagues to triangulate these findings, particularly for teams in the Global South (e.g., Indian Premier League).
Relatedly, studies should seek to expand from just professional teams to athletes and sponsors, key stakeholders in sport industry ecosystem. Another study could scrape data from emergent or unique social media platforms like Danmu or Discord to examine trends during similar periods; even though professional sport has resumed in many parts of the world, the core sport product does not always encompass full days, so it would be beneficial to look at engagement during the morning, workday, and night periods in other jurisdictions.
Finally, a future work ascertaining fan motivations of turbulengagement or engaging on social media when the sport product is not live would extend the current work, as well as a follow-up study to model turbulengagement over a longitudinal period or across multiple teams and leagues.
Note: The author would like to acknowledge Trevor Wakayama for his research assistance during this project.
Achen, R. M., Kaczorowski, J., Horsmann, T., & Ketzler, A. (2018). Exploring off-season content and interaction on Facebook: A comparison of U.S. professional sport leagues. International Journal of Sport Communication, 11, 389–413. https://doi.org/10.1123/ijsc.2018-0013
Brooker, P., Barnett, J., & Cribbin, T. (2016). Doing social media analytics. Big Data & Society, 3, 1-12. https://doi.org/10.1177%2F2053951716658060
Byers, T., Gormley, K-L., Winand, M., Anagnostopoulos, C., Richard, R., & Digennaro, S. (2021). COVID-19 impacts on sport governance and management: A global, critical realist perspective. Managing Sport and Leisure. Advance online publication. https://doi.org/10.1123/ijsc.2014-0083
Dees, W., Walsh, P., McEvoy, C. D., McKelvey, S., Mullin, B. J., Hardy, S., & Sutton, W. A. (2022). Sport marketing (5th ed.). Human Kinetics.
Golliver, B. (2020, July 29). Ridiculed and alone, Rudy Gobert was the face of coronavirus in sports. Now he’s ready to return. Washington Post. https://www.washingtonpost.com/sports/2020/07/29/rudy-gobert-coronavirus-nba-return/
Mastromartino, B., Ross, W. J., Wear, H., & Naraine, M. L. (2020). Thinking outside the ‘box’: A discussion of sports fan, teams, and the environment in the context of COVID-19. Sport in Society, 23, 1707-1723. https://doi.org/10.1080/17430437.2020.1804108
Naraine, M. L. (2019). Follower segments within and across the social media networks of major professional sport organizations. Sport Marketing Quarterly, 28, 222-233. https://doi.org/10.32731/SMQ.284.122019.04
Naraine, M. L., & Karg, A. J. (2019). Digital media in international sports: Engaging fans via social media and fantasy sports. In E. MacIntosh, G. Bravo, and M. Li (Eds.), International sport management (pp. 315-331). Human Kinetics.
Naraine, M. L., Wear, H. T., & Whitburn, D. J. (2019). User engagement from within the Twitter community of professional sport organizations. Managing Sport and Leisure, 24, 275-293. https://doi.org/10.1080/23750472.2019.1630665
Pedersen, P. M., Laucella, P. C., Kian, T. K., & Geurin, A. N. (2021). Strategic sport communication (3rd ed.). Human Kinetics.
Peters, K., Chen, Y., Kaplan, A. M., Ognibeni, B., & Pauwels, K. (2013). Social media metrics — A framework and guidelines for managing social media. Journal of Interactive Marketing, 27, 281–298. https://doi.org/10.1016/j.intmar.2013.09.007
Sheinin, D. (2020, March 14). The week the coronavirus ground the sports world to a halt. The Washington Post. https://www.washingtonpost.com/sports/2020/03/14/sports-cancellations-timeline-coronavirus/?arc404=true&itid=lk_readmore_manual_20
Wakefield, L. T., & Bennett, G. (2018). Sports fan experience: Electronic word-of-mouth in ephemeral social media. Sport Management Review, 21, 147-159. https://doi.org/10.1016/j.smr.2017.06.003
Wang, Y., & Zhou, S. (2015). How do sports organizations use social media to build relationships? A content analysis of NBA clubs’ Twitter use. International Journal of Sport Communication, 8, 133-148. https://doi.org/10.1123/ijsc.2014-0083
Watanabe, N. M., Shapiro, S. & Drayer, J. (2021). Big data and analytics in sport management. Journal of Sport Management. Advance online publication. https://doi.org/10.1123/jsm.2021-0067
Yan, G., Watanabe, N. M., Shapiro, S. L., Naraine, M. L., & Hull, K. (2019). Unfolding the Twitter scene of the 2017 UEFA Champions League Final: Social media networks and power dynamics. European Sport Management Quarterly, 19, 419-436. https://doi.org/10.1080/16184742.2018.1517272