Predicting Recommendation of Sporting Events: A Study of Satisfaction Levels Among Tennis Event Promoters and Non-Promoters

Resumen

Aim and Research QuestionsHow satisfied are the promoters and non-promoters of sporting events? How related are theirsatisfaction and their likelihood of recommending the event? This work aims: a) to determinethe differences in the degree of satisfaction between promoters and non-promoters of the event,and b) to establish a predictive model identifying the variables that provide a greaterprobability of not recommending the event.Theoretical Background and Literature ReviewNet Promoter Score (NPS) is an indicator created by Reichheld (2003). To calculate the NPS,a threshold of responses is established which classifies customers into promoters (scoresbetween 10-9), passives (scores between 8-7), and detractors (scores between 6-0).The NPS is a popular metric in various market sectors, including sports (Dalmau-Torres et al.,2022). This simple tool measures promoters and non-promoters through a single question,having gained popularity in managerial practice, but with some concerns among academics(Baehre et al., 2022). Its limitations are mainly related to the cut-off points adopted,questionable effect on sales, and the one-single question (Baehre et al., 2022). Some authorsalso defend the continuity to use NPS and identify it as part of the academic research agenda(Bendle et al., 2019), since NPS could be a valid predictor of future sales growth under certainconditions, including short-time application and interpreted as brand health metric and not ascustomer loyalty metric (Baehre et al., 2022).Research Design, Methodology, and Data AnalysisA convenience sampling of 1372 spectators (1% of the total) (average age=39.47±13.72;women=35.7%; men=65.3%) of an international tennis tournament held in Madrid, Spain inearly May 2022 were surveyed face to face during their stay in the tournament. A Likert scale1 to 5 was used to evaluate the degree of satisfaction with different attributes of the tournament:tournament environment, tournament security, comfort on the courts, shopping area, activitieswithin the complex, food stalls, variety of food, price of the food, waiting time (queues) in thefood stalls, sustainability of the tournament, ticket prices, level of the matches and level of theplayers. The NPS variable was measured on a Likert Scale from 0 (very unlikely) to 10(extremely likely). For that purpose, the question ‘How likely are you to recommend thistournament to a friend or colleague?’ was asked.Two types of analyses were carried out. First, the difference between promoters (n=1038) andnon-promoters (i.e., detractors and passives; n=334) was explored. Due to the size of thesample, the central limit theorem was applied and the comparison between both groups wasperformed through parametric Student's T test for independent samples. To measure themagnitude of the difference between the means of the groups, Cohen’s d Statistic was used.Based on Cohen’s d, the effect size was classified as: trivial (<0.20), small (<0.50), moderate(<0.80), large (>0.80) (Batterham & Hopkins, 2006).After the initial analysis, a binary logistic regression model followed. NPS was recoded as adichotomous variable (0=non-promoter; 1=promoter) and used as the dependent variable. Allsatisfaction variables were included as independent variables. Two preliminary logisticregression models were performed using the forward and backward procedures with the WaldStatistic. Collinearity issues were assessed through the correlation matrix. The final modelincluded ‘tournament security’, ‘level of the matches’, and ‘ticket price’. The cut-off point forthis model was set at 0.7 to improve its specificity. SPSS 26.0 was utilized for the statisticalanalysis.Results/Findings and DiscussionThe comparison between the satisfaction of non-promoter and promoter groups was analyzed.Each of the analyzed variables showed statistically significant differences (p<0.05) with amoderate effect size (ES=0.50 to 0.79), except for the variable ‘waiting time in the food court’(ES=0.41) and ‘level of players’ (ES=0.45), which exhibited small effect sizes.According to the binary logistic regression model, the ‘tournament security’ showed astatistically significant increase in the odds ratio of being a promoter of the event by a factorof 1.7 for each unit of growth in this variable. Additionally, for each unit of growth for thevariable ‘satisfaction in the food court’, the odds ratio in favor of the promoter category wasincreased by 1.396. Finally, the odds ratio of being a promoter of the event was multiplied by1.642 for each unit of increase in the variable ‘satisfaction with the ticket price’.These results align with prior research that identifies that satisfaction plays a critical role inword-of-mouth behavior (Kim et al., 2014), but contributes to understanding the event’sattributes that contribute the most to the event brand health.Conclusion, Contribution, and ImplicationThis study revealed significant differences in satisfaction levels between promoters and nonpromoters of the event. Moreover, a predictive model was proposed to identify criticalvariables that influence event brand health. This work has implications, enabling sportsmanagers to prioritize their available resources for enhancing the more critical aspects of theevents.