One of the most important questions for any sports franchise is how to put fans in the seats.
Conventional wisdom says that winning helps, but to what extent? How effective are a team’s marketing efforts at building attendance, and to what degree are they aligned with how people search for them on the internet? Is the relation between winning, attendance and search interest stable from one team to the next? Or are there subtle variations across the leagues that, in some way, reflect the unique persona of each fan base?
As I flopped down on the couch on Sunday to watch two big market teams battle it out in Game 4 of the World Series, I decided to pick up my laptop and investigate. In order do so, I created a data set comprised of the five teams in the National League West, with winning percentage and total attendance per home game (via Baseball-Reference), and search interest (via Google Trends). While it seems intuitive that all three variable pairs would have noticeable correlations – more wins should draw bigger crowds, as well as generate increased attention online – I was curious about how strong those correlations would be, as well as how those correlations would differ between the teams.
Before we get to that, however, I should touch on a few points regarding methodology. First, there’s the usual caveat that correlation does not imply causation. Demonstrating causation is tricky business and typically requires a controlled experimental design. Think of this instead as a quick and dirty analysis from a biased San Diego Padres fan who also happens to be a search marketing analyst.
Second, data from the first month of the season was excluded from this analysis due to both the potential confound of “early season excitement,” as well as erratic winning percentage on a small sample of games. The attendance for each game was normalized against the average attendance for that day of the week – a process that not only controlled for fluctuations in attendance based on what day of the week a game was played (e.g., Friday games draw bigger crowds than Tuesday games), but also allowed for comparisons between teams with different sized stadiums.
Finally, the attendance data was broken up into week-long increments, corresponding to the preset dates from the Google Trends export, and an average home game attendance was calculated for the week. Only weeks in which there were two or more games played at the team’s home park were included in the analysis.
Ok, now on to the results…
Overall, the biggest correlation is between search interest and winning percentage.
The R2 between search interest and winning percentage for all NL West teams is .20, which in plain English means that 20% of the variation we see in Google search queries for a team can be attributed to how well (or poorly) that team is playing. That’s remarkable! Especially when you consider the strength of correlation is almost three times that of wins and attendance!
It’s also interesting that the two teams with the highest Interest-Wins correlations were the Los Angeles Dodgers and the Colorado Rockies, both of whom had the largest mid-season turnarounds in the division. The Dodgers played miserable baseball through the first two and a half months of the season, posting a .417 winning percentage and trailing the division-leading Arizona Diamondbacks by 9.5 games. Los Angeles then went 58-23 over the next three months, marking the largest deficit that the club has ever overcome to win a division title. The Rockies, on the other hand, had a .563 winning percentage on May 24, and then proceeded to go 47-67 over their remaining 114 games to finish last in the NL West. Colorado was eliminated from playoff contention with a full two weeks left in the season.
Interestingly, though, Rockies’ fans showed up to games with seemingly little regard for how the team was playing, as evidenced by the extremely low correlation between wins and attendance. This leads to the next question…
What team had the most fair weather fans in the NL West? Oh, the Dodgers.
And it wasn’t even close. Roughly 42% of the variation in attendance at Dodger Stadium over the course of the season could be attributed to the team’s performance on the field. By the time LA had clinched the division, fan attendance was up 10% over the previous year, and the club’s television ratings had increased by 40%.
It shouldn’t be a surprise that attendance at San Francisco Giants’ games had little to do with the team’s winning percentage. Since the team moved into their new digs at AT&T Park, they’ve broken the 3 million mark in 12 of the last 14 seasons. In 2013, fans filled the stadium to 99.2% of its capacity – virtually unheard of for a third place team on the west coast. (Of course, it doesn’t hurt that that the Giants have won two of the last four World Series.)
Fans will take in a ballgame for any number of reasons, including the quality of the team, the weather, or what old timer’s bobblehead is being handed out to the first 5,000 fans. Knowing what works and what doesn’t is fundamental to understanding a team’s fan base, and should ideally drive a team’s online marketing program. Specific topics, trends and target personas should be identified in real time, as well as what measurable impact they have on engagement, loyalty, and ticket sales.
This leads us to our final question…
What team’s attendance was most closely aligned with its search trends? The Padres.
A whopping 44% of variation in attendance at Padres’ home games could be explained by fluctuations in search trends. Incredible, right? It’s a search marketer’s dream to have such a strong connection between an activity that’s so easily measured and actionable as search, and (what is presumably) the bread and butter of your offline KPIs. At this point, of course, you would want to further segment the data in order to better understand the kinds of searches that modulate the strength of that correlation (i.e., for ticket deals, videos of game highlights, blog commentary, etc.), but all of this data is easily captured and delivered with the right marketing analytics and insights.
In the end, an analysis like this can have significant implications on marketing strategies, over and above what teams can expect at the gate. With a clearer understanding of your fan base, you are being handed the opportunity to leverage and build long term relations and loyalty. And, as such, your search strategies and social monitoring require an approach to content that is timely, relevant, and responds to a fan’s particular need at a given moment.