Beyond the buzzers and the bright lights at every NBA game stands a global media company. In a typical season, fans get to see their favorite team play 82 regular season games before the summer playoffs create the epic moments and matchups that have defined generations of superstars and driveway reenactments for kids from Los Angeles to New York to Beijing.

With 30 teams (each valued at over $1 billion) playing 82 games, scheduling is an essential component of the NBA as a product. The television revenues alone produce over $2 billion dollars annually and stakeholders from fans and advertisers alike want their teams to be front and center for the world to witness the dunks and drama generated by these incredible athletes. The NBA’s success requires the organization to accurately estimate the viewership for every single game, using that information to make key decisions, including which games are broadcast on which networks and which games receive advanced promotion. Fortunately, like all major sports, there are decades of data about the NBA from a wide variety of sources to help with estimating viewership.

Perhaps no one knows just how much data is available better than Jonathan Berkowsky, Senior Manager of Global Media Insights for the NBA. Jonathan will be one of the first people to tell you that having data and generating insights from data are two very different things.

That is why the NBA partnered with Domo to turn their estimated viewership processes from a manual computation to an automated pipeline that can produce instantaneous assessments of estimated viewers with a much wider set of variables than Dr. James Naismith could have ever imagined when he first invented the game in 1891.

Revolutionizing How the NBA Estimates Viewership, One Dataset at a Time

Before partnering with Domo, Jonathan’s team had to use a manual process to compare historical data from Nielsen’s TV ratings and hope that the data would provide insights that could be projected for future games. Even though the data was coming from one source, the manual process was labor intensive and simply not as accurate as the NBA needed.

As a result, achieving clarity on some of the biggest questions was far more difficult than it should have been. What would happen to the viewership if a team was on a winning streak? What would happen if the team’s superstar was injured? Are some fans so loyal that they will watch their favorite team play knowing they were not projected to win? The NBA needed more data, and more speed and agility to put it together, so they could gain insights that would help the NBA succeed in an increasingly competitive media landscape.

With Domo, Jonathan’s team dove into a huge variety of datasets. Among their many insights, they found that the network an NBA game is shown on is an essential variable for estimating viewership. Additionally, which players are on the court is a key predictor. As Jonathan explains, “superstars make a big difference for viewership, and we have now built into the model when our superstar players will be missing because of injuries or otherwise so the estimates can be adjusted manually or automatically.”

Players are also a key part of matchup appeal. As the regular season comes to a close, Jonathan’s model can evaluate teams that are on a winning streak along with matchups that are important for the playoffs. After building the model, Jonathan can watch as it automatically updates with every player, team, and network performance night in and night out. According to Jonathan, “we have a lot of variables that can impact viewership of a specific game. And a lot of those variables can depend on each other. With Domo we are delivering estimates with a great deal of accuracy because Domo took this massive set of data and helped us pinpoint the variables that are the most impactful.”

Knowing the variables is an important step, but Jonathan’s team also worked with Domo to develop an organization-wide digital literacy that has improved decision making at all levels. As Jonathan explains, “the amount of detail that we’ve been able to implement that we didn’t know to look at before, or we didn’t have the resources to look at before has been helpful for everyone. The whole organization has access to the most up to date data so we’re able to make the best data informed decisions.”

Improving the Fan Experience Today and Beyond

The partnership with Domo has helped the NBA make better predictions about viewership for any given game. As Jonathan explains, those predictions radically improve the experiences of fans across the globe. Jonathan notes “with Domo and with the speed that comes with it, we’re able to be agile with the schedule changes and we have the flexibility to look at the games that are currently scheduled at any given time and consider which ones will drive the most viewership and be the most exciting for our fans.”

In addition to the agility that comes with having access to real-time data, the process of automation has allowed Jonathan’s team to focus on new ways to improve the overall product for the global community of fans. As Jonathan explains, “the speed that we’re able to do the estimations now is outstanding. We can upload an entire schedule of regular season NBA games, and be able to automatically estimate the viewership for each one on a given day at a given start time. In the past that was a manual process for each individual game.”

Knowing that the fans are getting a better product day in and day out is what drives Jonathan to continue to develop new insights through Domo. As he notes, “I am a part of something that is special, delivering basketball to the world and trying to make that game better. That is what drives me to apply my skill set and things that I’ve learned to help this organization to be as successful as it can be because I want to see basketball succeed globally.”

Digging Deeper: How the NBA is Continuing to Utilize Domo to Gain New Insights

Although estimating viewership is an essential part of the NBA success story, the partnership with Domo has helped the entire organization improve its decision making. As Jonathan explains, his team has developed a reputation as the people to go to when someone needs data because “we are able to be that single source of truth, while being flexible enough to provide the data however people need it. We now provide data insights for the whole organization and building a structure in Domo has been hugely helpful because with Domo, we can find interesting insights by looking at the data in a different way.”

According to Jonathan, an essential part of the partnership with Domo has been the way that non-technical employees have been able to use Domo to generate key insights. As he notes, “those of us on the team that worked on this project, we are not data scientists. And, even as a non-data scientist, the Domo team was great at explaining everything to me and showing me the ropes so that if I do need to make changes, I’m able to go in and make the key adjustments.”

What started as a project with Domo to incorporate existing data has evolved into a nuanced automated model that enables the NBA to make key decisions about its product for fans across the globe. Now the question for John and his team is just how far they want to go to examine more variables. One conversation his team has had recently has been, “does what jersey a team is wearing impact the viewership? Our teams have a few different types of jerseys and that’s the type of data that we could adjust to see if it matters. The potential impact of jerseys on viewership is the type of data that would be fun to know.”