(upbeat music) Today, I'm joined by one of AFL's biggest stars. No, it's not this year's Brownlow Medal winner, nor is it Collingwood captain, Darcy Moore. It's none other than AFL's Head of Data, Elisa Koch. Welcome to the show, Elisa. Thank you so much for making the time to chat to me today. Thanks, Janice, thanks for having me, excited to be here. So, Elisa, I think, when we've spoke in the past, I told you, I don't know too much about AFL, it's not something that I grew up watching in my household, although I did borrow a scarf from a 10-year-old Collingwood supporter I know. (Elisa laughing) Elisa, could I start by asking you to just share for our viewers, our show does go out globally, what is the AFL and why is it so loved by so many Australians? Absolutely, yeah, so just to clarify, I also did not grow up with Australian rules football, also known as footy or AFL. I'm American, so normally when I get asked, "Yeah, what is AFL? What does it kind of look like when it's being played?" the answer that I normally provide, "It's kind of a combination of soccer, football and a bit of American football, NFL football as well." Not many people know this, but it's played on a oval field, which is actually quite large. I think it's like two times the size of a normal soccer or football field. You can kick the ball, you can catch the ball, you can run with the ball, you can handball the ball. So it's incredibly free-flowing game, and I know anyone who watches it who's never seen AFL being played before are just amazed at how exciting and engaging game it can be. I live in Melbourne, Australia, and it is absolutely religion down here, and as you said, it's a much-loved sport all across Australia. Thank you for that background, I think that's fantastic. Now, Elisa, you've had a very diverse and wide-ranging career to date, from heading up marketing intelligence at Avon over in New York, to now, I guess, building out the data capability for the AFL right here in Melbourne, Australia. Can you tell me a bit more about your journey with the AFL and what you've created over there? Yeah, so I started at the AFL at the end of 2019. And as you mentioned, I've worked in data and analytics my entire career across many different industries, but this is the first time I'm working in sport. Evaluating the opportunity, I knew that sport was an industry with a lot of data about fans, but also a lot of data in terms of kind of what's happening on the field in terms of performance data and all of that. So it was an industry that I was incredibly excited about, and the opportunity that was put in front of me was about building a data capability from scratch. And it was building a capability not just for the AFL as the league, as the governing body, but also for all 18 AFL clubs as well. So it was a really unique challenge and opportunity that I kinda jumped into at the end of 2019. Fast-forward to 2023 and that's where we're at, a shared data capability, kinda end to end. So everything from data warehousing of all our industry data sources, and we make that data available to clubs, whether it's through reporting, visualization, all the way up to having machine learning capability for them as well. So it's been an incredible adventure to get this kind of built from scratch, but so rewarding to see the AFL and all 18 clubs on board. I know something you're really passionate about is that community aspect you've created with the data, I guess, like any competitive sport, we're used to seeing the competition between the various clubs on the field, how has that worked behind the scenes? Is there competition when it comes to accessing and using this data? Yeah, it's something I've really like to talk about and what I'm so proud of in terms of the capability that we've built for the AFL and clubs is just this aspect of building a data community. I think, with data teams, and this is really for any industry where a data team operates, it's really important to know that to really scale the impact of a data team, you really need to democratize your efforts and really train your end users to be able to train and upskill not only themselves, but for everyone around them. So we put a really big focus on having the data community that not only analysts and business users can be a part of at the AFL, but also all 18 clubs. And people are really surprised to hear this because they just assume that you have 18 sports teams, they're gonna be incredibly competitive with each other. And what we see is that, absolutely, on the field, (laughing) they are incredibly competitive, but when it comes to collaborating as analysts within the data community, we actually all wanna help each other out. And we understand the challenges that come with being a data analyst and knowing that you have a community around you to bounce ideas off of and learn from, it's just been amazing. And again, as I mentioned earlier, it's allowed my team to really scale our impact beyond just the 12 people on my team. We have evangelists out there evangelizing on our behalf. So again, I really do stress the importance of having a data community as part of your broader capability as a data team. That's fantastic, I think it really is data democratization at its best what you've created over at the AFL. You talk about, I guess, having this sort of shared resource when it comes to data, are there any particular tools or technologies that you've invested in that you think have really helped you achieve the outcomes at the AFL? Yeah, I mean, really just investing in technology that is centralized. Kinda cloud-first has been really important for us, knowing that, as we increase in scale, that the technology can scale with us. So investing in a centralized data warehouse, investing in a centralized reporting and visualization platform, and also, most recently, we've invested in a centralized kind of data prep and auto-machine learning platform as well. So again, this is all technology that we've centrally procured at the AFL, but then we allow clubs to have licenses as part of that industry solution. It's great to see you guys already moving into machine learning. Is there a particular business use case that you're proud of that you're able to share with us when it comes to ML? Absolutely, yeah, so really we thought about, our first machine learning use case, we wanted to know that there'd be ROI or dollars that we can attribute to it. So when you think about the AFL and clubs, where we derive a lot of our revenue is through club membership. So you have a Collingwood Magpie scarf on. (Janice laughing) So there are, I believe, close to 100,000 Collingwood members, and each club will have a very loyal kind of fan and member base. So our first use case for machine learning was to build membership churn models, which essentially takes all the behavioral data that we have on members, checks it all into a machine learning model, and out comes a churn propensity score, which enables clubs to be a bit more proactive in terms of, look, if they have a few members that are showing a high risk of churn, based on the churn propensity score, then they'd be able to proactively reach out or engage with them to reduce that churn risk. So leveraging machine learning around the consumer membership base was an initial use case, and we're now in year two of rolling that out to all clubs. Fantastic, I think that's something that I think a lot of sporting clubs could really learn from and use those sorts of business use cases. Now, Elisa, I'm conscious of time, I did wanna touch on, you've done some amazing work with data and analytics, but I think the AFL's use of technology in general has been really inspiring. I've heard all about your Just Walk Out stores at Marvel Stadium, using a combination of computer vision and machine learning to really give customers this seamless customer experience when buying something from the stores, to some of the work you're doing around accessibilities to really make the game accessible to blind and visually impaired fans of the AFL. What else can we look forward to? What's next on the agenda? Absolutely, so from a technology perspective, everything that you've mentioned have just been, yeah, real innovation for the AFL, but not only for us, but also across sports. So really exciting to be playing in those areas. I would say, within the data and analytics space, just I'm very appreciative that I'm in an environment where the business allows me to trial and test new approaches. Machine learning that had never really been leveraged before in the industry, but now we have a few use cases kind of up our sleeve. So I'm really looking forward to trialing in a few different areas. Kind of mapping has become incredibly important, being able to visualize our fans demographically and geographically, understanding penetration in different areas across the country, that's been a really interesting capability that has unlocked a lot for us. So really it's just been about understanding pain points, finding new innovative ways to solve them and just kind of trialing it through. So really excited about kind of what we have upcoming in the coming years. Amazing, well, keep up the fantastic work. I just can't wait to see what comes out of you and the team next. Elisa, thank you so much for making the time to join me today and share your experiences with the AFL with us. To our viewers, I just wanted to say thanks, once again, for tuning in. I hope you've enjoyed today's episode. Feel free to comment below. I'm sure I'll get a few comments on my choice of football scarf for today. And Elisa, thank you, once again, it's been a pleasure to speak to you. Fantastic, thank you for having me, Janice.