MONTY: THE WORLD'S FIRST PREDICTIVE COMMENTATOR

Gold Spike

Case Film

Presentation Image

TitleMONTY: THE WORLD'S FIRST PREDICTIVE COMMENTATOR
BrandFOXTEL
Product / ServiceFOX CRICKET
CategoryC02. Use of Real-time Data
EntrantMINDSHARE Sydney, AUSTRALIA
Idea Creation MINDSHARE Sydney, AUSTRALIA
Media Placement MINDSHARE Sydney, AUSTRALIA
Production GOOGLE AUSTRALIA Pyrmont, AUSTRALIA
Production 2 OOH MEDIA Sydney, AUSTRALIA
Additional Company OPTA SPORTS London, UNITED KINGDOM

Credits

Name Company Position
Jack Smyth Mindshare Sydney Head of Innovation
Simon Wallace Mindshare WW Data Science Partner
Drew Jarrett Google Australia Customer Solutions Engineer
Yuliya Kudryavtseva Google Australia Ads Solutions Consultant
Steve Cliffe oOh! Media Creative Manager

Why is this work relevant for Media?

The world’s first artificial intelligence (AI) sports commentator, ‘Monty’, has transformed the way Australians watch cricket. Monty watched every single ball in every single game, triggering dynamic creative across pre-roll video, mobile display ads and outdoor billboards when he spotted a wicket. Fans quickly recognised him as the season’s breakout star. The model ensured they never missed a play worth paying for. Monty delivered an 18% increase in average weekly sales and Fox Cricket became the number-one channel for share.

Background

Foxtel is Australia’s premium subscription TV network, home to the FOX Sports network and an aggressive player in the nation’s sports rights market. Foxtel’s AU$600m broadcast deal for cricket was a strategically vital investment to secure a competitive advantage over the next six years. And, with a price tag like that, expectations for the cricket launch campaign were extraordinarily high: • Acquire new subscribers, while also reducing CPA by 40% • Steal share of free-to-air viewing from CH7 • Become number-one channel on Foxtel This would be an intimidating brief at the best of times. To make matters worse, thanks to Australia’s arcane sport licensing laws, we were asking fans to pay for a product they could still partially enjoy free of charge through competitor broadcaster CH7. We had to create a cricket experience unique to Foxtel, a new way to watch the game that would set our subscription apart.

Describe the creative idea/insights (30% of vote)

The only way to hit our targets was to change the consideration set. We wouldn’t compete with free cricket - we’d create an entirely new cricket experience worth paying for. To understand what fans would be willing to pay for we built a detailed map of major matches where we could trace fan attention and engagement for each play, overlaid with social conversation data, revealing what mattered and when. The results were striking. Despite millions of Australians following the game every summer, few ever saw the best plays on the pitch. The fan’s favourite play – taking a wicket – accounts for only 18 seconds of the average 30-hour game and the majority only saw it through replays. This was our key insight – most key moments in our audience’s favourite sporting pastime had never been seen truly live.

Describe the strategy (20% of vote)

We knew we had to deliver more value for the subscription and ensure fans never missed a moment worth paying for. In short, we had to predict wickets. So we created the world’s first artificial intelligence (AI) predictive commentator for cricket: ‘Monty’. He tells you when a wicket is about to fall and makes sure you are watching when it does. Monty watched every single ball from last year’s season and tracks 83 variables in live games every time the bowler begins his run up (pitch condition, weather, speed of delivery, etc). With our creative idea coming to life through cutting edge tech, our media strategy was designed to turn every impression into a prediction: investing 3:1 into data against standard media. We prioritised channels that could adapt in real time and keep pace with the game: digital video, digital outdoor, display, app and voice.

Describe the execution (20% of vote)

Once Monty was trained, we let him loose on live games. When he spotted a wicket coming, he triggered dynamic creative across pre-roll video, mobile display ads and outdoor billboards, with a call to action to tune in to Fox Cricket and watch the wicket fall – making sure subscribers got maximum value and alerting the rest of the public to the fact that Fox (via Monty) could ensure they never missed a big play. Through a Google integration, fans could also ask Fox Cricket for ‘Monty’s Call’ through the Google Home Assistant during live play, and an API was also integrated into the Fox Cricket app for instant predictions and enhanced visualisations, which fans could use to help with suggestions on drafting their ultimate fantasy team. Rather than using data for targeting, we predicted big plays – putting the viewer ahead of the game for the first time ever.

List the results (30% of vote)

Monty correctly predicted more than 1,867 wickets across the season with an average accuracy of 79% for live deliveries. Our aggressive strategy to invest in data over inventory, and create an entirely new way to watch the game, paid off. Monty over-delivered against every core objective: Acquire new subscribers and reduce CPA by 40% – average weekly sales increased 18% post-launch and CPA was 61% below average (i.e. 53% above target). Steal share of free-to-air viewing – CH7 share of P16-54 dropped to 20.8% from 49.9% (CH9 – previous rights holder) year prior. Become number-one channel on Foxtel – We were number one for share, beating the full-year performance of every other channel … despite only launching in September.

Describe the use of data, or how the data enhanced the campaign output

Our primary source of data was Opta's international cricket feed, which allowed us to draw on historical matches to train the model and also receive accurate information on every single ball during a live game within seconds of the delivery on the pitch. A custom hosting environment was built to receive, filter and feed the information to the model. This dataset formed the model's core training material, with adaptation as required to adjust for the critical values consistent across all international fixtures. Over the course of the summer season the model received regular tuning based on performance against actual results to refine accuracy.