Inspiration
Super Bowl moments move fast. A touchdown or a momentum shift creates a tiny window where fans are paying maximum attention, and marketing is feel timely. We build a system that reacts to those moments automatically in a way that’s easy to prototype and demo. Fans don’t just watch, they predict. We connect that fan prediction behavior directly to real-time offer, so engagement becomes useful for businesses.
What it does
Fan Touchdown takes a Super Bowl “live feed”, detects special moments like touchdowns, score changes, and quarter changes. Every time a special trigger happens, it automatically generates and publishes an advertisement creative tailored to that moment for small and medium businesses. It monitors engagement on ensured ad posts. When users comment with a prediction about the game, the system sends them a personalized offer notification. If their prediction matches what happens next, they receive a better reward; otherwise they still get a smaller engagement reward. The goal is to make ads feel like part of the game rather than an interruption.
How we built it
We first standardized the game feed into a consistent event format so every moment could be handled the same way. We generate dynamic ad creatives by selecting a template for each trigger, and then publishing the ad. Consumer A listens for those ads, monitors comments for user predictions, stores the prediction with the current game state, and sends offer notifications tied to that interaction.
Challenges we ran into
The biggest challenge was defining what counts as a “special moment” and keeping the system from becoming noisy. The Super Bowl has a lot happening, and if you trigger too often, it stops feeling special. Another challenge was prediction parsing. User comments are messy and ambiguous. We handled this by starting with a small set of prediction patterns. We also had to be careful with timing. In a real live system, there’s delay between the play, the broadcast, and the data feed.
Accomplishments that we're proud of
We built an end-to-end pipeline that can take a live Super Bowl feed and automatically publish moment-based ads in real time. We also connected ad engagement to converting user predictions into personalized offer notifications, which is strong for retention and repeat engagement.
What we learned
We learned that prediction-based engagement works best when the rules are simple and transparent. Users should understand what counts as a prediction, when it closes, and what they get for participating.
What's next for Fan Touchdown
We want to improve prediction understanding with better parsing and richer prediction types, while keeping the UX simple. On the marketing side, we’ll add sponsor eligibility rules, frequency caps, and A/B testing for creatives and offers. On the fan side, we’ll add a lightweight prediction UI and not just comments so we can structure predictions cleanly and reduce ambiguity.
Log in or sign up for Devpost to join the conversation.