You’re the newest member of Team X’s BI team. Team X has been collecting data on its customers and their ticket purchases for several years now, but wants to better understand their customers’ behaviors–and better act on their tendencies. Teams of all sports around the world attack this same general problem, and now it’s your turn to dig in. Using the given representative datasets, what can you learn about your current customers and non-buying prospects–and what can you infer about their future behavior?
This problem is intended to be open-ended. Below we list several sub-prompts to get you started. You are free and encouraged to attack any or as many of the sub-prompts below as you would like; you’re also welcome to come up with brand new ideas and put them into action. You can dig deep into one or a couple ideas,or take a wider approach to get a start on several different ideas. The goal is to put you and the rest ofTeam X in a great spot to move forward after you recover from this Hackathon and work continues next week.
1) Identify customer clusters that could benefit Team X, then assign some or all known customers to your clusters. What actions couldTeam X take to “personalize”the experience of each cluster to drive revenue?
2) Build a model or models to predict:
a. probability of attendance of a given ticket buyer at a given game, and/or
b. total scans for a given game. Can you further predict attendance rates in specific sections or regions of your arena?
How can Team X leverage the output of these models–what actions or new offerings might you recommend to the ticket sales team?
3) Analyze historical arrival times via the provided scans data. Are there particular scanners or regions of the arena that are currently overloaded–and if so, over any specific time frames?Further, can you predict arrival times for different games at a micro-or macro-level–including specific seat and/or scanner locations?What would you recommend given your findings? Feel free to use the provided Team_X_Arena_Map.png file and corresponding Scanner_Locations_Map.csv file to visualize your findings.
4) Build a model to predict the ticket purchase behavior of a given customer. Are there particular games and/or ticket products you can recommend for a given customer?The term “ticket products” generally refers to the type of bundle–or lack thereof–of a given sale, such as an individual game, mini-plan (e.g., 4-11 games), half plan (e.g., 21 games) or full season ticket membership. Such products could also include perks like food or merchandise vouchers.For any of the above ideas–or any of your own–feel free to bring in any publicly available outside data you believe could be relevant.
You’re the newest member of Team X’s BI team. Team X has been collecting data on its customers and their ticket purchases for several years now, but wants to better understand their customers’ behaviors–and better act on their tendencies. Teams of all sports around the world attack this same general problem, and now it’s your turn to dig in. Using the given representative datasets, what can you learn about your current customers and non-buying prospects–and what can you infer about their future behavior?
This problem is intended to be open-ended. Below we list several sub-prompts to get you started. You are free and encouraged to attack any or as many of the sub-prompts below as you would like; you’re also welcome to come up with brand new ideas and put them into action. You can dig deep into one or a couple ideas,or take a wider approach to get a start on several different ideas. The goal is to put you and the rest ofTeam X in a great spot to move forward after you recover from this Hackathon and work continues next week.
1) Identify customer clusters that could benefit Team X, then assign some or all known customers to your clusters. What actions couldTeam X take to “personalize”the experience of each cluster to drive revenue?
2) Build a model or models to predict: a. probability of attendance of a given ticket buyer at a given game, and/or b. total scans for a given game. Can you further predict attendance rates in specific sections or regions of your arena?
How can Team X leverage the output of these models–what actions or new offerings might you recommend to the ticket sales team?
3) Analyze historical arrival times via the provided scans data. Are there particular scanners or regions of the arena that are currently overloaded–and if so, over any specific time frames?Further, can you predict arrival times for different games at a micro-or macro-level–including specific seat and/or scanner locations?What would you recommend given your findings? Feel free to use the provided Team_X_Arena_Map.png file and corresponding Scanner_Locations_Map.csv file to visualize your findings.
4) Build a model to predict the ticket purchase behavior of a given customer. Are there particular games and/or ticket products you can recommend for a given customer?The term “ticket products” generally refers to the type of bundle–or lack thereof–of a given sale, such as an individual game, mini-plan (e.g., 4-11 games), half plan (e.g., 21 games) or full season ticket membership. Such products could also include perks like food or merchandise vouchers.For any of the above ideas–or any of your own–feel free to bring in any publicly available outside data you believe could be relevant.