99P Labs & CMU’s Data Analytics Club Rideshare Challenge Fall 2022
During the 2022 fall semester, 99P Labs and Carnegie Mellon University’s Data Analytics student club at the Tepper School of Business collaborated on a two-week rideshare-themed data challenge. This blog post recaps the challenge.
The Kickoff
The challenge kickoff was a hybrid event that took place both in The Tepper Quad and online. In-person, we had a few dozen club members and were joined online by about twenty more. In addition, the 99P Labs team was able to make it to Carnegie Mellon University that evening so we could be in person for the kickoff of the challenge. The opening event was a great time.
Tepper Data Analytics Club leadership started the kickoff by introducing the 99P Labs team, then they went through a presentation laying out the details of the challenge. After that, they opened the floor for the club to ask the 99P Labs members questions. We spent the rest of the event answering the member’s questions and getting to know the students.
The Challenge
The 99P Labs team and the leadership of the Tepper Analytics Club met numerous times before the event to scope and fine-tune the challenge. The hope was to create a challenge that was broad enough to leave paths available for creativity but to have just enough constraints to give the competitors some idea of where to get started.
The official prompt was:
As we slowly return to a ride-sharing world like we had prior to COVID, we want to better understand associated trends and externalities. From a customer perspective, ride-sharing has environmental benefits, such as decreased CO2 production. There are, however, drawbacks like (potentially) increased travel time. So how can a ride-share company create a value proposition that accounts for the environmental benefits and potential drawbacks?
While using the 99P Lab’s Microtransit data set (and, optionally, other data sources), challengers should attempt to explain the ROI of weighing the environmental benefits of using ride-share against the time costs and make a final recommendation to a hypothetical ride-share company.
Providing Support for the Event
The Tepper Data Analytics Club leadership did an excellent job coordinating efforts for the 99P Labs team to help support the members regarding the challenge and the data. They set up two different options: an office hour over Zoom and a Discord server where students could ask questions, and the 99P Labs team could answer asynchronously.
The office hours were tremendous; they allowed us to connect with some participants and provide context to the problem statement for the students. There was substantial interest from the members, the whole hour flew by, and there was never a long pause after a question. An interesting observation was the game theory flavor, where some of the questions were asked in a way that attempted to give the least amount of information about their use case for fear of it being helpful to a competitor. But after the initial fear of “giving your competitor an edge” passed, the members started opening up, and we made good progress. My favorite part of the conversation was towards the end when we discussed the conceptual underpinnings of innovation.
We could also connect with the participants through the Tepper Data Analytics Club’s Discord server. Discord gave the members who could not make the office hours a way to ask questions. But, again, since there was a genuine fear of the other teams gaining something from hearing about other teams’ insights, most of these questions occurred over direct messages. Nevertheless, I had numerous fruitful conversations and could answer questions, so I think the Discord server was definitely a value add.
The Finalists
In the end, fifteen teams submitted entries and the Tepper Data Analytics Club’s leadership chose three finalists. The finalist teams were “Team Outlier,” “Coda,” and “Data-vengers.” Now I will go through each of their executive summaries.
“Team Outlier” analyzed trip data over seven months based on time of day, assessing utilization of current assets and identifying potential areas of improvement. Based on their analysis, they suggest two actions to lower emissions and enhance the sustainability of business operations:
(1): Shift asset usage from larger vehicles (>=9 seat capacity) to smaller vehicles during times of day with lower occupancy trips.
(2) Invest in Electric Vehicles (EV) to take advantage of federal tax credits and to lower emissions.
“Coda” started their executive summary by envisioning a company with a mission to reduce carbon emissions from transportation by helping people reduce their need for driving, whose primary business model is a scheduled commuter ride-sharing service.
Their main conclusion was that the data suggests the company is not operating as efficiently as it could because of inactive cars, long wait times, and a high no-show rate. They offered three suggestions, two strategic suggestions, and one business model adjustment.
- Increased efficiency in vehicle matching (use a smaller capacity car for trips with fewer riders)
- Improve the passenger experience by providing real-time traffic data and a policy to arrive 10 minutes early.
- The business model adjustment is to try to target small and mid-sized businesses that can’t offer company shuttles to increase demand.
The “Data-vengers” took a highly literal interpretation of the prompt and created an equation attempting to quantify the relationship between environmental benefits and time cost. Their executive summary walks through the math of their equation and ends by giving recommendations for three different zones.
- For marketing, they suggest the company should target the younger generation by partnering with schools and universities and create user growth campaigns like referral programs and an environmental contribution dashboard.
- For strategic partnering, the company should collaborate with NGOs, an environmental agency, and related alliances to create a more cost-efficient and environmentally friendly operating model.
- Finally, for investment, the company should consider replacing all or part of the current cars with Electric Vehicles.
The Final Presentations
All three finalists presented their submissions to a panel of judges consisting of Carnegie Mellon University faculty and 99P Labs team members. The finalists presented from a classroom on campus, and the judges watched over Zoom. The teams did a phenomenal job with the presentations; it was great to see them given a chance to explain and pitch their submissions and respond to questions from the judges. After all the presentations, the judges briefly deliberated in private to determine the ordering of the finalists. Although all three finalist teams’ submissions were excellent, there had to be an order determined in the end. “Team Outlier” came in first, “Data-vengers” came in second, and “Coda” came in third.
So to wrap up this post, we would like to thank the finalists for their presentations, all the competitors for the time, effort, and care they put into their submissions, and the Tepper Data Analytics Club’s leadership for helping run such a successful challenge. The 99P Labs teams had a great time and look forward to future collaboration with the club.
Conclusion
As always, if you’d like to stay in the loop for all our engagements, research, and news, we would love it if you reached out to connect with us on LinkedIn and gave us a follow here on Medium. Also, please check out our developer portal if you want to take a closer look at the datasets. Thank you so much for reading!
