4 Conclusion

4.1 Final Thoughts

The boxplot of mean ratings created questions around three main takeaways from our project:

• A significant difference in mean ratings between Raleigh, NC and New York City, NY.

• Determined important variables explaining why Raleigh boba shops are rated higher than NYC, using a linear regression model with stepwise selection and AIC metrics.

• The random forest model confirmed that the “city” feature was important in mean boba shop rating outcomes, which could be attributed in part to physical shop proximity, as seen on the Tableau dashboard.

In Raleigh, Asian fusion and coffee/tea were important Yelp attributions to secure high ratings. Conversely, in NYC, authenticity and the Taiwanese attributions played key roles. These findings suggest that customers in NYC seem to have higher expectations as compared with customers in Raleigh, which was further supported after visual comparisons of each location data set.

4.2 Recommendations

From our conclusion, we recommend that aspiring businesses considering a new boba shop could:

• Understand the preferences and expectations of potential customers.

• Prepare to face the competition in the area.

Our findings emphasize that what garners a high rating in one location may not necessarily translate to the same acclaim in another, especially in densely populated areas like NYC.

Looking ahead, this project encourages future comparative studies of bubble tea shops in different areas across the country or even the world. Expanding the analysis to include major metropolitan locations like Los Angeles or Taiwan could uncover broader trends in boba shop dynamics, providing entrepreneurs with valuable insights for navigating diverse urban landscapes.