Page 1 of 30
Culvers Location Analysis
Prepared for: Russ and Vicky Polsky
Prepared by: Amber Bovenmyer, Katie Yadro, Allison Spoden, Craig Penkert, Sifan Wu
General Business Intermediate Statistics 304
May 9, 2014
Page 3 of 30
3
Introduction
Our team has been studying statistical techniques and procedures together for the past 12
weeks. An emphasis in our studies has been put on practical data analysis to solve a business
problems. We performed a statistical analysis on data obtained from 89 Culver’s locations in
order to provide you with a reasonable explanation as to which location will be the most suitable
for your needs.
Culver’s wide variety of entrees, including their famous ButterBurgersR and custard,
make the franchise an ideal investment. However, choosing the optimal location to open your
new Culver’s is crucial in ensuring the highest return on your investment. Our analysis of the
three potential locations takes into account the cost to purchase the land, to acquire a franchise,
the fees and royalties due to Culver Franchising System, Inc., the cost of products, salaries, and
the demographics of the population surrounding each potential location. The demographic
variables included per capita income, marital status, and number of autos per household within
both a one and five-mile radius, traffic count, population within a one and five-mile radius, and
Culver Franchising System, Inc. suitability ratings.
We ran various statistical regressions using the data we obtained to determine which
variables of interest were the best predictors of total gross revenue from sales. Our analysis
showed that traffic, population located within one mile of restaurant, and suitability were the best
predictors of revenue. We used these variables of interest in our final model to determine which
location would be the most profitable with taking into account the expenses the franchise would
incur.
Although our initial data set was incomplete in regards to data given for certain stores, we
want to ensure you that we took the correct measures to account for this missing data. We