# Where to Open a Liquor Store in Iowa

Starting a business is never an easy endeavor but it certainly helps to have statistics guide an entrepreneurs early decisions. I chose to analyze counties in Iowa to see which counties were most conducive to opening a liquor store.

Obviously, the less competition the better. That is why a main feature I’ll focus on is population to store ratio. The graphs below show how population to store ratio correlates to the profit those stores saw. The titles are a bit of a misnomer, as the graphs actually show an inverse saturation level. The higher the population to store ratio, the lower the saturation. That said, there is a somewhat linear trend in the graph, especially when ignoring the two outliers seen on the left graph.

Another variable I looked at was the total residential value in each county. Because this is total residential value, it is actually a significant representation of population weighted by house value. I chose to do this as this was the only metric of income I could access and I was interested to see how income would perform as a feature. The combination of population and house value made for a significant predictor as illustrated below.

The final and most important variable I was able to engineer was liquor consumption per person in each county. While raw population is useful, we are also very much concerned with how much a person in that county consumes. There could be a highly populated area, however, if that area sports a highly religious population, consumption for each person will likely be relatively low. Conversely, if a town has a major university, each person will likely consume at a much greater rate, on average. You can see the strength of the linear relationship between consumption per person and profit per store below.

With those three variables I was able to build a simple linear regression to predict the profit per store in each county. Below is a heat map of Iowa representing the expected profit per store based on the regression model.

Dickinson was predicted to have the highest profit per store at \$161,297 followed by Polk county at \$137,861. Dickinson seems a bit out of place relative to the neighboring counties, however, quick research indicated that Dickinson county is home to popular vacation spots for local Iowans. Johnson county, home of Iowa University finished third with a predicted profit of \$119,782 per store.

As you’ve seen, the research focused more broadly on overall liquor market conditions by county. It’s important to note that these profit estimates don’t take into account overhead, a major factor of business profitability. Because real estate can be street specific, estimating rent costs at the county level would be dangerous. The analysis provided is best served as a guide to locate general areas that appear conducive to profitability. Further analysis must revolve around on site scouting of locations and concrete rent prices.