April 28: Starbucks Vs. Dunkin' Donuts


As we transition to our next series, we are making time for a few maps that do not fit into a series, but have interesting information.  Last week, we looked at dry cleaning services as an indicator of economic activity.  This week we are looking at coffee...



Starbucks left Downtown Newark in 2008, but returned in September 2015.  As you can see from the maps from 2015, before the arrival of the new Starbucks, there is a donut hole (pardon the pun) in Newark; in parts of Newark, less than 10% of residents have been to a Starbucks in the last 6 months.  Dunkin’ Donuts, on the other hand, has market penetration spread across Newark and the rest of Essex County.  But Starbucks has deeper market penetration in the suburbs than Dunkin, with over 25% of the population reporting that they visited a Starbucks in the past month.  Yet there is no Starbucks location in suburban Essex County.  Why would there be more Starbucks visits in suburban Essex County, where there is no store, than in urban Newark, which has at least one in Gateway Center?



One possibility is that because urban residents drive less and own fewer cars, they cannot get out and visit Starbucks as easily as suburbanites.  If this is the reason, the arrival of a new store in Downtown Newark may cause Starbucks usage to go up in that area.


Another possibility is that wealthy suburbanites have more disposable income to spend on Starbucks coffee, which is more expensive than other brands.  Coffee is a luxury item and not everyone can afford to buy a premium drink.  In this case, a new store would not cause an increase in use.


It could be that urban Newark residents just don’t prefer Starbucks coffee, even if there was a convenient store, perhaps because they don’t like the taste or the brand.  Starbucks generally more popular with suburbanites in Essex County.  Dunkin’ Donuts, by contrast, seems to be enjoyed by about 20% of urban and suburban residents alike.   If Newark residents don’t prefer Starbucks, a new store would not necessarily make Starbucks more popular.


One way to test these hypotheses is to compare Newark’s population to similar populations in other cities. ESRI, a geospatial data and software company, has a system that classifies the United States into lifestyle groups called “Tapestries.”  Newark’s urban population is classified as “Midtown Singles” in the ESRI Tapestry system.  The suburbanites in Essex County are classified as the “Affluent Estates.”  For this analysis, we created two searches for the two tapestry categories in the areas surrounding NYC, from as far west as Gladstone in New Jersey to as far east as Levittown in Long Island, and as far north as Stamford, Connecticut and as far south as Princeton, New Jersey.  Looking across the region, 10.14% of residents in the Midtown Singles zip codes go to Starbucks, and 22.45% of people in the Affluent Estates category go to Starbucks.  The percentage of Starbucks drinkers in the Midtown Singles demographic is 9.9%, very close to the 10.4% for the region.  Thus the disparity in coffee drinkers is not caused only by a lack of a store; urban residents visit Starbucks less than they do Dunkin’ Donuts.  If Newark is similar to other urban areas in the region, it is unlikely that a new store will have an impact on this divide.


Who is visiting this new Starbucks location, Newark residents or commuters?  Will Newark become a Starbucks town, or stick with Dunkin’ Donuts?  Leave your thoughts in the comments.


Analysis by Steve Scott