Over a decade ago National Public Radio (NPR) posted a segment covering the Bad Barista Index (BBI), the brainchild of communications consultant (and self-proclaimed acolyte of the caffeine cathedral known as Starbucks) Mark Rovner. The BBI uses barista service at Starbucks to gauge changes in the health of the economy. It is based on the concept that when economic conditions are unfavorable (i.e. slack labor market and high unemployment) skilled professionals are forced out of their respective industries and must take lower-paying positions to make ends meet. Service industries, faced with a growing pool of talented labor, then snatch up these competent workers. Naturally, this results in overqualified baristas who can contribute high skills, a stronger work ethic and consequently a better level of service. Conversely, when the economy bounces back these overqualified workers are the first to leave the service industry to find better paying jobs. They are then replaced by an influx of less skilled and less experienced workers, temporarily resulting in a decline in the quality of service. When the economy is holding steady, current workers learn more skills in their position and quality steadily increases.
In short, if that Starbuck’s barista can fully execute your grande venti iced skinny hazelnut macchiato light ice and no whip with one pump classic and an extra shot of expresso in no time at all… the economy might not be doing so great.
Theory and Methodology
Since Rovner’s description lacked the kind of concreteness we like to see in economic development, we have gone ahead and defined the BBI for this article as an informal measure of shifts in the labor market, both positive and negative, that result in sudden changes in the quality of service sector labor.
To examine this theory, we dug into the data to see whether the BBI was borne out of statistical fact or just caffeine-deprived musings. Publicly-available Yelp ratings scraped from the profiles of multiple Starbucks within selected metropolitan statistical areas (MSAs) were used as a proxy for barista quality, which in turn served as our metric for the BBI. We utilized Local Area Unemployment (LAU) rates from the Bureau of Labor Statistics over the same time period as our measure of economic health. We hypothesized that Yelp ratings would rise in response to high unemployment and decrease in times of low unemployment.
The graphs below show the annual average Yelp rating for Starbucks from 2008 to 2016 compared to LAU rates. The green and red boxes are used to highlight times that best represent the BBI theory. Green areas highlight where the average Yelp rating increases in tandem with unemployment, signifying a glut of over-qualified baristas. Conversely the red areas highlight where unemployment declined and the average Yelp rating followed, indicating the flight of skilled workers and their subsequent replacement by inexperienced and less-skilled workers.
Interestingly, we found that there may be evidence of a relationship between the quality of baristas and the level of unemployment. In almost all cases, average Yelp ratings increased and decreased in tandem with the level of unemployment. However, it also appears that during periods of extended decreases in unemployment, average Yelp ratings tend to deviate from this relationship and increase on their own. This is potentially indicative of lower turnover resulting in employees developing new skills, which in turn benefits the overall coffee-purchasing experience.
Overall, the average Yelp ratings for Boston Starbucks emulate our BBI theory. As unemployment rises from 2008 to 2010, the average Yelp rating increases along with it. Unemployment begins to decrease in 2010 and is matched with a decrease in Yelp ratings. In 2015, the Yelp ratings increased despite decreasing unemployment.
The Manhattan Starbucks’ yelp ratings followed the BBI unemployment trend, rising when unemployment rose and declining as unemployment declined for except for during the 2014 to 2015-time period. On an unrelated note, average Yelp ratings were far lower in Manhattan than in the other studied geographies. Average ratings in the other geographies regularly surpassed 4.0 and often neared 4.5, yet in Manhattan the average rating never exceeded 3.5. Deviations in average Yelp ratings from year to year were also minimal compared to the other studied MSAs.
The period between 2008 and 2012 follows the BBI theory. However, from the three-year span from 2012 to 2015 average yelp ratings across Miami Starbucks increased despite unemployment’s continuous decline.
Unlike the other selections in this list, the data examined for the Burlington MSA was not taken from Starbucks. As it turns out, there are simply not enough Yelp users in the Burlington-South Burlington MSA who were dedicated enough to review their local Starbucks. Instead, we decided to switch the analysis up and examine a single coffee shop called Uncommon Grounds, located along the main outdoor shopping district. Strangely, the data from this geography suggests an inverse (or lagged) trend compared to the ones seen above. As the graph shows, unemployment peaks in 2009; then, a year later, Uncommon Grounds sees its lowest Yelp rating ever.
Sadly, using Yelp reviews as a measure of service quality presents multiple limitations. Yelp reviewers are self-selected and therefore potentially biased compared to the average customer. Many are also prone to overreacting in their ratings, leaning towards the maximum and minimum ratings. When a business only has 20 reviews, per se, a disgruntled customer leaving a 1-star rating can tip the scales significantly.
Secondly, Yelp reviews are not specific to service alone but rather a combination of goods and services. Thus, if a customer is not happy with the goods produced by the business, irrelevant of the worker’s ability to prepare and deliver them, the customer may still leave a bad review. Yelp does its best to not recommend reviews that are fake, unhelpful, or biased, which helps ensure the reviews posted are helpful to viewers and thus helpful for this analysis.
Importance for Economic Development
Businesses in service industries can apply the Bad Barista Index concept by basing workforce development strategies on local economic indicators such as unemployment rate. Workforce training can be altered to account for variations skill-level of new employees. This helps achieve high quality service regardless of the candidate pool skill-level. Additionally, awareness of economic conditions can allow business owners to be best prepared for employee turnover and identify the need for hiring campaigns in advance of employees leaving. For example, business owners would identify that the local economy has been declining and increased number of training hours or provide additional tutorials for onboarding employees. Integrating the BBI concept to adapt business workforce development strategies can turn that “bad barista” into a “good barista”, helping a business to achieve a low BBI regardless of local economic conditions.