Starbucks Forecasting Labor Needs

The Roman emperor Claudianus purportedly said: “Change or Die.” In many ways, this epitomizes this week’s readings on forecasting, as the most appropriate forecasting processed not only depend on being continually reviewed, honed and improved, but also on responding to change (Prater & Whitehead).  My entry explores what can happen when companies over-rely on the time-series forecasting model.

The one company that continually struck me as being both successful and in constant procedural flux was Starbucks. However, Starbucks has recently come under fire for its inability to consistently create workable schedules for all of its employees, and has had to re-evaluate its labor forecasting methods (Rooney).  

Ruggless’ article “Restaurant Operations Watch: The dark side of automated scheduling” posits that overreliance on automated scheduling systems – which take into account time series data and employee availability – has caused this, despite the ability of individual managers to change the generated schedule.  Simply put, the qualitative aspect mentioned by both Prater & Whitehead and Wheelwright & Winslow is missing.

I find this fascinating, as Starbucks has long prided itself on its stores delivering the “Third Space” (Hanna), a space that is between work and home – a space that customers want to come back to. In essence, when you are at Starbucks, you are not only purchasing a beverage – you are purchasing an experience.

Presumably, the creation of this third space is dependent on having happy workers, as such a big part of creating the third space is connected to service delivery. However, this is hard to create, if baristas work the infamous “clopens” mentioned by Ruggless, as these can result in tired and “grumpy” workers, which in turn affects the end product.

It is clear that it would have been in the best interest of the company to have processes in place to train managers to take into account all four forecasting methodologies mentioned by Prater & Whitehead – qualitative, causal, time series and simulation – long before these articles surfaced.  It is my belief that Starbucks upper management (company level) were aware of the value of appropriate forecasting, as evinced by Jargon’s piece on the lean work-flow process for crafting beverages and by Hanna’s article on Starbucks’ (inclusive) company culture and strategy, prior to the media hubbub.  

My question then becomes, if upper management is aware of the value of forecasting and of creating an inclusive company culture, how can there be such a big disconnect in communicating this at the store level, when other efforts – such as the “no more than two beverage at a time” (Jargon) – have been successful?

 

References:

Hanna, Julia. "Starbucks, Reinvented: A Seven-Year Study On Schultz, Strategy And Reinventing A Brilliant Brand." Forbes. Forbes Magazine, 25 Aug. 2014. Web. 2 Sept. 2014. <http://www.forbes.com/sites/hbsworkingknowledge/2014/08/25/starbucks-reinvented/>.

Jargon, Julie. "At Starbucks, Baristas Told No More Than Two Drinks." The Wall Street Journal. Dow Jones & Company, 13 Oct. 2010. Web. 2 Sept. 2014. <http://online.wsj.com/news/articles/SB10001424052748704164004575548403514060736>.

Prater, Edmund, and Kim Whitehead. "Forecasting." An introduction to supply chain management: a global supply chain support perspective. Harvard Business Publishing - Digital Chapter, 2013. 47-57.

Ruggless, Ron. "Restaurant Operations Watch: The dark side of automated scheduling." Starbucks to revise automated scheduling procedures. Nation's Restaurant News, 15 Aug. 2014. Web. 2 Sept. 2014. <http://nrn.com/hr-training/restaurant-operations-watch-dark-side-automated-scheduling>.

Rooney, Ben. "Starbucks to ease scheduling hardships for baristas." CNNMoney. Cable News Network, 14 Aug. 2014. Web. 2 Sept. 2014. <http://money.cnn.com/2014/08/14/news/companies/starbucks-schedule-changes/index.html>.

Wheelwright, Steven C., and Ann B. Winslow. "Forecasting." Operations Management. Online: Harvard Business Publishing – Digital Chapter, 2013. 3-34.