Starbucks Coffee - T-Log Process Analysis

[pic 1][pic 2]

[pic 3][pic 4][pic 5]

[pic 6]

[pic 7]

Table of Contents

OverView        

Company        

Data Collection        

Process Map        

Process Analysis and Insights        

Recommendations        

General        

Design a three tier operational strategy        

Launch ‘Pre-order’ App        

Case Specific - Village        

Case Specific - Old SUB        

References        

Appendix        

Appendix 1 – Gantt chart        

Appendix 2 – Raw Data (Village)        

Appendix 3 – Raw data (SUB)        

OverView

Company

Starbucks has “5500 coffeehouses in over 50 countries”. It is a company that is commonly associated with efficiency and quick customer service. As a group we thought it would be interesting to study their process and learn how Starbucks is successful operationally, given such a large customer base. We were also curious as to why some Starbucks locations are extremely efficient while others are significantly slower. This report will focus on the process of receiving a Starbucks beverage from point of sale to the point of receipt at the University Village location. We also studied the process at the old SUB location as a point of comparison. Given the small scale of the report, we will assume that customers will only ask for the most popular beverage, standard latte with/without syrup, as opposed to other variations.  

We found that the process is a Job Shop, which means one barista handles all the activities involved in making the beverage from start to finish. Although there is a cashier who takes orders and payment, he or she is not specialized and also doubles up as a barista when needed.  

• Input:

Espresso shots from the espresso machine

Milk: Machine steams and pours milk

Syrup

• Output: Coffee

Data Collection

We collected data by visiting two Starbucks locations, UBC village and the student union building. We recorded the customer arrival/exit rate and the time it takes to complete each step in making the drink. We also spoke to employees and managers at these two Starbucks location to gather additional information.

Process Map

The flow units in the process are the orders. Based on the data we collected, we found that customers were arriving at a rate of 98/hr at the Village Starbucks - comparatively the rate at the Old SUB was much higher at 134/hr.

The process starts with the customer queuing up at the cashier. According to the raw data we collected at the SUB, it takes the cashier 37 seconds on average to take an order and to accept payment. The order (cup) is then in a buffer until a barista fills the espresso handle and queues shots. It only takes the barista 2 seconds to press the button however the queue shots activity takes 25 seconds on average. In the meantime the barista can continue with the process (this overlap is depicted in the Gantt chart). The barista then adds syrup (3 seconds) if the order requires it and steams the milk for an average of 11 seconds. The espresso and the milk is then combined (5 seconds) and delivered to the customer, which takes about 3 seconds on average.

[pic 8]

Starbucks franchises function with a base level of 2 workers at any given time, usually 1 cashier, who may also double as a barista in busy periods, and 1 full-time barista that focuses on fulfilling orders. From this base point it is easy for Starbucks to scale-up the operation by adding in additional Baristas at busy locations or during busy periods, which allows for multiple orders to be worked on simultaneously- as was with the Village location where there was 1 worker stationed at the cashier and, as it was a fairly busy weekend, 3 Baristas working on orders. Additionally, the shift manager was also available as a floating resource to provide help throughout the process - particularly as a stand-in cashier to help utilize any additional capacity within the resource pool further along the process by taking in additional orders. The other resources at work are the espresso machine, the coffee grinder and the milk steamer.

Process Analysis and Insights

At the Starbucks in the village, the bottleneck resource is the cashier with a unit load of 30 seconds putting the capacity rate of the process at 120 units/hour. On the other hand, the bottleneck resource at the old SUB location is the queue shots activity with a unit load of 36 seconds putting the capacity rate at 100 units/hour.

An important performance measure for a business like Starbucks is average queuing time. The inventory in our analysis refers to the customers waiting in line rather than the work-in-progress orders. Using the PK formula we found that there are approximately 4 people (Iq = 3.64) in the queue at all times at the Village location. Consequently, a customer visiting the Village location can expect to wait 2.23 minutes on average.  However, it is important to note that these observations were made on a Sunday morning so they are not representative of the peak hours. The graph below compares the two locations based on its performance measures.  

[pic 9]

Both the flow time and the queue time at the SUB Starbucks is higher than the Starbucks at the Village. The flow time is 135% higher at the Old Sub, whilst queue time is only 23% longer. The total time it takes for a customer to get their coffee from when they enter the queue is 296 seconds, which is almost 5 minutes, whilst it is 178 seconds at the Village, which is almost 3 minutes, which means it takes 66% longer for someone to get their coffee at the Old Sub compared with the Village.

We found 2 main reasons for the discrepancy between flow time and queue time: firstly, even though the total process was significantly more efficient at the Village compared with the Old Sub, the cashier took longer to take an order at the Village compared with the Sub (30 seconds vs. 23 seconds). Most notably, the cashier at the Village spent more time talking to the customer compared with the Sub (this may have been due to cultural or age differences), and thus became the bottleneck in the entire process.