When John and Scott went into Starbucks stores to actually see how work was done, what they encountered was somewhat of revelation to both of them. John, an independent consultant for Starbucks at the time, remembers the chaos of the store backrooms and managers hauling materials in their cars from one store to another. Scott, who was VP of Strategy, had shared with John some information he had learned through many store visits: 15% of the time a store would be out of brewed coffee (three varieties were offered daily), which put baristas in a “moral dilemma” of how to satisfy the customer (attend to the brewing process when a timer chimed or attend to the line of customers in front of you?). And there was approximately $50 million in actual coffee waste each year.
Corporate was tracking an accounting proxy for actual waste termed “coffee variance.” It was a derived dollar value of waste given sales and many assumptions like average batch size and a “standard” waste amount vs. actual amount of brewed coffee used as revealed through periodic inventory counts. No one really understood what this accounting measure represented other than an expert in accounting who shared with Scott when he investigated the measure how inaccurate it was for any store. That didn’t stop field leaders and executives from using the measure to evaluate performance and make decisions.
“Although I was at the time a store manager, running the store, spending 80% of my time on the floor as a barista serving customers, I was totally oblivious to all those problems in the work,” said Josh. Baristas tended to deal with stockouts as best they could. Not until those on the front lines and managers learned how to see the work in an objective manner and collect facts were they motivated to solve the problems.
This story of brewed coffee offers a lot of learning for a lot of businesses, and so much of what Scott, John, and Josh discovered is still relevant today. Scott assumed the stores knew how to deal with the problem, but John encouraged him to spend more time in the shops, seeing what was occurring. The seemingly simple effort of going to gemba and observing the work provided learning that literally no one else at corporate had, and they learned the accounting construct of “coffee variance” was an incorrect way to get at the problem. Josh had been told to find problems through the P&L, and so even surrounded by the problem he was blind to it.
Once they focused their attention on the traditionally mundane brewed coffee business in the stores, they saw a lot of problems: frequent stockouts of coffee when the customer wanted it; lots of brewed coffee discarded after 30 minutes because that was the company standard as taste deteriorated; and unused ground coffee (ground each morning) for the day but not brewed and discarded at closing. One major cause of the brewed coffee outages that had gone unnoticed was the stores were mandated to use two coffee shuttles (large thermoses) for the high-volume Pike Place Roast (PPR), and dedicated one shuttle to the Bold variety and one to Decaf. There was a 30-minute count-down timer attached to every shuttle, which was started when the coffee brew cycle was initiated and ending when the coffee had “expired” (according to the Starbucks Coffee team at the Support Center). This mandated standard process resulted in many of the customer stockouts that stores experienced, but no one saw that connection.
They also saw the toll on baristas. “As we learned to focus on the work, we saw all of this physical burden,” added Scott. “Every batch, which is roughly every eight minutes, the barista had to bend down … and kneel while they scooped this coffee. If you do this five hours a day, it’s very burdensome on the body and a lot of reaching and stretching.”
They were focused on the simultaneous problems of waste and customer out-of-stocks, coffee quality for the customer, and barista physical burden. Scott called the inventory issues the “too much and not enough” problem, and since seeing it at Starbucks he now sees in virtually every retail store visit. Unlike past efforts to solve in-store problems, there would be no corporate directive this time. It was an opportunity to introduce the problem and a problem-solving process.
“We certainly did not want to just tell stores what to do because we recognized that every store was in a different situation,” said Scott. “We wanted to instead engage stores in first discovering and then solving their problems as it relates to the brewed coffee story. We also wanted to make sure our leaders and our leadership hierarchy above the stores — district manager who oversees 10 stores and regional director/manager who oversees about 100 stores — that they also fully understood and experienced these changes. Many of our district managers and regional directors had never brewed a batch of coffee before and certainly did not understand the business in great detail. We felt for them to support this learning activity and change, they needed to understand it themselves first and learn before they went and led their teams through it.”
The vehicle to engage managers and directors was Starbuck’s annual manager conference in New Orleans in 2008, with approximately 10,000 attendees. At the conference they presented the problems to the directors and managers as well as a cascading approach to solve them in ways unique to each of Starbucks thousands of stores.
Lean learning and a problem-solving process moved down from regional managers to district managers and then to store managers and store partners. Ideas moved from the frontlines back up through the managers. Out of the conference, regional managers took a few weeks and went through the waste problem themselves working with a “seed store,” brewing coffee and seeing what took place; they learned a lot, said Scott, and could think about countermeasures to implement that would alleviate problems of waste. Then they led the roughly nine other district managers in the same go-and-see process.
“This process, while taking longer than just sending out a corporate memo, proved to be very effective,” added Scott. “In a very short time we saw waste cut in half, something we had never seen before.” A coffee brewing process was emerging, one that was consistent, easy to learn, and easy to do. Grinding would occur per batch for freshness (not big-batch grinding a day’s worth of coffee in the morning), making each variety available all the time. This necessitated grinders located closer to brewing equipment and in a position that eliminated the bending and reaching. A very different approach to brewing coffee for the store teams had emerged, one that required a significant mental shift: From starting a brew cycle when a batch of coffee “expires” (chime beeps at 30 minutes) vs. brewing a different type of coffee every eight minutes on a cycle: PPR, Bold, Decaf, repeat.
The time to brew a batch fell from 67 seconds to 53 seconds. All of the changes, which were being trialed in various regions and stores, culminated in a problem-solving guide.
It was during this time that Josh had migrated to the “Starbucks lean team.” Josh said that Scott had received permission to build out the team to support the problem-solving effort just prior to the New Orleans event, and he ran an “experiment” store and took those tests across the country to a variety of store environments. Just solving the problems technically was only part of the challenge. He had to convince 19-year-olds to accept they had a problem to solve and practice and try new methods on their own (behavioral change); they simply did not believe the new approach to grinding and relocating equipment would reduce brew times.
“The real test became effectively proving that to the baristas who, as you can image, were skeptical that good ideas could come out of corporate, and for good reason they had that skepticism,” said Josh. Many of the problems that were now being taken on were designed into the work by corporate. The stores had been doing what they were told.
The objective for the lean team, according to Scott, was to create and enable problem solving at every store. “I have always believed people have great capability in their environment to make improvements. So how do we guide those improvements based on a set of principles and a process of practical problem solving that will help them get to better places. It was by design we introduced problem solving as an early thing because we recognized that was going to be an important capability that we needed as we introduced other changes for stores to experiment with.”
The improvement effort asked for substantial behavioral changes in addition to the technical process changes, said Scott. “Getting people engaged in behavior change was, in itself, a huge challenge.”
“Yes, there was a lot of learning that we got and effort we put into how to effectively change behavior. When people then became willing and did try the new behavior, it worked. The technical work changes worked. They had been well-tested and vetted… It was, in fact, a better way to brew coffee, a better way to make mochas.” Looking back at this time, Josh said since then he’s rarely seen the thoroughness of what was done at Starbucks in preparing people for both social and technical changes.
Read more articles in our Starbuck’s Coaching and Co-Learning series:
Scott’s presentation that illustrates the Starbucks coffee problems and countermeasures developed to address them.
Bring Co-Learning to Your Organization
LEI has spent decades developing coaching methodologies that create lasting transformation. Our coaches work alongside your leaders to build internal capability through structured co-learning experiences.
The Lean Management Program
Build the capability to lead and sustain Lean Enterprises
