"… perhaps the queen of all proverbs. The road to hell is paved with good intentions." So said Richard Chenevix Trench long ago, and he may well have been talking about Starbucks and their lean deployment. For that matter, he may well have been talking about the vast majority of companies that pursue lean.
Most of the time when I link to an article I don't really expect anyone to read it. I just put the link there to prove I'm not making this stuff up. This time is different, however. I really hope you read the whole article on the lean foibles at the coffee company.
"The company has deployed a "lean team" to study every move its baristas make in order to shave seconds off each order." I'm sure Starbucks had good intentions when they decides to send lean teams into the coffee joints to shape things up; and I'm sure the lean 'experts' who trained and led the teams had every intention of using lean thinking and lean techniques to make things better for everyone. They all missed the point, though, and now they are grappling with a false and wholly unnecessary apparent conflict between cost and quality.
Lean tools such as 5S and standardized work – the two lean things Starbucks seemed to have cherry-picked to built their lean deployment around – were designed to compress the cycle time of the process, and thereby drive out unnecessary waste and cost. The value stream map of the process to improve does not follow the customer – it should follow the product – the coffee beans and the milk, in their case. Customer experience mapping is an entirely different deal.
It doesn't look as though they even followed the customer, however. Rather, they jumped right into the direct labor element – the 30-60 seconds of the process where some guy actually makes a cuppa joe for someone. Then they used a couple of lean tools that are really not the exclusive purview of lean at all.
Lean manufacturing and good old fashioned industrial engineering are hardly mutually exclusive. Moving the bins of coffee beans from below the counter to a more convenient place may well be something someone does in the course of a 5S effort and creating standard work. It is also something a classical industrial engineer with a stopwatch would have done in 1953. It is a lot like sweeping the floor as part of a 5S effort. Just because you use a broom in the course of the 5S, don't make the mistake of thinking Toyota and Shigeo Shingo invented brooms.
The tried and true IE techniques that trace back to old FW Taylor and the systematic abuse of people were created and perfected to support the old business model centering around direct labor optimization.
Lean techniques were developed and are still being perfected in support of the Lean Business Model centering on flow. The woods are full of 'experts' who don't really understand the difference, and they are full of business leaders who either don't understand the difference either, or who refuse to let go of their old labor cost centered understanding of business economics. Those folks pursue lean in that gray overlap area in the diagram above. It enables them to look lean and talk lean without really letting go of the old manufacturing paradigms.
What Starbucks did was hardly lean at all. It was just focusing on the small part of lean that fits with old school thinking. Starbucks is making more money not because of this misapplication of lean, but because it reduces its fixed capacity (number of stores), rolled out new products and provided a better value for its customers with the additional perks in the loyalty card program. Their lean effort didn't help them at all. It only created the cost versus quality dilemna described in the article.
So what should they do now? Truly understanding and pursuing the Lean Business Model is the best thing they can do in the long term. In the short term my advice is to get rid of the 'lean teams' and buy everyone a copy of The Goal. They are not in a lean situation at all – they are just an old time business model with a constraint problem.
Tom Southworth says
I did read the article…and just shook my head. Speaking of Shook (John), wasn’t he working with Starbucks a while back? I don’t see any mention of his name; perhaps he just got fed up with this L.A.M.E. attempt at Lean.
Bob Emiliani says
Bill – I agree with your analysis that “What Starbucks did was hardly lean at all.”
But regarding “…old FW Taylor and the systematic abuse of people…” This is an inaccurate characterization of Taylor’s intentions and of his work. Abuse of people was the result of other narrow-minded managers and unscrupulous consultants who misunderstood and misapplied Scientific Management and turned it into a mean-spirited, zero-sum exercise in maximizing efficiencies at workers’ expense. The parallels to modern-day Fake Lean management are stunning.
In 1913, Morris Cooke, a close colleague of Taylor’s wrote “The Spirit and Social Significance of Scientific Management.” It is a wonderful description of Scientific Management in which Cooke makes clear the importance of the worker and how they must be respected. It is a must-read for Lean practitioners, in part to better understand the lineage of Lean management. The paper can be found here at http://www.bobemiliani.com/oddsnends/morris_cooke.pdf
The best description of Taylor’s work can be found in his testimony to Congress 25-30 January 1912. It can be found in the book Scientific Management: Comprising Shop Management, Principles of Scientific Management, Testimony Before the House Committee, F.W. Taylor, with foreword by Harlow S. Person, Harper & Brothers Publishers, New York, NY, 1947.
In it, Taylor said: “It ceases to be scientific management the moment it is used for bad.” Ditto for Lean.
Bill Waddell says
Bob,
I think you are being a little bit too forgiving of Taylor. He wrote of a gamut of men ranging from “First Rate” at the top of the ladder down to “Stupid Men” at the bottom; and that pig iron, for instance, should be handled by a man “so stupid and so phlegmatic that he more nearly resembles in his mental make-up the ox”; a man “so stupid the word ‘percentage’ has no meaning to him, and he must consequently be trained by a man more intelligent than himself.” He wrote that “dirt handling” should be assigned to Hungarians and Italians. This hardly smacks of someone deeply imbued with respect for workers.
Regardless of Taylor’s later attempts to redefine himself when labor so passionately rebelled against his ideas that the Congressional hearings you cited became necessary to consider whether Scientific Management should be outlawed, everything I found out about him from his own pen convinced me that he was an arrogant, self-serving bastard of the highest order.
In any event, I agree with you that, regardless of his beliefs and intentions, it was the latter day application – or misapplication – of his principles that did the greater harm.
Bill
david foster says
Someone told me that a kind of coffee cake or such was frequently out of stock at his local Starbucks. When he asked them why, the guy said the store had no discretion over re-orders but that the product was pushed to them by some sort of centralized inventory control system.
If this is true, it’s really bad. One can imagine degrees of badness–a system which assumes a consistent product mix across stores being very bad, one that attempts to “learn” local patterns but not allowing manual override being a little less-bad. But there’s really no excuse for not allowing the store manager to override the system when particular products have a local following.
Chris says
Bill, you might want to consider whether the author of the article really paid attention and got the details of the implementation right. From what I’ve heard they DID include real baristas in the initial testing of the new process. Furthermore, my wife works at SBUX and was recently trained on the new method. Part of the training was to make 6 drinks the old way, then 6 the new way – letting the barista experience the difference in the process and see that indeed the drink to drink time decreases to a regular drum beat of around 45 seconds immediately after the first drink is completed. She was both impressed and a bit disgusted with the training – it proved the new process to her, but wasted 12 drinks!
Pete says
I read the article at the WSJ. It reminded me of the analogy between the “experts from corporate” and seagulls. Both fly in, eat your lunch, crap all over everything and then fly on to somewhere else. Nobody wins.
Tom says
To build on a prior comment, I would serious challenge your assumption that the journalist in this case got the story right. We’ve seen her mangle lean in her previous piece, and she continues to do so in this article. It’s a fair assumption that she is as faithful to the facts about how SBUX proposed improvements as she is about the nature of lean. This is not to defend Starbucks per se; but to suggest that we would all do better to spend more time on the gemba of this problem, since there are few apparent facts in the Journal article.
Martin_B says
I read the article and I don’t see anything wrong with what Starbucks is trying to do.
They are, firstly, trying to make their product at a certain level of quality, consistent across all stores. Secondly, they are trying to shave cycle times, but not at the expense of quality.
How else are they supposed to do it if not with experts with stopwatches?
If Toyota ran Starbucks they’d give baristas lifetime employment and encourage them to continually improve their cycle time, while adhering to very strict quality standards.
Problem is, baristas are youngsters and turn over several times a year (just guessing), so the Toyota method is a non-starter. A short training course in standardised methods is the logical solution.
FW Taylor gets a bad rap on this blog. You can’t condemn him because he expressed himself in the language typical of a hundred years ago. If he was writing today I’m sure he’d be more PC. And the workers of those days were often unschooled farm hands and immigrants, unlike industrial workers of today.
Where do today’s youngsters joining the work force acquire skills with wrenches and power tools and computers? By messing up Dad’s tools and equipment. But people who grew up in rural households where a kerosene lantern was considered high tech don’t have a clue about such things. Unless you’ve worked with such people you have no idea how badly they need training in the simplest skills.
And unless you do the work yourself, you have no idea how badly designed some tools and workplaces are. An uneducated worker won’t tell an educated manager his problems or ideas — he’s too afraid of being put down or made to look stupid. Employing a third-party expert to study the situation isn’t a bad idea.
Mark Graban says
All other comments aside, Starbucks didn’t just jump immediately to the store direct labor. My understanding is that they’ve been working on the supply chain to the store first (so they’ll quit running out of particular flavored syrups, etc.)
The article in the WSJ makes this seem more like Tayloristic industrial engineering than lean, but the writer could have it wrong. She had leaked documents and talked to a few baristas, including one whose store isn’t using this method yet.
My blog post on this:
http://www.leanblog.org/2010/10/lean-at-starbucks-or-tayloristic-industrial-engineering/
It’s also worth revisiting, maybe, that John Shook wrote about this last year:
http://www.lean.org/shook/ColumnArchive.cfm?y=2009#Col1085
Bill Waddell says
Mark, I’m sure the reporter got quite a bit wrong and/or out of context – seems to be the norm for journalists these days. That said, the layoff-happy management at Starbucks isn’t making much of a case for their understanding of the basic principles of lean.
Metal Stamping Man says
I caught that article on another site. It was quite entertaining. All I know is, is that as long as I get my frappuccino quickly and double blended… I’m a happy camper.