Action and reaction

How do companies learn?

We've discussed this question on and off for years, and mostly it's not easy. Two years ago, in the middle of a series of posts on Boeing, I wrote about data falsification scandals at Toyota and suggested that it can be phenomenally difficult for a company to learn new attitudes. Two weeks later I backed off to say that it's possible but not easy. And just last month I wrote about Robert Cole's study, Managing Quality Fads, about how American companies learned quality methods from the Japanese during the 1980's. So yes, it's possible but not easy. 

That conclusion just brings us back to the original question: If it's possible, then how? And here it can be useful to look at success stories—companies that did learn a lot from others, and improved their operations accordingly. One advantage of Cole's book is that he gives a lot of detailed data about the period he describes, including a number of case studies. Of these, the first big success in adopting the new quality methods was Florida Power and Light (or FPL).

What made their success possible?

Cole sets the stage by explaining: "FPL [began] its quality improvement ... in 1981. In 1985, it became the first large American company ... [to start] learning directly from the Japanese. It entered into a deep exchange relationship with Kansai Electric and other ... Japanese companies[,] and worked closely with Japanese professors...."* But there were a number of special factors that supported FPL's quality initiative.

• FPL was in the same line of work as Kansai Electric, but they were not direct competitors. So neither company feared accidentally revealing trade secrets to the other.
• FPL and Kansai Electric used the same equipment—they both used Westinghouse-designed nuclear plants. So FPL trusted Kansai's reported metrics, rather than waving them aside as "apples and oranges." When they learned that Kansai's Mihama plant reported 10% as many quality problems as FPL's Turkey Point plant, they looked right away for management and operational differences that could explain such different performance from the same infrastructure.
• FPL and Kansai set up an extensive program of travel and visitation, so that personnel from each company could build personal contacts with their counterparts.
• Crucially, the CEOs of the two companies built a strong personal rapport. It soon became flatly unacceptable to express anti-Japanese sentiments at FPL.**   

What did they do?

With this background as a framework, FPL adopted Kansai's Total Quality Control (TQC) measures almost intact. The successes were dramatic enough that four years later (in 1989), FPL was the first non-Japanese company to win the Deming Quality prize.

Perhaps more to the point, FPL's success got the attention of other American companies. Soon FPL was giving seminars on quality improvement; and in 1986 they spun off a subsidiary (Qualtec) as a for-profit consulting firm. In the end, Qualtec could not keep up with the demand for its services, so FPL sold it to the Marshall Group in 1995.

What happened next?

Newton's Third Law of Motion famously says that to every action corresponds a reaction, and I'm not the first author to apply the same phrase metaphorically to human organizations as well. In 1989, James Broadhead took over as FPL's new CEO; and after the company won the Deming Prize, he began to dismantle much of the quality superstructure and documentation requirements. He argued that this structure and these requirements were too extensive to be cost-effective, and therefore that they were inconsistent with FPL's overall business objectives.

But it is important to understand that Broadhead did not rip out the company's quality measures root and branch. And FPL's overall quality levels did not suffer under his administration.*** What he did was to integrate the quality measures into normal operations, so that they became "just how we do things" rather than adding them as special steps administered by the Quality Department. As that integration took place, he could also shrink the size of the Quality Department because there was no longer a need for so many extra staff. He found that FPL had implemented a lot of documentation requirements because they were demanded by the Deming Prize; but in fact, no one ever looked at the documents again after the prize committee left. So he eliminated those requirements that had no use. He continued to support benchmarking, self-managing teams, process reengineering, and employee empowerment. So while the American business press described Broadhead's administration as a massive rejection of Japanese quality principles, it was nothing of the kind. He kept what worked—all the measures that reduced downtime and outages and disruption of service—because they worked! All these measures were profitable. He just made them unremarkable enough that they no longer drew special attention.  

Points to take away

The original question was, How do companies learn? So concretely it is fair to ask, How much of FPL's experience can apply to another company?

Obviously FPL faced special circumstances that gave it almost a perfect framework for learning from someone else: for starters, the partnership with Kansai, who was not a competitor and who used exactly the same equipment.

But I think there are three points that have a wider application, and it is important to remember all three.

1. Any learning initiative—really any change initiative of any kind—has to have strong and consistent executive support. The forces opposing change are always powerful; so without reliable pressure from the top, it is likely that nothing will happen.
2. 
   The more widely you spread information about the initiative or the new way of thinking—and the more personal you make it—the better your success will be. One of FPL's big successes was their travel initiative, that brought so many of their employees face-to-face with their counterparts at Kansai. Once they made friends, it was easier to talk about problems at work—and to listen to the answer.
3. Unless there are external measures to guarantee the change is permanent (for example, if one company has bought another), there will be a backlash in time. Don't waste effort trying to fight or avoid it. The better strategy is to make sure that all the most important parts of the change are fully embedded into normal practice by the time that the backlash arrives. Let the new managers pull down some posters or eliminate other symbolic reminders of the change initiative. But make sure the new methods themselves are so routine and so profitable that no one thinks of changing them.   

If you remember all three points, is that enough to make it easy for your company to learn? Probably not. But they are sure to help.

__________

* Robert E. Cole, Managing Quality Fads: How American Business Learned to Play the Quality Game (New York, Oxford: Oxford University Press, 1999), p. 66. This whole story comes from the same work, pp. 66-71. 

** If I remember the early 1980's correctly, FPL's attitude could not yet be reliably assumed across the country as a whole. 

*** So James Broadhead's career at FPL does not foreshadow that of (let's say) Harry Stonecipher at Boeing, a few years later.   

           

"Hardening in the field"

Years ago, I worked for a small tech startup. We were scrappy and energetic, and we hadn't quite decided how finished a new product had to be before we could ship it. 

• Did it have to be bug-free? There were always more bugs. 
• How about if it had No serious bugs? That sounds nice, but what counts as "serious"?

In all these discussions, our head of engineering usually wanted to ship now and not later. (Of course he also saw the financial statements, and knew that we needed the revenue!) His argument was that if the product was basically good enough, then what we needed was for it to operate in a real-world environment so that we could identify which remaining defects really mattered. Then when we fixed those, the product would be ready. Other nominal defects might exist, but they would be merely cosmetic. He called this process "hardening in the field."

"Let's see: eggs, cheese, filling. I guess it's ready to serve!"
Or maybe not.

One of our project managers remarked that shipping a product with the hope that it will get better at the customer site is like a restaurant serving up raw ingredients and then hoping that the meal will get fully cooked once it's at the table. But the idea isn't quite as bad as that. In fact, this line of reasoning is exactly why the tech industry introduced the concept of beta testing. Admittedly it is a dirty trick to ship beta-quality product to a paying customer who expects something finished. But companies frequently do need to see their products operate in a real-world environment, and some customers are so eager for new technology that they will accept the risk that the beta product might fail unpredictably. Once my startup matured enough to establish regular beta programs for our new releases, we stopped talking about "hardening in the field."  

"But wait—this is OK?"

So companies developing new products face competing demands. The need for real-world data pushes them to release sooner; customer expectations about those products may push them to wait until the basics are solid. I assume that nobody will release a beta-version automobile whose brakes don't work yet. (Though I might be wrong about that. See for example the discussion in this post, and the linked news articles.) Likewise most restaurants won't serve uncooked food, unless the customer ordered sashimi or carpaccio. But the high tech market is more confusing, because the expectations conflict.

Rapid innovation is a more or less constant feature of the high tech market landscape. Everybody knows that brand-new implementations of new technology are usually full of bugs; stable, reliable implementations take longer. So what do you do? Partly it depends on the inherent risks of the exact product you are designing. Is it a car or a rocket that can hurt people if it fails? Or is it a toy, where failure will just disappoint them? Is it easy to recover? And what does the regulatory environment look like? Obviously you have to take account of all these factors.

Beyond those factors, though, you may just have to decide where you want your organization to fit in the ecosystem of high-tech products: do you want to be first to market with innovative technology, or are you willing to trade speed of innovation for product reliability?

And then, if possible, you would like to design your Quality Management System so that it supports your decision—so that it nudges you into being the kind of company you want to be.

If it is important for you to be first to market, you should measure your development process with KPIs that track (among other things) how fast new releases reach the field. Since your initial releases are likely to be buggy, your customer support process should monitor KPIs that track the speed with which customer issues are resolved. You may wish to implement an Agile development model, or offer customers the opportunity to work with you as partners in exchange for providing their feedback as active members of the development process.

Conversely, if it is more important to you that your products be fully reliable before they reach a customer, then you should not measure speed of delivery as one of your development KPIs. What you measure is what you optimize; if you are willing to sacrifice speed for reliability, don’t measure speed. In this case, you are more likely to set metrics around the extent and comprehensiveness of testing, and the number of known bugs at time of release. You might also choose to use a waterfall development model (instead of an Agile one) so that testing is done on one version at a time, thus reducing the number of variables in the development process and presumably some quantum of risk. 

It's interesting to realize that "Quality" doesn't always mean the same thing—or rather, that it can mean two different things (in this case both speed of innovation and reliability of performance) which are incompatible, and which you have to choose between. And that single choice can have ripple effects across your metrics, your processes, and your strategy.

           

The fight to define Quality

Five years ago, in one of my first posts for this blog, I argued that all the definitions of Quality that have been proposed in the literature mean effectively the same thing. Whether you talk about "conformance to requirements," or "fitness for use," or "excellence in goods and services," I argued that fundamentally you are talking about getting what you want out of those goods or services. So I proposed the umbrella-definition, "Quality means getting what you want."¹ And at a workaday level, I still think that's fair. 

But it turns out there were reasons behind the fight over a definition, reasons besides academic posturing. Robert Cole explains in his Managing Quality Fads (a book that I discussed two weeks ago) that American businesses saw a profound shift during the late 1970's and early 1980's in the understanding of Quality, and the different definitions served as banners or rallying cries for the contending models. So while they may all mean (more or less) the same thing today, they meant very different things in the past. Thus it can be useful to understand what these definitions meant at the time, and what the parties were fighting over.

Old model

Because we know how things turned out in the end, some commentators have blamed American managers for failing to adopt Japanese Quality methods faster than they did.² But Cole explains clearly that such a view is profoundly unhistorical. It is simply not true to allege that American managers were stupid (most of them clearly were not) or that they were ignorant about Quality. What is true is that they understood a great deal about Quality according to a framework that Cole calls "the old model"—but this framework proved unsuccessful in head-to-head competition with the "new model" practiced by many Japanese companies. So before they could learn the new model, they had to carry out two preliminary tasks first: they had to understand that there are more than one way to think about Quality; and then they had to unlearn much of what they already knew.³ 

What was this "old model"? It was a way of thinking about manufacturing processes that had been responsible for a string of remarkable successes, starting with the industrialization of the American economy, and culminating—within the living memory of many managers still working in the 1970's and 1980's—with unconditional victory in World War Two as the "arsenal of democracy." And the basic principles behind the model are plausible:

• 

  Cole's table of "quality-hostile
  assumptions"⁴ (Click to enlarge.)

  Economic success comes from the division of labor, where everyone can focus on doing his own work as well as possible. This means that engineers should design products, managers should make decisions, and workers should do what they are told. Anything else means inefficiency and poor performance.
• For most jobs, there's pretty much just one way to do them. Once you've figured out what that procedure is, and have implemented it, there's nothing more to do. Once a procedure is defined and in place, the only way to make it less expensive is to pay your workers less. (If your workers are unionized, that means moving overseas to a cheaper country.)⁵
• These two facts—the division of labor, and the one-and-only way to do most jobs—mean that once an engineer has finished creating a design, most of the important decisions have already been made. At that point, the only degree of variability remaining is in how strictly the design is implemented, and in how carefully the procedure is followed. Stick to the design, and you'll have a good product; deviate, and you won't.
• Products have to be "good enough," which means there is a "quality floor" beneath which they must not be allowed to sink. But no customer is going to pay for a gold-plated shovel, so it is foolish to throw money after a chimerical quest for perfection.⁶
• We know what customers want. We've seen time and again that they want products they don't already have. So if we make the products—and history shows that we are fantastically productive at making things!—customers will buy them.  

There's more, of course. And I have included a summary of what Cole calls "quality-hostile assumptions" in the inset picture above. But it is important to understand that these assumptions were not stupid! In the event, they were indeed overturned by the new model of Quality. But each of the principles I listed has a superficial plausibility to it. And the overall model had been so successful in recent history that it looked like madness to question it.⁷ 

New model

The new model of Quality reversed many of the assumptions of the old model. In some ways this is unsurprising, because it grew out of a historical situation that was nearly the opposite of the confident, successful American experience. Japan, after all, lost the Second World War, and her industrial plant had been largely reduced to rubble by American bombers. Japanese management responded to this catastrophe by mobilizing all Japanese—and in particular all employees—to support a collective effort to rebuild and regrow. In the face of a national calamity, everyone was asked to contribute, in any way they could.⁸ 

Cole's table comparing the
old and new quality models.¹³
(Click to enlarge.)

But if everyone is asked to contribute, there is no room for a narrow-minded adherence to the division of labor. Naturally the engineer knows a very great deal about how to design a product; but he may not know everything about the particular set of machines and tools that we have to do the job here and now. Teams on the factory floor therefore help improve both the design implementation and the manufacturing processes.⁹

Early consequences are that industrial processes are never permanently fixed. They are always subject to improvement.¹⁰ Also, individual employees are not locked into a single role forever: all employees are assumed to be learners and potential problem-solvers.¹¹ 

And then feedback effects begin to appear. Improved methods mean less rework. But less rework means less cost, so high-quality manufacturing processes can be cheaper than low-quality ones—in direct contradiction to the expectations of the old model.¹²

The old model argued that continual improvement would run aground on the law of diminishing returns: it would take ever more cost and effort to make ever smaller improvements in anything. But the new model understood that there are always many targets for improvement. You can make this aspect of the job a little better, and then turn around to improve a different one. Briefly, you can make improvement continual by regularly picking new things to improve.¹⁴ 

What's more, it turns out that customers are willing to pay more for products that work better and last longer! So an emphasis on superior product quality entails greater market share as well as lower costs. In other words, increased product quality drives improved corporate profitability. And an overall focus on customer satisfaction provides an effective framework to organize all these other efforts.¹⁵   

What about the definitions?

With this background, it is easier to see the competing definitions of Quality as slogans or ideological commitments to one model or the other.

The most obvious of these is Philip Crosby's definition: Quality is conformance to requirements.¹⁶ This definition is foundational to the old model. Remember, in that model the engineer designs the product and makes all the important decisions; once the design is complete, the best outcome is produced by conforming exactly to the design and the process ... in other words, to the requirements. Nothing more is needed in order to achieve Quality, and nothing more is wanted.

ISO 9000:2015 agrees with Crosby, in definition 3.6.2: "degree to which a set of inherent characteristics (3.10.1) of an object (3.6.1) fulfils requirements (3.6.4)."

The other pole is represented by Joseph Juran's definition: Quality is fitness for use. "Use" is what the customer does with the product or service, and "fitness" just means that it works. So Juran's definition means, in effect, that Quality is whatever the customer says it is. This is an extreme statement of the new model that Quality depends on customer satisfaction.

W. Edwards Deming and the American Society for Quality try to bridge the gap between the models. Deming calls out both sides by saying, "Good quality means a predictable degree of uniformity and depend­abil­ity [old model] with a quality standard suited to the customer [new model]." The ASQ, for their part, define that, "Quality denotes an excellence in goods and services, especially to the degree they conform to requirements [old model] and satisfy customers [new model]."

In any event, the battles are over. For the time being, the new model of Quality has won the day. So we can afford the magnanimous conviction that everyone who talks about Quality probably means more or less the same thing. But I think it can be valuable to understand how we got where we are.

Gosh, in all this discussion I never explained my own position. In case it's not obvious from everything else I have written here for the last five years, I throw my support firmly behind the new model. There have been too many cases where a company or an organization worked out a detailed plan, executed it flawlessly, and the results were just no good for anyone. (*cough* Edsel! *cough*) So I think that Quality ultimately has to mean goodness—and that for something to be good, it must (among other things) be good for someone in particular.¹⁷ Therefore I take the side of the customer in this debate between models.

At a theoretical level, I'm fond of the simplicity of the "innocuous truism" formulated by Robert Pirsig: "Quality is what you like."¹⁸ You can probably see the influence of Pirsig in my own definition at the top of this essay.

__________

¹ More exactly, I went on to explain, "Whether you are talking with your manager or your auditor or a technician on the line, if there's a question about the relevance of this or that element of the system just ask, 'Are we getting what we want? Do we need this element in order to make sure we continue to get what we want?' If yes, the element belongs in your QMS .... If no, not." 

² See, for example, this famous satirical pseudo-advertisement. 

³ In fact, Cole argues statistically that workers with deep backgrounds in traditional quality control functions were mostly passed over when companies began to create Vice President of Quality positions in the 1980's and 1990's. It was easier for their companies to "learn" new methods by promoting new men than it was for the employees themselves. "[The] overwhelming majority of the 174 vice presidents for quality did not have a career in quality prior to assuming their vice presidential role.... [Of] the Fortune 500 quality leaders surveyed, only 9% had quality responsibilities in their third previous job." Robert E. Cole, Managing Quality Fads: How American Business Learned to Play the Quality Game (New York, Oxford: Oxfor University Press, 1999), p. 44. 

⁴ Ibid., p. 76.

⁵ Ibid., p. 50.

⁶ Ibid., p. 47.

⁷ Ibid., p. 48.

⁸ "Various authors have noted that Japanese companies tend to magnify even small challenges as a strategy to mobilize all employees on behalf of aggressive new corporatewide goals .... [One] can suggest that large growth-oriented Japanese manufacturing firms throughout most of the post-World War II period have been characterized by a weakness orientation, which is designed to reveal problems in current performance to a broad range of employees. By contrast, comparable American firms throughout much of the postwar period were characterized more by a strength orientation." Ibid., p. 55. 

⁹ Ibid., p. 30. 

¹⁰ Ibid., p. 31.

¹¹ Ibid., p. 30.

¹² Ibid.

¹³ Ibid., p. 26.

¹⁴ Ibid., pp. 77-78.

¹⁵ Ibid., pp. 29-30.

¹⁶ References for this and the following definitions are given in this blog post. 

¹⁷ Cf., for example, William Blake, Jerusalem, plate 55, lines 60-63: "He who would do good to another, must do it in Minute Particulars, | General Good is the plea of the scoundrel, hypocrite, & flatterer: | For Art & Science cannot exist but in minutely organised Particulars, | And not in generalising Demonstrations of the Rational Power."

¹⁸ Robert Pirsig, Zen and the Art of Motorcycle Maintenance: An Inquiry into Values (New York: William Morrow & Co., 1974, 1999), p. 232.