FIVE

Early Wins, Squandered Gains

Pioneering the PD Process at Merck

IN THE SUMMER OF 2005, MERCK, long a darling of Wall Street, was feeling the heat. Its decades-long track record of introducing new blockbuster drugs had hit a five-year dry spell. One product, Vioxx, highly significant to its earnings stream, had been voluntarily withdrawn after publicity on possible side effects and FDA scrutiny.²

Merck was under enormous pressure to deliver earnings. A high stock price to corporate earnings ratio was essential to acquire promising companies with stock and use their patents to fill its shrinking pipeline. These strategic calculations were felt through the ranks. Only a handful of drugs still retained patent protection (thereby commanding high profits). Several would soon come off patent and compete with generic alternatives. High priority was assigned to grow market share to compensate for shrinking margins.

Merck Mexico had a particular problem. Poor performance of the osteoporosis “miracle drug,” Fosamax, stood out like ink stains on a white dress shirt. Mexico ranked last among all countries in which the product was sold, and management seemed incapable of correcting this performance shortfall. Headquarters called constantly. Tensions rose. Jobs were on the line.

Readers, having been exposed to a litany of intractable societal ills—from FGM eradication to malnutrition—might easily dismiss the “Fosamax problem” as an insignificant corporate hiccup. All is in the eye of the beholder. Shifting Mexico’s worst-in-class Fosamax standing had been as resistant to improvement as reducing MRSA infections in hospitals. Careers depended on it.

A far more radical challenge lurked unseen. Precisely because the problem could not be solved through business as usual, Merck Mexico would have to tackle its cultural orthodoxies and constitutional framework of top-down, command-and-control solutions. Let there be no doubt, this undertaking incurred political risk for those leading the charge. The corporate immune defense response would mobilize itself to repell the invaders at the beachhead.

Some at headquarters were casting about for outside-the-box answers. An article on positive deviance in Fast Company magazine caught the attention of Grey Warner, vice president of Merck’s pharma business in Latin America. His territory extended from Mexico to Tierra del Fuego. Warner had concluded that the pharma industry’s traditional business model was broken and needed to be reinvented. With blockbuster drugs coming off patent and generics cannibalizing revenue streams, new product introductions could no longer be counted upon. A breakthrough in sales and marketing might offer a way outside the box. Could Merck pull off a disruptive change in its relationships with hospitals and doctors and gain market share? Could the company move from being just a typical provider of pharmaceuticals to a trusted member of the health care club? In the sweltering summer of 2005, Warner posed this challenge to his extended management team and offered the idea of positive deviance as a possible way forward.

We have previously noted that an invitation is essential to the integrity of the PD process. Warner made it clear to those assembled that they could opt in or opt out. Most of those present abstained. Tentatively, the director of sales and marketing for Mexico, Andres Bruzual, expressed interest. His predicament was a layup for PD. Bruzual had exhausted every trick in Merck’s repertoire to address the Fosamax sales problem. PD beckoned as an alternative of last resort.

In February 2005 he broached the idea of pilot testing PD at the quarterly district manager meeting in Mexico City. The austere corporate meeting room was arranged in a classroom-style configuration. Bruzual, a six-foot-tall Venezuelan with a crew cut, is charismatic and tightly wired. A take-charge kind of guy, he fills the room with energy. His kickoff was distinctly out of character: “This will be driven by your passion, not by me,” he said. Quizzical exchange of glances. There was no precedent whatsoever for Bruzual deferring to the whims of subordinates or being guided by their “passion.” None could recall being invited to take the ball and run with it, and most couldn’t believe it would happen. Many were skeptical as to how this would play out. Of forty-one district managers overseeing a field sales force of over 224 local sales reps in Mexico, half volunteered to stick around and learn what PD was all about.

An orientation was held the next day, introducing the concept and its successes in the developing world. Then Bruzual concluded with an unprecedented invitation: “So there it is,” he said. “I leave it up to you to decide whether or not you want to try the PD idea and what to try it on if you choose to do so.”

“It is almost impossible to overstate the extent to which this isn’t Merck,” states David Raimondo, chief of staff for the region. “It’s a glass-half-empty place. In the face of poor sales, the conventional solution is to target something as ‘broken’ and impose a top-down get-well program.” Raimondo continues: “You couldn’t find a more unlikely personality to sponsor a process like this. Andres is central casting’s iconic manager—General Patton with a Spanish accent. His persona elicits commitment and followership. But he relented to PD’s cornerstone principles of a bottom-up process.”

That’s because Bruzual was desperate. Having taken a leap of faith, he played by the rules. He was fastidious in waiting out the long, pregnant silence that followed after his offer. Finally someone spoke. Eventually the participants came up with their own topic and how they would organize to tackle it.

Unsurprisingly, the focus of the PD initiative would be Fosamax, which, as previously noted, had proven unresponsive to a sequence of campaigns to spur sales and market share. Notwithstanding global volume of $3.2 billion, the half-life of the product was ebbing away. Its patent was expiring in 2008. Fosamax was one of the few blockbusters in Merck’s offering where proven product superiority guaranteed high margins—if the company could just sell more of it. Bruzual needed to seize the moment before the window of opportunity was closed by generic alternatives.

It is relevant to grasp that Merck is a highly systematized, results-oriented place. The firm has metrics for everything. One is a “product evolution index.” A score of 100 means a product is growing as fast as the market. Fosamax Mexico was scoring in the seventies—in other words, losing ground to competition. Of forty-one districts (each district consisting of a half dozen reps and one district manager), only six were meeting quota and maintaining market share. From a PD perspective, these positive exceptions were observable fact. With similar marketing budgets, sales coverage, and demographics, they were outperforming the pack. This discrepancy was the hook that mobilized the district managers (DMs) to test the PD concept.

“PD certainly unleashed a novel approach,” confirms Bruzual. “For the first time, I was looking for answers from the grass roots instead of giving them an answer. But it was slow and painful. Only half of the district managers came back the second day, and most of those seemed reluctant to take the reins. I reflected on why. We’re products of classic conditioning. Over two decades here, truth is most solutions come from the top. We had to approach things differently if we expected to discover different answers. But I was upending the culture of the place.”

Inertia and the Prevailing Psychological Contract

Time for the pause button. Bruzual’s narrative calls attention to the district managers’ hesitation to buy into a process that might give them greater voice and control of their destiny. Why? Companies foster an implicit employment contract in which employee contributions and deference to authority are traded in exchange for compensation, a career path, and some measure of job security. In effect, we trade personal discretion, freedom, and latitude for independent action as the price of membership in a company. As witnessed by the reluctance of DMs to take initiative, this “contract” is deeply embedded. And it takes considerable effort to alter it—not just by executives but by employees as well. That’s why an authentic bottom-up process (such as the positive deviance approach) is not an easy sell. Employees are conditioned to go with the flow and follow directions. Accordingly, they need to witness a great deal of management conviction (and behavior consistent with words) to overcome disbelief and fear of retribution for breaking the old contract for a yet-to-be-proven new one.

Bottom-up change is far easier if the authority structure allows leadership to be exercised from the ranks below. One witnesses this at high-performing retailers such as Nordstrom and in the transplants of Japanese automotive manufacturers such as Honda in Marysville, Ohio, or Toyota’s Nummi facility in Fremont, California. But it is a rare condition in most Western firms and a blatantly heretical one at Merck.

There is also political risk for leaders such as Bruzual who elect to experiment in the “participative” domain. Countercultural happenings are hard to hide. They generate disequilibrium that reverberates far from the epicenter. Word gets around: “Andres is doing something weird in Mexico!” It is essential to protect such guerrilla leaders who pick up the baton. Those in authority—in this case, Grey Warner (Bruzual’s boss) and Bruzual himself—need to provide air cover, insulate the initiative from the skeptics, and pass the no-bullshit test from dubious subordinates that their commitment is authentic. Management must resist the impulse to take charge. They must legitimize the challenge as important and be convincing that they can’t solve it themselves. This, by the way, is often experienced by managers as a tacit confession of inadequacy (given implicit expectations that “you’re the boss and get paid a lot for having the answers”). A bottom-up process is an unnatural act within a traditional hierarchy. It disturbs equilibrium. And payoff is far from guaranteed.

Tentative Beginnings

Following the workshop, a core team of five district managers met to discuss next steps. One remembers: “We sat there looking at each other. What had we gotten ourselves into? Yes, we had an idea of how PD worked, but would it work here? Skepticism arose from Merck’s long history of management gimmicks. Was this the flavor of the month?”

Belief systems foreclose the possibility of change. A defining encounter with this obstacle arose in Mali when tackling the high incidence of malnutrition and illness among children (see “The Sorcerer?”).

“We rolled up our sleeves,” recalls one team member. “Timing-wise, we were at a critical juncture. Mexico had a chance to redeem itself. PD appealed because it focuses on our best performers. Strange idea: celebrate them rather than condemning those who are not doing as well. The magic of PD is that it isn’t best practice (where superiors bludgeon everyone into doing something based on somebody else’s success somewhere else). PD appealed because we could figure it out for ourselves.”

“Putting us in charge forestalled our usual passive resistance,” states Hector Ruiz, another member of the group. “When we’re subjected to top-down edicts, we routinely ask for relief, negotiating with Bruzual to offload other work. But in this case we owned the initiative, and had no leverage to ask for offsets in exchange for the extra workload. As it was up to us to decide who would do what, we absorbed the extra workload into our budgets and calendars.”

A plan and a timeline were established. Then the timetable began to slip. Skepticism blossomed at headquarters like algae in a stagnant fishpond: “The district managers aren’t interested” was the corridor conversation. But Bruzual held steady. “Be patient,” he said, “they just have conflicting priorities.” By May, one team had conducted its focus groups. “This lit a fire under the others,” states district manager Takeshi Nakauma. “When we wrapped up phase one, we had reached three-quarters of the 224 sales reps in the country. Data provided a solid baseline of common practices. We began to hear recurring mentions of reps who might be potential PDs.”

A month later, two hundred sales reps from around the country gathered in Mexico City. A mixture of anticipation and skepticism filled the repurposed hotel ballroom. Following a very brief description of the concept of positive deviance and its application in Vietnam, tables of coded data for Fosamax sales were distributed with sales rep names and district numbers blanked out. “Data showed the good, the bad, and the ugly,” recalls one of the DMs present. “No one knew which rep was which. It soon became evident to the reps that the district managers had no predetermined answers. In self-organized groups of eight, the sales reps were asked to identify the high-performing districts and contemplate practices that correlated with better results. Salient findings were written on Post-its and organized into clusters.”

Merck was engaged in a process that tapped the ingenuity and practical field knowledge of its people (i.e., the individual nodes of intelligence). This kind of engagement takes hold when a quest is sufficiently aspirational or urgent that people come to believe their participation is critical to success. Once engagement kicks in, the social system (heretofore frozen into patterns appropriate for business as usual) can, through the PD process, harness energy and challenge orthodoxies. In effect, the group looks in the mirror, decides what makes sense, and determines what to do about it.

Meaningful change must bridge between intellectual awareness “that change is needed” on one hand and the new behavior that makes change real on the other. The second part, as previously noted, requires changes in the social context. Without realizing it, Bruzual was reconfiguring the social system by letting the DMs take the lead. They, in turn, gave the work over to the sales reps. It became apparent: “Management has no answers up their sleeve. It’s up to us!” The reps began to believe that their help was really necessary to interpret the data.

With the sales reps now in the loop, it was time for a second round of field conversations. This time the district managers targeted potential PDs. Pairs of district managers traveled to each city. Their mission was to zoom in on unusual practices.

Takeshi Nakauma from the core group recalls: “Interviewing the PD reps was like following a trail of breadcrumbs. Whereas the average salesman was adhering to standards (i.e., management’s sales productivity checklist), exceptional people were throwing the rulebook away. Amazing. They were improvising to stay in synch with the doctors and deepening those relationships.”

A back story is warranted here. Merck’s sales management embraced a set of orthodoxies venerated as the “Ten Commandments” in running its geographically dispersed sales force. Two, in particular, were tracked obsessively: (1) sales volume by drug category (as the indicator of sales rep performance), and (2) the “Rule of Seven” (i.e., reps must make seven sales calls per day to meet the company’s productivity standards). District managers were the enforcers of this code. (That’s why they—not sales reps—were made the centerpiece of the PD process and did the interviews. Andres Bruzual was convinced that the district managers, more than the reps themselves, needed to change mind-set and behavior.) Interviews were confirming that scattered across Mexico were a few standout sales reps with novel shortcuts to productivity, imaginative ways of positioning products, and personal touches that made them a trusted partner with the doctors.

People in organizations make thousands of decisions, and these sum to define an organization’s trajectory. The arc of these internalized “truths” is invisible but inexorable: How “big” does an opportunity have to be to be “interesting”? How are budgets negotiated? How is analysis translated into financial projections and how do projections in turn preconfigure the aperture of the possible? What are the consequences for missing financial targets? For making mistakes? These interpretations reinforce the conventional wisdom about “the way things work around here.”³ When adaptive change becomes necessary, subtle undercurrents embedded in the social system can inflict death by a thousand cuts and undermine needed change. Companies—particularly successful and well-established ones—are often ensnared in a belief system that worked brilliantly in the past but becomes incongruent with new realities.

Core team member Fa Leo adds: “Initially, the reps felt that our interviews were aimed at ‘getting dirt on their boss.’ Participation in Mexico City began to change that. They could see our focus was on field outcomes, not supervisory behavior. They began to speak up. ‘Hey, this is a chance to tell it like it is.’ We began to hear about behavior at variance to the rules. They told us what they were really doing versus what their managers thought they were doing. We realized how little the district managers knew. Example: reps are supposed to use a handheld computer to track sales and profile customers—a Merck standard. But they said things like, ‘We don’t use them. It’s just a reporting device to headquarters.’ Or, ‘Time spent entering data into handhelds takes time from working with doctors.’ Scales were falling from our eyes.”

Somewhere in this second round, many district managers had an epiphany. Some reps were exhibiting extraordinary ingenuity and achieving extraordinary results. That’s when many became believers. They were discovering useful tricks of the trade that had been invisible. A district manager expressed his surprise: “Very interesting stuff. At first I was shocked. ‘Handhelds not important!’ That’s heresy! ‘Marketing material is irrelevant’—wow! But as it sank in, I was getting insight into the counterproductive side effects of our productivity measures.”

Having interviewed fifteen possible PDs, district managers narrowed the list to twelve whose approaches seemed most replicable. But instead of declaring these “best practices,” they reconvened all sales reps for a second time. The venue was the ballroom of the Royal Hotel, Mexico City. More than two hundred attended, arranging themselves in small groups. There were no tables. Each group had a flip chart. Findings of common and uncommon behaviors were debated. Throughout the day the DMs emphasized: “This stuff belongs to you. The practices belong to you.”

The biggest surprise to many—both sales reps and district managers alike—was that the “Rule of Seven” was getting in the way. PDs making three calls per day were routinely outperforming those who made seven. The rule was actually hurting Fosamax results.

Some subgroups had a PD in their circle, but neither they nor the PD knew it. When discussion turned to one of the uncommon practices, the unknowing PD might say, “Well, I do that. I give presents to the doctor. Isn’t that normal?” The rest would reply, “No, not at all.” Another was doing a great job with client education. Instead of just sending out obligatory invitations to doctors for Merck-sponsored lectures, this individual called doctors before the agenda was set, solicited ideas for topics, enrolled them in the event’s design, and invited them to participate on a panel. A big “wow” for many. The key to success in the relationship with the doctor is ingenious “little things”—sending birthday cards, attending a doctor’s son’s graduation, bringing a favorite food from another territory. Reps were learning from each other. There was a lot of give and take. District managers observed from the sidelines and did not intrude on group discussions.

The session was designed to conclude with time for all reps to write up their takeaways for their return to the field on Monday. It was pouring rain, and given rush-hour traffic in Mexico City (ugly on the best of days), the sponsors decided to adjourn the session so people could go home at 3 p.m. Yet most stayed until 5:30 and completed their personal action plans. The meeting ended with two hundred individuals committing to an uncommon behavior that was appropriate to their situation.

Results? Extraordinary! By the end of the year in which the project was launched, all districts met quota. One year later, the same population of sales reps was 30 percent above quota, an exceptional achievement. District managers witnessed veterans adopting the PD behaviors voluntarily, not out of obligation or compliance. Many formerly “exceptional” behaviors have since become commonplace. (One of the attributes of the PD approach is that the exceptional behaviors stick.) The “Rule of Seven” is now only a guideline. A number of the PDs have been promoted, a by-product of their results and visibility through the process. Of the twenty-five district managers who participated extensively, most report a change in their outlook and management philosophy.

Early Wins, Squandered Gains

Given the impressive results, readers may be startled to learn that Merck’s successful application of PD remained confined to the Fosamax sales force in Mexico. Why? As Bruzual notes, sponsorship requires executives who are truly willing to empower those in the ranks below them and hold steady during the time it takes to discover things. Mustering commitment isn’t just a matter of waking up one morning and deciding to give positive deviance a try. It requires political will to engage in adaptive work. “For one brief moment,” states Fa Leo, “Andres had all the ducks in line: Grey Warner was an advocate, as was his team at corporate. All bought in. But counterparts elsewhere in other businesses responsible for other pharmaceutical lines or geographies didn’t buy the idea.”

“We cannot say the big uptick in Fosamax sales that followed was 100 percent the result of PD,” observes Bruzual. “We had simultaneously released Fosamax Plus—an improved formulation. This muddies the metrics a little. Still, the PDs continued to do better than the pack, even with Fosamax Plus. In any case, our pre-PD baseline was 39 percent market share. Post PD, we had 45 percent market share—highest in the world.” Today, Mexico’s Fosamax share still ranks near the top.

Political Capital and Organizational Drift

The gains were significant—but so were the costs. “This kind of process needs a lot of noise, news, attention, buzz, and enthusiasm,” reflects Bruzual. “If the senior sponsors aren’t deeply committed, you can’t sustain something like this. Most organizations have a big, powerful constituency for ‘what is’ but almost no constituency for ‘what could be.’

“It’s very energy intensive,” he continues. “It took some doing to convince my immediate boss to get a hotel for the district managers and to pay travel costs for reps to attend the session. We had to justify hijacking teams from their day jobs on the street. Mexico is a huge country. Travel time and travel costs add up. Consider the scale—forty-one district managers and over two hundred sales reps. All this took eight months. We had to piggyback off routine cycle meetings to reach people and disseminate what we were learning. If my general manager hadn’t bought in, it wouldn’t have happened.

“When you try to do something like this,” Bruzual concludes, “you draw on precious resources: money—and especially time. Taking sales reps out of the field for a day means they’re not calling on doctors. People get nervous. Merck (and the pharma industry as a whole) is a very old-style, traditional business. We’ve been doing things one way for one hundred fifty years. Until recently, Merck has been very successful. You become arrogant. There is little latitude for self-examination—and little openness to change.”

There is a term for the undertow that draws organizations toward traditional patterns. It is called “organizational drift,” and it exercises influence subliminally. Existing political arrangements usually ensure that novel initiatives rarely get a fair day in court. The weapon of choice for protecting the status quo is the “precaucus” veto (a stealth device for sidelining bold new ideas), often framed in economic terms. As the value of money spent is more quantifiable than the value of opportunities lost (e.g., the opportunity cost of taking a salesman away from the field for a day is more measurable, at least in the short term, than any potential gain), the burden of proof lies with the initiators of change. Unsurprisingly, the status quo prevails. Remember that those on top have made it in the current system, and they see little personal value in changing what they know and can succeed in—especially when challenged by ideas like positive deviance that are detached from the orthodoxies that undergird their historical success.

People in authority do not generally see themselves as simply defending the status quo or avoiding the risks of change. At Merck they saw themselves as protecting the operating system and cultural norms upon which a large, global enterprise was built. Destabilizing this essential discipline is not taken lightly. The challenge, of course, in companies as in nature, is to foster sufficient adaptation to the changing environment to remain competitive while at the same time conserving those qualities that remain relevant.

Perspective on differing contexts for the exercise and alignment of authority is warranted here. In Vietnam, Egypt, and the Pittsburgh VA hospital, more attention was given to achieving alignment—or at minimum acquiescence—from those in charge. And in most of the examples in this book, authority was diffused. Village-level buy-in was sufficient to get a PD process up and running. Ministerial authority in Hanoi, Cairo, Kampala, or Islamabad was unlikely to stand in the way of success once efficacy was proven. In hospitals, alarming MRSA transmission rates were actually killing patients—a burning platform that few could ignore. At Merck, by way of contrast, Grey Warner did not attempt to enroll the broader authority structure above him or even to syndicate this initiative with functional and geographic peers at his level. Within his own span of control, only one of his subordinates, Bruzual, volunteered to try PD. Others opted out. Further, no effort was made to evangelize any of these stakeholders as the initiative gained momentum. Add to the mix Merck’s command-and-control tradition, far more pervasive and rigorous than generally encountered in villages, countries, and even hospitals, and the lack of uptake becomes more understandable.

Barriers to the Diffusion of Innovation

Though Bruzual was instrumental in the success achieved in Mexico, it is noteworthy that when rotated into a new assignment in Chile, he didn’t repeat his initiative there. Why? “Support from my Chilean country manager wasn’t in place,” he explains. “And my successor in Mexico was, let’s say, ‘a different kind of manager.’ He let the approach languish, notwithstanding the Fosamax beachhead. My district managers in Mexico have been hesitant to apply it elsewhere as well.”

When pressed, Bruzual elaborated on why PD hadn’t diffused beyond Fosamax and Mexico. “‘Not invented here’ is always latent in an organization’s immune defense system,” he states. “Even after we got a great write-up in the London Times (which we disseminated), people said: ‘Yes, but that’s Mexico. We’re not Mexico; we’re different.’ The general manager of Argentina read the report on our results. ‘Is this really true?’ he asked, betraying his skepticism. We sent him audited documentation of results. No response. Nothing crystallized in other countries. A few tried but failed for lack of sponsorship.”

David Gasser, a senior staff member of Bruzual’s executive team, elaborates on the dilutive effect of incessant job rotations. “You crush the momentum of a process like this when you change senior leadership and transfer the champions. Success was a brilliant but fragile flower in the Merck jungle. It needed prolonged nurturance. Job rotations break up the symbiotic networks you’ve created. The fragile ecology that fostered change never reaches the critical mass necessary to become self-perpetuating.” Gasser makes an important point here. Prior to PD, successes had occurred in intact communities (including hospitals) where the stakeholder group was relatively stable. The argument for job rotation in big companies, of course, is that it cross-trains talent and helps break down functional silos by introducing different perspectives. All true. But it is absolutely toxic to efforts aimed at building a critical mass of Jesuits dedicated to the new gospel. Constant churn ensures that the advocacy of individual converts will be trumped by institutional memory.

Lessons Learned

“What did I learn personally?” Bruzual asks rhetorically. “In this industry you are told what to do. By the time you get to my level, it is an unnatural act to free up people to do what they think they need to do. PD has had an impact on my leadership approach. I have gone from being directive to being more empowering. It was a personal learning experience.”

Bruzual’s subordinates noted a change in his style. (They initially worried Bruzual was ill, as he was so atypically nondirective.) “Today he listens more and asks more questions,” states Raimondo. “He has learned how to mobilize the troops. Vestiges of the experience are embedded in his behavior. But he is a tiny minority in this huge enterprise; indeed he was unique among his peers in Mexico.”

The man who started the whole thing, Grey Warner, reflects on results both bitter and sweet. “I’ve probably done more with many of these tools than anyone else. But overall, not a lot has happened since. I’m certain that the people involved speak to each other in a different way (peer to peer, peer to boss). They approach work in a different way. At the finale in Mexico City we went around the room asking people for their impressions. One rep, not a PD himself, said, ‘I’ve seen how small things can make a big difference.’ PD has manifested itself in the way individuals are more involved and in the way they go about working.”

Reflections

The sustained turnaround of Fosamax sales at Merck confirms that the PD process can work in companies. But the aftermath raises intriguing questions. Why the lack of diffusion elsewhere? Is there an inverse correlation between tightly managed companies like Merck and PD’s prospects? Is there something about corporate management that confines successes to isolated beachheads?

Merck epitomizes the classic standard model: socially engineered, top-down, authority driven. As mentioned briefly in chapter 1, this worldview holds that uncertainty and risk can be mitigated by meticulous planning, direction, and control. This framework promises greater predictability and reduces executive performance anxiety. Authors, consultants, and management gurus have invested decades railing against the limitations of this approach with little headway. Peel the onion, and we discover a root system entwined around orthodoxies that hold it fast in our repertoire. Not least among these, as noted earlier in this chapter, is the psychological contract that undergirds employee service to an enterprise.

The standard model has a lot going for it. It seems to be the most efficient way of getting from here to there. For technical problems involving “known knowns” (e.g., launching a marketing campaign) or “known unknowns” (e.g., the protocols for human trials for a promising new compound), it is exactly the right approach to take.

Alfred North Whitehead once said: “Civilization advances by extending the number of operations which we can perform without thinking about them.”⁴ One of the many virtues of the standard model is just this: the more that can be accomplished through productive routines, the less potential waste from improvisation and reinventing the wheel.

Bruzual’s comments underscore that Merck, like most corporations, has a well-honed set of such protocols that meet the frugality test. When a choice arises between relying on a proven top-down approach (albeit even in situations where it may be stretched to the limits of applicability) versus an iffy, amorphous, hard-to-predict, bottom-up process, the familiar seems the better bet. It takes less time. Why abandon the proven in favor of the problematic?

On the whole, companies are remarkably efficient organisms. (And the market makes quick work of underperformers.) Sophisticated tools used in a modern enterprise give management reasonable basis for confidence that decisions from above translate into proximate action. Realtime information, financial controls, and operating tools (like Six Sigma, business process engineering, pay for performance, and other incentives) are, all in all, efficient and effective.

Like many companies, Merck operates under assumptions of scarcity. Almost universally, scarce resources are allocated via central authority. It is an effective way to make trade-offs, resolve conflicts over competing priorities, and lower transaction costs. Recall that Bruzual was sensitive to his use of investment dollars, head count, and sales force time. Scarcity is built into Merck’s genetic code. Start with this premise and many of the inhibitions against experimenting with an approach like PD become self evident. Scarcity, not abundance, is as foundational to management hierarchy as royal bloodlines have been to monarchies since Babylon. It preserves the order of things.

Mitigating Risk and Avoiding Failure

Corporations are averse to failure. And to their credit, its occurrence is episodic—an outlier condition. When failure occurs, causes are analyzed and preventive measures implemented. (See “The Spectrum of Process Standardization.”) Relying on contemporary management tools and processes, employees can be depended upon to execute a corrective plan. Executives have reasonable confidence that if they declare “rudder hard right” the ship will turn to starboard. One triumph of modern management is this legacy of tightly coupled operating systems that usually work.

Consider this: maybe failure is less the exception in noncorporate settings. (Some might argue it is more the norm.) Corporations work on the premise that they can weed out infrequent failure via rational analysis and the disciplined application of management science. More loosely coupled organizations (e.g., villages, NGOs, universities, etc.) experience these corporate management tools as problematic. Positive deviance (focusing on those who succeed against all odds) is appealing when one is not as certain the organization can march directly from A to B. The PD process starts with what is succeeding and works backward.

Limitations of the Standard Model

Einstein once said: “Everything should be made as simple as possible—but not simpler.” And there’s the rub. Yes, there are many good reasons to rely on the standard model most of the time. But, true to Einstein’s cautionary quote, simple remedies can at times be simplistic. There are circumstances where the standard model doesn’t work. Trouble is, it is hard to spot these cases ex ante.

A bottom-up process interjects diversity. This enlarges what we call “the solution space.” Up to a point more diversity means more adaptive capacity. Small and obscure components—easily overlooked—can be vital (we witnessed this at Genentech and Merck). Attention to detail matters, and attention to small details matters a lot.⁵ In nature, such detail is manifested in gene pools and tiny mutations. Organizational mutations occur among those closest to the action via small adaptations in the face of adversity or opportunity. Those closer to the bottom of the organizational pyramid (for example, in companies where delivery of services or the manufacture of products actually takes place) participate in “experiments” every day. But the standard model isn’t especially well geared to listening, let alone learning through this bottom-up channel. As described earlier, Merck Mexico discovered, as part of a positive deviance inquiry, that a few aberrant sales reps had experimented, learned, and established innovative pathways to success. They had made microadaptations to the microclimate of Mexico—notwithstanding Merck’s systemic guidelines for its global sales force. The missed opportunity, of course, was that Merck needed myriad microadaptations of this kind to optimize in each microclimate and compete. But high-decibel directives of the standard model drowned out quiet discoveries from below. Unsurprisingly, these remained invisible to management until the PD process brought them to light.

The biggest limitation of the standard model is that it takes process standardization too far. It loses its potency when local knowledge and aesthetic judgment matter. Adaptive work is judgment intensive and craft dependent. Variability between one situation and the next matters. One-size-fits-all protocols fail. Grey Warner’s ambition to reinvent Merck’s relationships with the doctors grasped the importance of all this in his quest to differentiate Merck from competition.

The case for tapping the distributed intelligence of the community that needs to change is not advocated as an end in itself or inspired by an ideology of “empowerment,” “employee participation,” or “workplace democracy.” The success in Mexico demonstrates that the rank and file need to be engaged because they often discover ingenious ways to deliver better results. It is also the best way not only of capturing what they know but of enabling them to change what they do.

Can PD Work in Companies?

These reflections highlight many factors in for-profit enterprises that make adaptation of the positive deviance approach difficult. But it is not impossible. The next section describes a highly successful adaptation of PD at Goldman Sachs. It offers a provocative contrast to Merck because both firms were: 1) trying to understand what made some salesmen more productive, and 2) spread successes cultivated in one region to other geographically dispersed units. Sales forces are ideal for testing assumptions about performance and change. Salespeople typically exercise more latitude for independent action day to day than their counterparts on assembly lines or in office cubicles. Owing to the propensity toward greater variation in performance, commission-based incentives are employed to align corporate and individual goals. All this extra leeway creates fertile ground for positive deviants. We observed this at Merck. Goldman Sachs provides a more encouraging counterpoint, engaging its wealth advisors in identifying the PDs in their midst, learning what they were up to, and giving individuals the freedom to opt in or opt out of applying lessons learned. Few declined the invitation. Within two years, these ideas had spread across all eleven regions in the United States.

Leveraging the Positive Deviants at Goldman Sachs

Goldman Sachs’s Private Wealth Management unit had experienced a string of controversial top-down change initiatives.⁶ On one hand, the field force of three hundred investment advisors felt railroaded by New York’s radical policy shifts and top-down edicts. On the other, headquarters felt thwarted in achieving the pace of change needed to stay in step with the marketplace.

Investment professionals (IPs) in the field historically operated independently or as two- or three-person teams. Each unit evolved highly idiosyncratic approaches to the work of persuading high-net-worth individuals to entrust the teams with the management of their money. Success in this work depended on performance, of course, but also on building deep, trusting relationships with wealthy clients that often lasted for decades. These clients invited their IPs to weddings, bar mitzvahs, and graduation ceremonies that extended relationships to their heirs. For this reason, IPs remained notoriously independent of the company itself. When they moved to another firm, their clients would typically follow. To an uncommon degree, power favored the field force—not the traditional corporate hierarchy.

On the other side of this classic field-versus-headquarters schism was Goldman Sachs’s top management, deeply concerned that the industry was undergoing a transformation of disruptive proportions. Investment firms were under pressure to deliver greater transparency and compliance oversight while simultaneously reducing their brokerage fees. How could Goldman Sachs retain its existing wealthy clients, improve its profitability, and grow its assets in a turbulent and increasingly competitive environment? Management’s solution? Transform the IPs’ approach from a process that relied heavily on brokerage income to one focused on fee-based advice. But the IPs resisted the policy shift—and this was maddening to their seniors. The business case for change was compelling—it was all there in PowerPoint. Why didn’t the field force’s knowledge of the SWOT analysis translate into an appetite to actually change what they were doing? Management tweaked and rejiggered their pitch—a simple application of knowledge management. The assumption was: if we just get the right communication formula, we can inoculate the many disbelievers.

David Dechman, cohead of the U.S. Private Wealth Management (PWM) unit, found himself at the epicenter of the impasse. By late 2001, the PWM unit was feeling the full impact of the gridlock of the stakeholders. Dechman’s chosen path was to relinquish his role as an authority figure and to let go of top management’s deep attachment to “better knowledge management” or “more directives” as the solution. He exercised leadership by asking the individual IPs one arresting question: “Are some teams, with similar territories and prospects, able to thrive in this difficult climate?” In short, Dechman engaged the reps in tapping their own wisdom.

David Dechman had never heard of the positive deviance process. But the stubbornness of the problem he faced and the independence of those he sought to mobilize drew him intuitively to design a process that was identical in every important respect.

A six-person council of influential IPs spearheaded a “sales force effectiveness” inquiry. The council’s task was to catalog common practices and subsequently identify exceptionally successful approaches. Its members assured the rank and file that any findings would be subjected to the acid test of relevance and scalability (e.g., what was working for the best team in Boston would have to be transferable to Dallas).

Phase one invested two months in focus groups involving IPs across the system. Interviews of peers established a baseline of common practices and occasionally surfaced an unusual angle that seemed to be working. Phase two focused on discovery, talking in depth to potential PDs. Phase three coalesced these findings. Teams (composed in part of individuals already using the practices as well as those who had become advocates along the way) carried the momentum forward. Five practices seemed especially relevant. The teams were then charged with coming up with an action learning design such that every team in the country could be exposed to the ideas, debate them, and implement them on a voluntary basis.

When it was time to disseminate, teams that had identified the PD practices became the pointy end of the spear. They visited each office and described why and how their particular practice worked. There was one person on each team from each office, so one of the presenting IPs could double as a local resource on the topic. When local teams had questions later, they could turn to their resource person. The approach generated an amazing buzz—primarily because each team could witness what others “just like me” were doing that was yielding better results.

Phase three of the process involved building a system to measure progress toward goals and to track trends over time. Dechman ranked each of the eleven regional offices on their incorporation of the five practices, and he publicized results. The process relied exclusively on transparency and peer review. No sanctions for nonadoption were imposed. People automatically felt good about being on top or bad about being on the bottom. This sustained attention when backsliding might have otherwise set in.

Goldman Sachs management had tried and failed to communicate the threatening market changes using analysis-based, classroom-like briefings. Sometimes this technique works. But explicit knowledge, conventionally delivered like pizza (neat boxes with toppings of concepts, theories, best practices, and war stories), is consumed by the brain but not metabolized into action. The learning we call intuition, know-how, and common sense gets into the bloodstream through osmosis. It is shaped by social context. Management needed to tackle this to progress.

During the course of this endeavor, old rivalries between formerly competing teams of investment professionals subsided. For the first time in memory, a sense of “we win together” emerged as the new ethic. Skepticism gave way to conviction as the wealth advisors overcame their own exceptionalism. The PD approach, implemented over eighteen months, profoundly altered behavior, practice, and performance. The business unit got a jump on the competition. Three years later, the Private Wealth Management unit had doubled in size and was singled out by the firm and its shareholders as the business it wanted to grow most aggressively. Average productivity per wealth advisor nearly doubled, team size has increased, and the fee-based business model has achieved near-universal acceptance.

Results

The case study at Merck and the contrasting example at Goldman Sachs describe parallel challenges facing a sales force and their efforts to amplify positive deviance. The outcomes, however, were highly divergent. Neither was a minor initiative. Merck utilized PD to address below-quota sales of its blockbuster Fosamax across its sales force in Mexico. Goldman Sachs employed a methodology that mirrors PD to increase productivity of its three hundred wealth advisors distributed across the United States. Both initiatives were highly successful, documented by dramatic empirical increases in performance that have proven sustainable over time. Nevertheless, Merck has yet to utilize the process again. At Goldman Sachs, by way of contrast, the approach has been internalized as a way of running the business.

Does the success of the positive deviance approach at Goldman Sachs offer hope that the standard model is losing its grip on the corporate mindset? Probably not. Why? Goldman’s Private Wealth Management division operates more like a loosely coupled community than a traditional company. Its wealth advisors are quasi-independent franchises, each with a set of loyal clients. (Perceived independence is important to providing impartial and objective financial advice to high-net-worth individuals.) Accordingly, upper management can’t successfully “dictate” to these entrepreneurs. More analogous to the village stakeholders of Vietnam and Egypt, management recognizes that the best course is to shepherd intelligence and entrepreneurial energy in the ranks. Goldman Sachs adopted a strategy for herding foxes. Its unique adaptation of positive deviance fit the bill.