Typically change agents look outside the organization to see so-called "best practices." Maybe we should be following the example of Save the Children that tackled the malnutrition problem in Vietnam by looking inside to find healthy children.

By Linda Rising

There are many things I like about searching for positive deviants. First of all, the Save the Children story is a great one. Second, it makes me stop and think whenever I want to try to "improve" an organization. In the old days, I would, usually at the behest of a high-level manager, research current "best practices" and then spearhead a task force to "make it happen." After all, if it works for <name your favorite big organization here>, it should work for us, right?

By finding models for change inside the organization, you are getting rid of the traditional organizational development model where you and some outside consultants go in and focus on a community’s deficits. You are the expert and the people involved in the actual work don’t know anything. The positive deviant model says, "We’re not the expert, we don’t know anything. Within your community, you have the answers, so let’s understand together what they are."

Let’s start with the story—it’s not a software tale at all but one that should concern us. It’s about malnourished children in Vietnam and a guy named Jerry Sternin and his wife, Monique. They were part of the Save the Children organization responding to a request from the Vietnamese government. And, here’s something they shared with software projects, they were given an impossible task—nearly half the children in Vietnam were malnourished—and a severe deadline—they had six months to produce results.

The problem of malnutrition is pretty well understood. It’s the result of ignorance, poverty, poor sanitation, faulty food-distribution, and cultural influences. These are tough problems to solve on any level, but impossible to treat in six months. The typical approach is to bring in some engineers and nutritionists with clipboards. They institute a few radical changes, things appear superficially better, and then the experts leave. Since there has been no real change, the experts take their expertise and their good ideas with them and it isn’t long before things are just like they were—maybe worse!

In contrast, here’s a step-by-step outline of the method of positive deviants:

1. Never assume you have the answer.

The Sternins didn’t know a great deal about Vietnam but they believed that the only way to solve the problem was to discover the solution within the Vietnamese villages. They were ready to listen, not to talk. They were ready to learn, not to prescribe. They were ready to find some children who were not malnourished and help others learn from their success.

2. It’s not a cocktail party.

In many organizational change texts, the idea of mixing across departments is encouraged. When identifying positive deviants, the opposite is used. The solutions you find will most likely be community-specific. You’re not trying to brainstorm creative ideas—you’re trying to identify what works. If people can’t identify with the deviants, they won’t adopt the solution, even if it has a proven track record. Stay within the community while you look for solutions.

3. Help them do it themselves.

It’s hard for us to follow this one! I lead retrospectives for companies and the challenge is to allow the teams to develop their own solutions. Since I’m being paid to lead the exercise, many managers want me to simply apply the proper band-aid to each difficulty as it arises. It would save time! But would it? It’s been my experience that when we are handed a solution, it won’t work. When we "own" that solution, because we discover it ourselves, we are much more likely to see it through. That’s exactly the message here. There is research to support this. Allow the community to examine the behavior of the positive deviants. Help them ask questions. Help them understand the deviant behavior. Support their observations with research but let them draw their own conclusions. This takes time but will produce the support you need. Respect and appreciation for local intelligence is required.

The Sternins started with 4 villages. They asked women if there were any children in poor families who were well nourished. The women knew who those children were. The next step was to discover what the families of those well-nourished children were doing that was different from the rest.

4. Learn the culture.

You can’t identify unusual behavior unless you understand typical behavior for the community. Find out what most people do. Ask questions to bring out assumed conventions.

Most Vietnamese communities regarded certain foods as low-class, even though the foods were nutritious. Mothers in the villages did not actively encourage eating. They believed that it wasn’t healthy to feed children with diarrhea.

5. Analyze deviant behavior.

As you learn how the community behaves, the behavior of deviants will surface. It will also become clear that the deviants have found a better way. The results will prove it. The people who need to change can see how to do it if you help them. The people within the community won’t feel that an outside solution has been imposed on them. They will discover the new way themselves.

In the Vietnamese villages the deviant mothers used alternative sources of food, some that was considered low-class. They fed their children even while the children had diarrhea, feeding children more frequently; and making sure that the children actually ate, rather than hoping that the children would take it upon themselves to eat.

6. Allow the community to adopt the deviant behavior on their own terms.

Don’t simply tell the community, "OK, guys, here it is. This is the answer to your problem. Make it happen!" You don’t want to solve a problem; you want to change behavior. Enable a learning environment where people can discover the deviant behavior.

Sternin makes a point of emphasizing the distinction: Don't teach new knowledge—encourage new behavior. Let the people who have discovered the deviations spread the word. Don't require adherence to the new practices, but do offer incentives for it.

In Vietnam, a health volunteer would invite 8 to 10 mothers into her home for medicinal-food training. As a price of entry, the mothers were required to bring a contribution of shrimp, crabs, and sweet-potato greens. They would use the ingredients to cook a meal for the group. After a couple of weeks, most of the group would continue to gather shrimp and greens, and their children would recover. Mothers could re-enroll and go through the two-week process again, over and over, until their children recovered and the behavior became habit.

7. Measure success and share the results.

When you share success, others will be curious. Learn from working with diverse groups and see how they have changed. This retrospective process is always useful and is a "best practice" for software development. To avoid making the same mistakes over and over and to learn from what worked well, take time for reflection to gather lessons learned and share those across the organization.
http://www.ddci.com/news_vol3num5.shtml#Postmortems

Sternin reports, "We saw malnutrition drop 65% to 85% throughout the villages in a two-year period. The Harvard School of Public Health came to the four original villages and did an independent study. They found that children who hadn't even been born when we left the villages were at the same enhanced nutritional levels as the ones who benefited from the program when we were there. That means that the behavior sticks."

8. Repeat.

Once the group has learned from the positive deviants, go back and repeat the process. It’s likely that other positive deviants will have created new solutions and the group can continue to learn. This is the goal of continuous improvement that has long been held up for software development without much real guidance as to how it can be achieved.

Sternin took his positive-deviance program to a total of 14 Vietnamese villages after succeeding in the initial 4. As the program grew, it uncovered new solutions in new localities: sesame seeds, peanuts, snails. The answers were never quite the same, but the process remained the same: Discover original local answers to the problem, and then give everyone access to the secrets.

The most powerful application of this final step was Sternin’s idea of a "living university." The 14 villages became a social lab. People could visit these villages and then return to their own villages and implement ideas they had learned. They could return to learn more. They could build on this success, as their home villages became positive deviants for others. The program ultimately reached 2.2 million Vietnamese in 265 villages.

Over the past decade, positive deviance has been applied to the problem of malnutrition in more than 20 countries through Save the Children. Other non-governmental organizations have applied it in many countries as well, including Bangladesh, Bhutan, Bolivia, Cambodia, Egypt, Ethiopia, Haiti, Myanmar, Nepal, and Sri Lanka.

News of Sternin's work in Vietnam spread rapidly among a variety of non-governmental organizations. Positive deviance is now being applied around the world to change behavior in a variety of other social and organizational situations, such as the spread of AIDS in the Third World and ethnic conflicts in Africa.

Let’s hear a business story. A successful pharmaceutical company had one unit that far outsold all of the other groups. The company believed that the more sales reps you had and the more calls you made on customers, the more you would sell but the positive deviants had fewer salespeople and made fewer calls. By applying the methods described in this article, they found that the successful reps were spending more time with individual doctors, educating them on the benefits and the uses of the products that they sold.

Steven Miller, Managing Director of Shell Oil Products Business Committee: Top-down strategies don’t win many ball games today. We need a different definition of strategy and a different way to generate it. The top can’t possibly have all the answers. The leaders provide the vision and are the context setters but the actual solutions—how best to meet the challenges—have to be made by the people closest to the action.

Barbara Waugh, Worldwide Change Manager for Hewlett-Packard Laboratories: The 60s way of doing things was to dig into a complicated problem, pick out the worst elements, and go after them. Today, I take the opposite approach. I seek out the positive deviants and support them. Give them resources and visibility.

Here’s a brief comparison of positive deviance with traditional organizational development:

Asset model vs. a deficit model. Look for what’s right instead of what’s wrong. It puts you in a completely different frame of mind.

Based on indigenous wisdom within the system vs. expert opinion from the outside. You’re the experts. We’re not coming in to tell you how to do your job.

Cheap vs. expensive.

Goes with the flow vs. disrupting the status quo.

Finds those who are successful and what they are doing differently vs. finding out what’s wrong and trying to fix it.

Stirs up less resistance because we’re learning from our colleagues vs. outside experts and new ideas from the outside.

Produces change that lasts vs. change that doesn’t last.

The biggest benefit comes at a personal level, as one villager told the Sternins, "Let us tell you about the changes in our lives. We were like seeds locked up in a dark place, and now we have found the light." http://www.savethechildren.com

References

Dorsey, D., "Positive Deviant," Fast Company, December 2000.

Pascale, R.T., M. Millemann, and L. Gioja, Surfing the Edge of Chaos, Crown Business, 2000.

Sternin, J. "The Power of Positive Deviancy," Harvard Business Review, January-February 2000.

Waugh, B. with M. Silk Forrest, The Soul in the Computer, Inner Ocean, 2001.

About the Author
http://www.lindarising.org
risingl@acm.org

Linda Rising has a Ph.D. from Arizona State University in the area of object-based design metrics. Her background includes university teaching as well as work in industry in telecommunications, avionics, and strategic weapons systems. She is the author of numerous articles and has published three books: Design Patterns in Communications, The Pattern Almanac 2000, and A Patterns Handbook. She is currently writing a book with Mary Lynn Manns: Introducing Patterns (or any Innovation) into Organizations, to appear in Fall 2002.