The hidden workplace (pg. 2)

Roads in India, airports in Latin America: There's a global building boom going on, and for Fluor (Charts, Fortune 500), one of the world's largest engineering and construction companies (ranked No. 174 on the Fortune 500), these are among the best of times. Except for one thing: With engineering talent in high demand, the competition to recruit and retain the best is intense. So Fluor, based in Irving, Texas, has to work hard to keep its employees happy, says Karen Vari, Fluor's global leader for diversity and inclusion. "If people feel included and knowledgeable and inspired and motivated," she says, "they are more likely to stay."

A companywide survey in 2002 showed that Fluor's employees wanted more access to senior-level executives. In response, the company started a mentoring program that matched one executive with anywhere from five to nine lower-level employees drawn from different parts of the company. Vari, who led the pilot, kept the "mentoring circles" as unstructured as possible. Participation was voluntary.

The groups met for roughly 90 minutes once a month over the course of a year. Members decided what to discuss at each meeting - from why Fluor was entering a given market to career development. "We're mixing the mentoring and the networking together," says Vari. The program started with 32 employees in three locations. Since then 1,070 employees in 11 spots have participated.

In Fluor's case the mentoring circles are meant to give employees a sense of belonging to a broader organization, one whose different parts fit together a certain way. Designers, for example, learn the perspective of sales executives; marketers see the challenges faced by the finance department. "Each of us was from a different background," says Susana Suárez González, a human resources director in the Houston office who has been involved in four mentoring circles in three countries. "The thing we all had in common was that we really, truly wanted to know more about the company."

Fluor surveys participants at the end of every cycle on whether the goals were met and whether they'd recommend the program to others; so far, 80% have rated almost every category as good or excellent. The company has tracked the circles' effect on retention only in Houston, but the results there have been dramatic: The turnover rate for those participating in the program has been half that of the rest of the office. Perhaps these were the more motivated employees to begin with, but Vari is inclined to give the circles some credit. "Attention equals retention," she concludes.

With 135,000 employees working in 80 countries, Procter & Gamble (Charts, Fortune 500) understands the importance of sharing information. In the past decade the number of people in its R&D operations has stayed about the same but has been dispersed across 25 technical centers (up from 12). Brice Westring, R&D global knowledge-management leader, worried about the strain this put on P&G's information networks. In the mid-1990s R&D had started its own communities of practice (COPs) - clusters of people with similar competencies, such as skin science or packaging. The question was whether those communities remained effective as the offices multiplied.

So in October 2006, P&G embarked on what Cross says is the largest social network analysis ever conducted in the private sector. The company asked 8,000 people who had been members of a COP to name their top ten connections, resulting in an incredibly complex web of 17,000 names. The data are still being digested, but P&G has come to some preliminary conclusions.

First, to no one's surprise, people were most likely to list the names of colleagues who worked in the same location, followed by colleagues elsewhere in the business unit. The COPs came in third, proof to Westring that they still knit people together.

The analysis also spotted some groups isolated on the periphery, such as R&D in China - a critical finding considering the importance of that market. "A lot of the conversation was one-way," says Westring. "They [R&D China] were seeking information but not necessarily being sought after," because of geographic and cultural barriers. Another ill-connected group: new hires. Westring is working on a plan to improve the integration of new R&D employees.

A more targeted use of network analysis took place at Raytheon, the Boston-area defense and technology company. In 2004 Karl Arunski became engineering director at the Colorado-based Rocky Mountain Engineering Center, which helps design satellite systems. He decided to run a network analysis to help him understand how the place worked.

On the organization chart, the 2,000 employees worked in groups sorted by traditional engineering disciplines like systems and software, but that didn't account for technology-focused expertise. Specialists in mission management, for instance, were dispersed among the various groups. So Arunski asked two executives to give him the names of five to ten experts in areas that were falling between the cracks of the formal structure. From those names he formed a group of about 50 people and ran a social network analysis that showed where their knowledge resided. Then Arunski created five "centers of excellence": groups of engineers who communicated via their own e-mail lists and occasional face-to-face meetings.

The analysis revealed trouble spots. Arunski noticed one person on the map who was connected to several engineers but was isolated from the rest of the team - a "Pluto," to use Arunski's term, "with a bunch of little moons around him." This group was especially valuable to Raytheon because its expertise was in systems architecture, a growth area. To put Pluto into play, Arunski asked him to head up his own steering committee. "I said, 'Go grab a few people who understand [systems] architecture, and I'll buy lunch."' At first a dozen engineers would gather to trade strategies; the pool eventually grew to 75 people. Since the group started in 2004, their expertise has helped shape three proposals, says Arunski.

The hidden workplace works because it is outside the traditional structure; therefore, it can't be treated the same way. Companies must resist the temptation to try to micromanage it. They must also avoid making the opposite mistake - relying on it to solve problems that need structure, such as major cost reductions. The informal organization "is most helpful when you're trying to influence behaviors that are more emotional than they are rational," says Katzenbach. It "won't do everything for you."

The best managers hit the sweet spot between the two structures by trying to install loose parameters around organizations that often operate independently. At Fluor the mentoring circles are voluntary; so are Bell Canada's Pride Builders. Both groups are effective, however, because they have an explicit connection to the work.

Another key to striking the right balance is finding the right ambassadors to shuttle between the formal and the informal. Knowing who those people are is a challenge: Cross's data show that half the time, managers don't know who the best networkers are. So while smoking may be bad for your health, hanging out with the smokers - or others like them - may actually help keep your company in the pink.

Do you know your way around a water cooler? Take this quiz and find out. 

1 2