Apr 01, 2009
When Internal Collaboration Is Bad for Your Company
Internal collaboration is almost universally viewed as good for an organization. Leaders routinely challenge employees to tear down silos, transcend boundaries, and work together in cross-unit teams. And although such initiatives often meet with resistance because they place an extra burden on individuals, the potential benefits of collaboration are significant: innovative cross-unit product development, increased sales through cross-selling, the transfer of best practices that reduce costs.
But the conventional wisdom rests on the false assumption that the more employees collaborate, the better off the company will be. In fact, collaboration can just as easily undermine performance. I’ve seen it happen many times during my 15 years of research in this area. In one instance, Martine Haas, of Wharton, and I studied more than 100 experienced sales teams at a large information technology consulting firm. Facing fierce competition from such rivals as IBM and Accenture for contracts that might be worth $50 million or more, teams putting together sales proposals would often seek advice from other teams with expertise in, say, a technology being implemented by the prospective client. Our research yielded a surprising conclusion about this seemingly sensible practice: The greater the collaboration (measured by hours of help a team received), the worse the result (measured by success in winning contracts). We ultimately determined that experienced teams typically didn’t learn as much from their peers as they thought they did. And whatever marginal knowledge they did gain was often outweighed by the time taken away from their work on the proposal.
The problem here wasn’t collaboration per se; our statistical analysis found that novice teams at the firm actually benefited from exchanging ideas with their peers. Rather, the problem was determining when it makes sense and, crucially, when it doesn’t. Too often a business leader asks, How can we get people to collaborate more? That’s the wrong question. It should be, Will collaboration on this project create or destroy value? In fact, to collaborate well is to know when not to do it.
How Collaboration Can Go Wrong
In 1996 the British government warned that so-called mad cow disease could be transferred to humans through the consumption of beef. The ensuing panic and disastrous impact on the worldwide beef industry over the next few years drove food companies of all kinds to think about their own vulnerability to unforeseen risks.
The Norwegian risk-management services firm Det Norske Veritas, or DNV, seemed well positioned to take advantage of the business opportunity this represented by helping food companies improve food safety. Founded in 1864 to verify the safety of ships, DNV had expanded over the years to provide an array of risk-management services through some 300 offices in 100 countries.
In the fall of 2002 DNV began to develop a service that would combine the expertise, resources, and customer bases of two of the firm’s business units: standards certification and risk-management consulting. The certification business had recently created a practice that inspected large food company production chains. The consulting business had also targeted the food industry as a growth area, with the aim of helping companies reduce risks in their supply chains and production processes.
Initial projections for a joint effort were promising: If the two businesses collaborated, cross-marketing their services to customers, they could realize 200% growth from 2004 to 2008, as opposed to 50% if they operated separately. The net cash flow projected for 2004 through 2008 from the joint effort was $40 million. (This and other DNV financial figures are altered here for reasons of confidentiality.)
The initiative was launched in 2003 and run by a cross-unit team charged with cross-selling the two types of services and developing new client relationships with food companies. But the team had trouble capitalizing on what looked like a golden opportunity. Individual business unit revenue from areas where the existing businesses had been strong—Norway for consulting services, for example, and Italy for certification—continued to grow, exceeding projections in 2004. But the two units did little cross-pollination in those markets. Furthermore, the team couldn’t get much traction in the United Kingdom and other targeted markets, which was particularly disappointing given that the certification group had established good relations with UK food regulators in the years following the outbreak of mad cow disease.
As new business failed to materialize, the consulting group, which was under pressure from headquarters to improve its overall results in the near term, began shifting its focus from the food industry to other sectors it had earlier targeted for growth, weakening the joint effort. The certification group continued to make the food industry a priority, but with the two groups’ combined food industry revenue lagging behind projections in 2005, DNV abandoned the initiative it had launched with such optimism only two years before.
Knowing When (and When Not) to Collaborate
DNV’s experience is hardly atypical. All too often plans involving collaboration among different parts of an organization are unveiled with fanfare only to collapse or fizzle out later. The best way to avoid such an outcome is to determine before you launch an initiative whether it is likely to yield a collaboration premium.
A collaboration premium is the difference between the projected financial return on a project and two often overlooked factors—opportunity cost and collaboration costs. In simple form:
Collaboration Premium Equation
The projected return on a project is the cash flow it is expected to generate. The opportunity cost is the cash flow an organization passes up by devoting time, effort, and resources to the collaboration project instead of to something else—particularly something that doesn’t require collaboration. Collaboration costs are those arising from the challenges involved in working across organizational boundaries—across business units, functional groups, sales offices, country subsidiaries, manufacturing sites. Cross-company collaboration typically means traveling more, coordinating work, and haggling over objectives and the sharing of information. The resulting tension that can develop between parties often creates significant costs: delays in getting to market, budget overruns, lower quality, limited cost savings, lost sales, damaged customer relationships.
Including collaboration costs makes this analysis different from the usual go/no-go decision making for proposed projects. Obviously, such costs can’t be precisely quantified, especially before a project is under way. Still, with some work you can arrive at good approximations. Given how much time managers already spend estimating the return on a project—and, occasionally, the associated opportunity cost—it makes sense to take the additional step of estimating collaboration costs, particularly because they can doom a project.
If, after going through this exercise, you don’t foresee a collaboration premium—or if a collaboration penalty is likely—the project shouldn’t be approved. Indeed, this sort of analysis might have helped DNV steer clear of a promising but ultimately costly business venture.
Avoiding Collaboration That Destroys ValueIn calculating the collaboration premium, it’s important to avoid several common errors.
Don’t overestimate the financial return.
Whether because of enthusiasm for collaboration or the natural optimism of managers, many companies place a mistakenly high value on collaboration. Especially when a team’s work appears to be a model of collaboration—the parties freely share resources and cooperate in resolving differences while coming up with nifty ideas—it may be easy to overlook the fact that the work is actually generating little value for the company. Never forget that the goal of collaboration is not collaboration but, rather, business results that would be impossible without it.
When a team’s work appears to be a model of collaboration, it may be easy to overlook the fact that it generates little value for the company.
In numerous well-known instances, collaboration premiums failed to materialize. Daimler’s $36 billion acquisition of Chrysler in 1998—with its promise of synergies between the two automakers—and the sale nine years later of 80% of Chrysler for a pitiful $1 billion constitute only the most conspicuous recent example. But collaboration’s benefits are usually overvalued in much more mundane settings. Recall how the experienced sales teams at the IT consulting firm that Martine Haas and I studied shared expertise as a matter of course during the preparation of project proposals—never stopping to seriously consider whether they, in fact, benefited from doing so.
Don’t ignore opportunity costs.
Executives evaluating any proposed business project should take into account the opportunities they will forgo by devoting resources to that project. If the project requires collaboration, it’s important to consider alternative noncollaborative activities with potentially higher returns. The opportunity cost is the estimated cash flow from the most attractive project not undertaken.
DNV didn’t overestimate the potential financial return of its food initiative, but it did fail to assess the opportunity cost. “There was no consensus at the top level that food was interesting or a priority,” said one senior manager. “We had not evaluated the food opportunity against other industry segments.” In fact, food was only one of several sectors—including information technology, health care, and government—that DNV’s consulting unit had targeted in 2001 as offering growth potential for its risk-management services. The opportunity in IT, which the consulting unit could have pursued on its own, undoubtedly had more potential. The unit made progress in 2004 generating new business in this sector, but it was constrained by a shortage of qualified consultants, some of whom were tied up with the food initiative. To pursue the food initiative, the consulting unit had to forgo additional business from the IT opportunity. I estimate the cost of this forgone opportunity at $25 million or more in lost cash flow.
Don’t underestimate collaboration costs.
In most companies, it’s difficult to get people in different units to work together effectively. Issues relating to turf, such as the sharing of resources and customers, often make groups resistant to collaboration. Individuals may resent taking on extra work if they don’t get additional recognition or financial incentives. Even when collaboration delivers obvious and immediate benefits to those involved (for example, one unit’s software package solves another’s current problem), blending the work of two units that usually operate independently creates impediments.
These costs, which should be assessed before committing to a cross-unit project, can be tough to identify and quantify. And they will vary depending on the collaboration culture of an organization. But although they can be reduced over time through companywide efforts to foster collaboration, it’s a mistake to underestimate them in the hope that collaboration can be mandated or will naturally improve during the course of a project.
It’s a mistake to underestimate collaboration costs in the hope that collaboration can be mandated or will naturally improve during the course of a project.
As DNV decided whether to move forward with its food initiative, the project managers failed to consider the substantial collaboration costs the company would incur because it wasn’t set up to collaborate. Mistrust between the consulting and certification units escalated as they tried—unsuccessfully, and with much quarreling—to build a common customer database. “All the team members tried to protect their own customers,” one manager in the certification group admitted. Because of the reluctance to share customer relationships, the team had to significantly reduce its estimates of the revenue to be generated by cross-selling.
Individual members of the cross-unit team were also pulled by conflicting goals and incentives. Only one team member was dedicated to the initiative full-time; most people had to meet individual targets within their respective units while also working on the joint project. Some people got a dressing down from their managers if their cross-unit work didn’t maximize their own unit’s revenue.
Even those who saw the benefits of the initiative found it hard to balance their two roles. “We all had personal agendas,” said one senior manager in the certification group. “It was difficult to prioritize the food initiative and to pull people out of their daily work to do the cross-area work.”
Although assigning a financial number to collaboration costs is difficult, I estimate that the cash flow sacrificed as a result of tension between the two groups, which scotched probably one in two cross-selling opportunities, was roughly $20 million.
Had the likely opportunity and collaboration costs of DNV’s food-safety project been estimated, the project would have looked decidedly less attractive. In fact, managers would have seen that, rather than a collaboration premium, it was likely to yield a collaboration penalty of something like $5 million—that is, the projected return of $40 million less an opportunity cost of $25 million and collaboration costs of $20 million.
How Collaboration Can Go Right
That’s not the end of DNV’s story, however. Several months after the firm abandoned the food-safety initiative, Henrik Madsen was named CEO. He had seen firsthand the poor business results, wasted management effort, and ill will spawned by the initiative, having been head of the certification unit at the time. But he also believed that performance could be enhanced by collaboration at the traditionally decentralized DNV.
Madsen quickly reorganized the firm into four market-oriented business units and began looking for collaboration opportunities. His executive committee systematically evaluated all the possible pairings of units and identified a number of promising opportunities for cross-selling. The unit-by-unit analysis also revealed something else important: pairings that offered no real opportunities for collaboration—an insight that would prevent wasted efforts in the future.
The disciplined process prompted the committee to assess the potential financial return of each opportunity. Estimates totaled roughly 10% of the company’s revenue at the time. The projected returns helped the committee prioritize options and assess the opportunity cost of choosing one over another. On the basis of these findings, along with an assessment of likely collaboration costs, the company launched a round of collaboration initiatives.
One of these involved the maritime unit, which provides detailed classification of vessels for companies in the shipping industry, and the IT unit, which specializes in risk-management services for IT systems in many industries. Because ships today operate using sophisticated computer systems, someone needs to help shipping companies manage the risk that those systems will malfunction at sea. There was a clear opportunity to sell IT’s services to the maritime unit’s customers—if effective collaboration could be achieved between the two units. That opportunity has already borne fruit: The IT unit won a contract to develop information systems for a huge cruise ship being built by a longtime customer of the maritime unit.
The IT unit has also collaborated with the company’s energy business to jointly sell services to oil and drilling companies—another opportunity identified in the executive committee’s review. That effort enhances the IT unit’s service offering with the energy unit’s oil and gas industry expertise, a package that most IT competitors can’t match. The two units split the revenue, which creates incentives for both.
In pursuing opportunities like these, DNV has worked to reduce some of the typical costs of collaboration. Annie Combelles, the chief operating officer of the IT business, says there was an obvious market for her unit’s services among customers of the maritime and energy units. “My concern was that those units understand what we could deliver,” she says. “My concern was internal, not external.” The IT group appointed a business development manager who had worked at DNV for 12 years, including a stint in the maritime unit, and had a broad personal network within the company. This made him a trusted and knowledgeable liaison to the maritime and other units, reducing potential conflict between them and the IT unit.
What’s more, the IT unit has moved cautiously in trying to capitalize on opportunities for internal collaboration. Although the maritime group’s longtime relationship with the cruise ship operator provided entrée for the information technology group, maritime didn’t want any missteps from IT to jeopardize that valuable relationship. IT therefore initially proposed a risk-assessment project in nonvital areas of the ship such as the “hotel” function, which included the Wi-Fi network, gambling computers, and the 5,000 personal computers to be used by guests. It evaluated each of these systems and identified 30 risks. This success led to a project involving vital areas of the ship, such as the power-management and positioning systems.
DNV’s renewed effort to encourage cross-unit collaboration is a work in progress that has nonetheless already produced some hard results: The portion of the IT unit’s sales that came from cross-unit collaboration climbed from almost nothing to 5% in 2008, and is projected to be 10% in 2009 and 30% the following year.• • •
Business leaders who trumpet the benefits of working together for the good of the organization are right in seeing collaboration’s tremendous potential. But they should temper those exhortations with the kind of analysis I’ve described here, which provides needed discipline in deciding when collaboration creates—or destroys—value. Ideally, as organizations become better at collaboration, through incentives and shifts in corporate culture, the associated costs will fall and the percentage of projects likely to benefit will rise.
Although the collaboration imperative is a hallmark of today’s business environment, the challenge is not to cultivate more collaboration. Rather, it’s to cultivate the right collaboration, so that we can achieve the great things not possible when we work alone.
A version of this article appeared in the April 2009 issue of Harvard Business Review.