Stop messing around with collaboration. Many innovations are ruined by "putting people together".

Is breaking down departmental walls and enabling close teamwork the only way to innovation? Research reveals a counterintuitive answer: excessive collaboration may backfire. The success or failure of innovation doesn't hinge on "connection" itself, but on whether the right collaboration model can be matched to different types of innovation tasks. True leadership lies in making accurate judgments: when full - scale close collaboration is needed and when the independent thinking space of the team should be protected. This is not only a test of management wisdom but also a redefinition of an organization's innovation ability.

Break down departmental barriers, establish end - to - end connections, and achieve efficient collaboration.

This is a common piece of advice regarding innovation. Sometimes it really works. For example, during the COVID - 19 pandemic, Ford and GE Healthcare needed to produce 50,000 ventilators in 100 days. This intensive cross - functional collaboration was undoubtedly the right choice. At that time, every decision was interdependent, every insight was shared, and every problem was solved jointly.

However, applying the same method in different contexts won't lead to innovation; instead, it will systematically hinder it. When NASA needed to predict dangerous solar energetic particle events, the breakthrough didn't come from traditional collaboration. It originated from an open - innovation challenge. Bruce Cragin, a semi - retired engineer, explored a solution path in the countryside of New Hampshire that the NASA team would never have considered. For the same urgent need, opposite methods were used, and both were successful. This is not a contradiction.

After analyzing 294 empirical studies on collective innovation, we found that connecting people for collaboration has inconsistent effects on innovation outcomes. Sometimes it has a strong positive impact, sometimes a negative one, and sometimes no impact at all. In some cases, this relationship is not even linear; in situations where excessive interconnection inhibits innovation, maintaining a certain degree of "isolation" may be more constructive.

The reason for these different effects is that most innovation advice overlooks this: the structure of the collective itself. Innovation emerges under different conditions, which depend on how the collective is structured, how its members interact, and the degree of goal alignment among them.

Different Types of Collectives

Managing collective innovation starts with a question that most leaders skip: What kind of collective are you actually working with?

Organizations often assume that all collaborative innovation should adopt the same approach. However, our review research found that collectives differ in two fundamental aspects, which determine how they should be managed:

Search dependence: To what extent is the process of finding solutions interdependent? Do people need to learn from each other when exploring possibilities, or can they find solutions independently?

Goal alignment: To what extent do participants work towards the same overall goal rather than pursuing their own priorities?

Based on these two dimensions, three distinct types of innovation collectives emerge: Aggregate (interdependent search + common goal), Divergent (interdependent search + non - common goal), and Focused (independent search + common goal). The remaining combination - independent search and non - common goal - is not really a collective; it's just individuals operating in parallel.

1. Aggregate Collectives

When an innovation problem requires the close integration of expertise, an aggregate collective (either physically or virtually) brings the right people together in the same space to solve the problem. Members learn from each other's insights in real - time, provide feedback, challenge assumptions, and co - create solutions. This is the typical model for cross - functional product development teams or interdisciplinary R & D work. For example, when Alphabet's DeepMind needed to predict protein structures - a problem that had puzzled scientists for nearly 50 years - they gathered machine - learning engineers, biologists, and physicists to build AlphaFold. This breakthrough required the continuous integration of interdisciplinary insights, which no individual or independent group could achieve.

2. Divergent Collectives

Sometimes, the best innovation comes from multiple entities pursuing their own paths in a shared ecosystem. In a divergent collective, participants have different goals but benefit from the flow of information. They may compete, selectively cooperate, or simply learn from each other's experiments. Johnson & Johnson's JLABS network is an example. At 13 sites, more than 650 biotech startups are "unencumbered" in pursuing their own ideas but all benefit from the shared infrastructure, investor network, and knowledge of the broader ecosystem. Each company follows its own agenda, yet the innovation power of the entire network exceeds the sum of its parts.

3. Focused Collectives

For problems with clear goals and multiple possible solution paths, a focused collective can leverage independent work on a large scale. Participants share a common goal but work in parallel, exploring different methods without coordinating their ideas. The best solution emerges through screening rather than collaboration. When Meta needed to deal with deepfakes in 2020, they launched a $1 million reward challenge and received more than 35,000 solutions. Each participant worked independently, testing different detection models. The winning solution had an accuracy rate of 82.56%, and Meta also obtained thousands of alternative solutions that could never be generated internally.

How to Promote Innovation in Aggregate Collectives

Aggregate collectives can unleash innovation that no single contributor can achieve alone, but they place higher demands on managers: more resources, more coordination, and more deliberate design. Success depends on two key processes: acquiring diverse knowledge and integrating it into a coherent solution.

This starts with the composition of the collective. Most studies show that collectives composed of people with a wide range of disciplinary backgrounds and cognitive styles can produce stronger creative results. However, it's important to prioritize the diversity of professional backgrounds and expertise rather than relying solely on demographic differences.

Take how IDEO builds its teams as an example. This global design and innovation company assembles teams that mix business strategists, engineers, anthropologists, and designers. This cognitive diversity allows the team to approach problems from multiple perspectives, promotes the exchange of ideas, and helps IDEO create breakthrough products for clients ranging from Apple to the Mayo Clinic.

However, the value of this diversity depends on the ability to bridge differences. Research shows that cognitive diversity can enhance creativity when the collaborative dynamics encourage the constructive use of different viewpoints. To promote this integration, managers must create an environment of trust, psychological safety, and open communication.

Leadership is crucial in creating such an environment. For example, multiple studies show that transformational leadership and rotational leadership - rotating decision - making power among team members according to task needs - can more effectively utilize dispersed knowledge, thereby significantly promoting team innovation. This is mainly due to the strengthening of trust, engagement, and collaboration within the team.

Another effective method is to introduce incentives based on collective performance, such as profit - sharing or team rewards, to reinforce these behaviors. For example, a study of 22,000 European companies shows that such incentives promote organizational innovation by enhancing psychological ownership, information sharing, and teamwork.

Haier organized itself into thousands of "micro - enterprises," where any employee or team can initiate a startup and hold shares - creating value and directly sharing profits. This model has incubated 183 startups within the company, accelerating the launch of new products and achieving far better results than traditional structures.

How to Promote Innovation in Divergent Collectives

The advantage of divergent collectives lies in loosely coordinated exploration. Semi - autonomous units - often spanning companies or departments - pursue their own innovation paths while being embedded in a broader network. This model shows unique advantages when the direction forward is unclear and the challenge can benefit from multi - domain, parallel trial - and - error (such as in the face of emerging technologies or industry - wide transformation).

The management challenges are three - fold: building the right network, avoiding information overload, and developing the ability to absorb incoming information.

This is usually a delicate balance. More partners mean more perspectives, until it turns into chaos. If left unchecked, a large network will overwhelm the team, drowning it in redundant information, dispersing the focus, and slowing down the decision - making cycle. Therefore, leaders must carefully plan relationships that can bring in new information, protect the collective from overload, and build internal capabilities to absorb incoming information.

The most valuable innovation contributions come from partners with different ways of seeing the world. Establishing strategic partnerships with entities that can bring novel insights rather than redundant knowledge is crucial. For example, a study of the Japanese supply chain found that companies with close connections to geographically distant suppliers outperformed those only embedded in local networks. These long - distance connections introduced new thinking. In contrast, an analysis of more than 2 million inventors in 337 regions in the United States found that companies confined to dense local clusters had less access to new information, resulting in a decline in innovation productivity.

But there are limits. Spreading innovation efforts too widely can also pose risks. A study analyzing more than 500,000 patents from 1,127 companies found that although geographically dispersed R & D teams can access more diverse knowledge, they also face high integration costs, often reducing the quality of innovation output.

Take BMW's deliberate selective strategy in its "Startup Garage" as an example. The company screens more than a thousand startups every year and only quickly incorporates a few into its network. This strategy has been successful: it has completed cooperation projects with 220 companies in 26 countries, and 30 of them have become long - term suppliers.

One way to overcome this burden of breadth is to develop absorptive capacity - the systems and skills to process external ideas. Research consistently shows that companies with stronger absorptive capacity benefit more from external collaboration. For example, Novartis' "Biome" center can quickly connect with external partners, pair them with internal experts, and grant them access to a selected "sandbox" of clinical trial data. This enhances Novartis' ability to digest its large inflow of data, enabling faster and more informed innovation.

How to Promote Innovation in Focused Collectives

To unleash the full potential of focused collectives, managers must design an environment that can attract diverse contributors while maintaining the independence on which this model operates.

In focused innovation, external people are crucial. A study of 166 scientific challenges involving more than 12,000 participants found that contributors from outside the core field are more likely to produce breakthrough solutions. However, novelty alone is not enough. Another study of 230 scientific competition entries found that outsiders can produce better results only when they first have a basic understanding of the problem domain. The implication is: invite people from non - traditional backgrounds but provide them with enough background knowledge to guide their thinking.

To maintain creative independence, leaders must avoid over - structuring the process. Research shows that exposing contributors to others' ideas too early may backfire, as it introduces cognitive anchoring, thereby narrowing the solution space. In a series of experimental studies, increasing peer visibility reduced originality, especially when individuals conformed to early - submitted solutions. The best results often occur when contributors are not influenced by others' thinking.

Incentives are also important, but in a different way than many people think. A study of 646 problem - solvers on a large crowdsourcing platform found that intrinsic motivation (such as the joy of solving problems) and extrinsic motivation (economic rewards, recognition) play complementary roles in driving the quality of innovative solutions. The most effective design usually takes both into account.

In terms of extrinsic incentives, the traditional "winner - takes - all" competition may suppress participation and reduce quality. According to a study of 261 innovation competitions, a more effective model is to offer multiple smaller awards and allow multiple wins. This method increases participation, diversifies contributions, and brings higher - quality solutions over time. Google's bug bounty program is an example: in 2024, it paid $11.8 million to 660 researchers, and many people received multiple bonuses for different discoveries.

Feedback, when used well, is also a powerful lever. In two longitudinal experiments and a survey of more than 1,500 innovation managers, constructive feedback focused on feasible improvements significantly increased early participation and led to better final results. Encouraging iterative improvement can transform isolated ideas into mature solutions. Even observing feedback to others can help. A study of more than 55,000 participants in a T - shirt design competition and on the e - commerce platform Threadless found that simply seeing how others were evaluated could help contributors calibrate their responses. In the long run, this visibility of feedback promotes learning, leading to better solutions.

Designing the Right Collective for Innovation

Our research reveals what most innovation advice overlooks: there is no one - size - fits - all strategy for collective innovation. Management practices that drive breakthroughs in one type of collective may systematically hinder them in another.

Take information sharing as an example. When a problem requires the integration of expertise, making everyone's ideas visible is crucial - real - time integration is how breakthroughs occur. But when a problem requires independent exploration, the same visibility can anchor thinking and narrow the solution space.

Then there's coordination. Aggregate collectives require close alignment because the work itself is interdependent. However, focused collectives rely on autonomy - the diversity of solutions comes from contributors exploring different paths without influencing each other.

Both the Ford - GE Healthcare ventilator project and NASA's solar particle prediction achieved success through collective innovation. However, they required opposite management methods because they had different collective structures.

Most organizations default to the aggregate approach because it feels efficient - bringing smart people together, maximizing the flow of information, and promoting consistency. Sometimes this intuition is completely correct. But when the problem actually requires independence or loose coordination, the same intuition becomes a constraint.

Exceptional innovation leaders understand this well: sometimes "breaking down barriers" is the solution, and sometimes it's the problem itself. The difference - managing collaboration at a general level or strategically structuring the collective - distinguishes innovation performance from innovation ability.

Keywords: #Organizational Management

Oguz A. Acar, Aybars Tuncdogan | By

Oguz A. Acar is a professor of marketing and innovation at King's Business School, King's College London. Aybars Tuncdogan is an associate professor of digital innovation and information security at King's Business School, King's College London.

Zhou Qiang | Edited

This article is from the WeChat official account "Harvard Business Review" (ID: hbrchinese). Author: HBR - China. Republished by 36Kr with permission.