In early 2000, a pharmaceutical giant and an international food group created a joint venture to develop and market a range of healthy foods. Analysts viewed it as the optimal platform to take advantage of the emerging functional food market, estimated to be worth $40bn in the US alone.

Newly developed products were scheduled to hit the shops before the end of the year. In reality, test marketing did not start until the following summer. Indeed, two years after being formed, with the market still waiting for a first product, the joint venture was dissolved. The basic problem, according to analysts, was the drug company’s inability to approach the market with a consumer mentality.

Such outcomes are by no means rare. Our research on co-innovation shows that partner companies can have a winning strategic fit, and yet fail to make the collaboration work. Of course, co-innovation is a particularly uncertain form of alliance, both in terms of the outcome (gains) and the process. Such uncertainty can be heightened by the starting conditions. Sometimes the partners know each other well. They might have a solid customer-supplier relationship, like the co-development agreements between Bang & Olufsen and its system suppliers. But the partners might also be competitors, such as GM and Ford, who combined to conceive fuel-efficient gearboxes. At the other extreme, the parties may have little prior awareness of one another, like Colgate-Palmolive and Nestle, who joined forces to produce plaque-fighting dental gum. Or they may have asymmetrical knowledge of each other, as in the case of a small software company and a large electronics business.

Regardless of the nature of the union, the dynamics of collaboration are tricky. Routine misunderstandings can quickly destabilise the relationship, leading to mistrust, missed milestones, lack of innovation, mutual dissatisfaction and, ultimately, collaborative failure.

To help avert – or, indeed, rescue such situations – it is critical to understand how these dynamics escalate.

On thin ice

Partners may have an initial feel for each other’s trustworthiness based on their past records and respective reputations for fair dealing, and any direct prior experiences they may have had with each other. But the basis for trust at the outset of a co-innovation agreement – possibly fuelled by eventful negotiations – is normally fragile. This is especially true if, as is increasingly the case, the partners are also competitors.

An extreme example is the partnership between tyre makers Michelin and Goodyear. When they signed their “historic” collaboration on run-flat tyre technology, Goodyear’s chairman, Robert Keegan, noted wryly: “Right now, outside this room, across the globe, Goodyear’s and Michelin’s sales and marketing organisations are beating each other’s brains out for new business – as they should.”

The rivals may have serious misgivings about the unintended transfer of skills or telegraphing of future strategies. And clearly, in such partnerships, the levels of mutual vigilance may be intense right from the outset.

Collaborative doubts

Working jointly to develop new products is an unpredictable process. There are bound to be delays, communication issues and adaptation difficulties, which can easily raise doubts about the capabilities or intentions of the chosen partner. There may be concerns both about the size of the prospective cake and how it will be apportioned.

In situations where trust is low, deviations from expectations – in terms of behaviour or outcomes – may take on disproportionate significance. Indeed, there is evidence that in social transactions the human mind is extremely alert to indications of cheating. First proposed by the sociobiologist Robert Trivers, this insight was later supported by researchers who gave people logical puzzles to solve. The participants cracked the problem far more easily if the scenario involved the possibility of someone benefiting at someone else’s expense – compared with framing the puzzle in an emotionally neutral way – even though the two ways of putting the problem were logically equivalent. Thus, partners may be oversensitive to potential or perceived breaches of trust, which are liable to generate anxieties and misunderstandings.

The potential for these types of anxieties is especially high if there is also a cross-cultural dimension to the alliance. For example, in the Volvo-Mitsubishi joint venture, the Volvo designers infuriated the Japanese production engineers by proposing last-minute changes. The problem stemmed from different understandings of the endpoint of the design and development task, and the fact that this was not “the Japanese way”.

Culturally different approaches to work are reflected in everything from scheduling and decision-making protocols to etiquette, and are easily mistaken for attempts to frustrate or take advantage of the other partner. As the chairman of a US company involved in steel-processing asserted: “Two years ago, a large and prestigious German company proposed a joint venture, and we bit. Now, I’m convinced their motivation was to keep us out of the market because we had something they didn’t. Of course, they don’t say so, it’s always something like: ‘This isn’t quite what the European market wants,’ or ‘we must redesign this part of it.’”

Different cultural grids, combined with wariness, are more likely to lead to negative interpretations of ambiguous situations.


When one partner starts to suspect a lack of commitment, goodwill or competence from the other party, various actions may ensue. Typically, monitoring and controls will be stiffened, contractual agreements may resurface and full reporting requirements may be enforced. Alternatively, expectations may be revised downwards, with corresponding cuts in resources and/or the quality of personnel assigned to the project. Or, fearing opportunistic behaviour, a partner may become more guarded about sharing information or ideas.

Unfortunately, such actions often fail to resolve the perceived problem. Instead, they further undermine the relationship by indicating a lack of trust and commitment, thus encouraging the other party to respond in kind. For example, one might expect the partner to hit back by adopting a strictly contractual view of the agreement, refusing to help on areas beyond the scope of the original provisions, such as adapting to partner requests or to market changes. Perversely, even a positive response may not allay the other partner’s fears. Research suggests that when we supervise others closely and see them do nothing wrong, we tend to imagine that the only reason they aren’t doing things wrong is because we are watching them – and so we start to trust them even less. Either way, we have the makings of a vicious cycle.

From bad to worse: an illustration

In one disintegrating collaboration studied in detail by Anca Metiu of Insead, software developers in a US company became very critical of their counterparts at an Indian partner. A dramatic increase in time-to-market pressures had led the US team to neglect the (previously productive) relationship. The Indian developers started grumbling about the lack of guidance, severely delayed feedback on their work and sudden changes in direction. Their junior status in the relationship made them especially sensitive to any perceived lack of respect or consideration. Predictably, the quality of their contributions and practicality of their suggestions suffered.

Oblivious to their own role in the problem, the US developers sensed that their Indian counterparts were not fully engaged. The US developers became increasingly critical of the work coming out of India and started manoeuvring to take control of all the modules of the project. As the ownership of the project shifted to the US, the need to communicate with India decreased, intensifying negative interpretations of the US partner’s intentions and encouraging some Indian developers to leave the project. This re-enforced the feeling among US developers that the project needed to be brought completely in-house.

When suspicions mount, collaboration partners often start sticking labels on one another, like unreliable, unreasonable, profiteering or paranoid. In this case, the Indian partner came to be seen as uncommitted, while the US partner was regarded as unhelpful, but also devious in its attempts to re-appropriate the project. Unfortunately, viewing each other through such lenses has a tendency to distort whatever new information is received – even if it happens to be contradictory.

Labels orient people’s attention and the way they make sense of data or events. In particular, the US developers focused on any work from India that seemed below par – and systematically brought these lapses to attention. If one Indian developer put a foot wrong, the whole US team knew about it. Moreover, US explanations for inferior work invariably focused on individual factors such as effort or ability, while neglecting external factors entirely. These biases were accentuated by steady turnover on both sides of the project.

More surprising still, reports that the Indian developers were working long hours further re-enforced their poor image. It merely fuelled US suspicions that the Indian developers must be working on parallel projects. In reality, deprived of US input, they were spending inordinate amounts of time trying to figure out the documents, instead of building on the work they had received.

Such information processing biases and stereotypical thinking intensify under conditions of stress – and clearly the US developers were under severe pressure to bring the product to market. The irony is that their behaviour actually delayed that objective and nearly sank the whole project.

Once the partners no longer expect the collaboration to last, each tends to take a short-term view and to disguise or limit its contribution – until the relationship loses its reason for being. There are two ways of addressing such problems: through intervention and prevention.

Back on track

The difficulty in trying to restore a floundering relationship is that each partner blames the other for being the source of the malaise and views its own efforts as responses to “bad behaviour”, rather than contributing factors.

If the alliance retains its strategic logic, the partners need to consider jointly how the relationship has veered off course. We sometimes compare this to a medical intervention.

Symptoms First, the partners need to agree on the shortfalls. What are the unmet objectives, and where has each partner experienced difficulties in working together? The two sides may actually highlight different starting points to the conflict and different pivotal events along the way. Indeed, some of the alleged “harms” by the other party may not even relate to actions, but rather to sins of omission – such as a failure to give sufficient advance warning regarding changes.

Diagnosis Both parties will have contributed – albeit in different ways – to creating the situation, and they need to explore how. Divergent work practices – mismatched budgeting cycles, incompatible decision tools, different decision-making styles, different work routines, communication and conflict resolution styles – may have contributed to mix-ups or delays. Their respective contexts or interests may also have altered over time. Uncommunicated pressures and changes may have driven “the other side” to behave badly (from your perspective) or rationally (from theirs).

Treatment The resolution must be seen as based on principles that both partners can accept as reasonable. This may mean taking a step back to discuss the principles being used. By doing this, the partners are more likely to arrive at joint decisions that they can both understand, accept and implement.

Going forward, the two parties need to rebuild trust, keep the lines of communication open and remain alert to their partner’s perspective of the alliance. They also need to agree on a process for handling future issues and bringing them up for discussion before they escalate. One possibility is to appoint an individual dedicated to the collaboration whose role it is to maintain communication, to shepherd the project along and to get senior management involved if necessary.

This was one of the solutions adopted in the US-Indian alliance described earlier. The US company appointed a manager specifically to liaise with the Indian partner. Other measures included increasing personnel exchanges. Thus, a few US engineers visited India to help put faces to names and to get a better feel for the context. Similarly, several Indian engineers were seconded to the US for overlapping periods of several weeks. The partners also introduced weekly conference calls (at rotating hours) and daily updates.

The relationship between the US business and its Indian partner was restored, but at no small cost. First, several experienced Indian developers had been lost to the project. Then, the US managers had to devote significant time and effort to getting the remaining Indian team members to re-engage with the process. They also had to work hard to make sure that scheduled conference calls were properly attended on the US side. Finally, they were obliged to grant matching holidays to the US developers for each day spent at the partner site.

Ultimately, the product was successfully launched. But a painful and costly process of reconciliation could have been avoided.

Set up to succeed

In terms of prevention, an awareness of the mechanisms described in this article is already half of the battle. Taken individually, these mechanisms are far from insurmountable. But their real power comes from their interaction, which triggers self-fulfiling and re-enforcing processes. The resulting vicious cycles mystify cause and effect and conflate task conflict and relationship conflict.

People working across organisational boundaries (including internal ones) need to know about these processes. It helps them to think about their actions from the other party’s perspective – and to monitor their own behaviour. It makes executives more resistant to applying loose labels and snap judgments – and mindful of their polarising and self-perpetuating effects. It promotes not just frequent contact, but also the importance of understanding each other’s contexts to avoid personal attributions. And it underlines the need for feedback forums where budding problems can be expressed and dealt with early in the partnership.

Beyond awareness, if there is one overriding principle for avoiding such situations, it is to invest in them from the start of the collaboration. The partners need to devote time to framing the relationship, clarifying the conditions and rules and sharing information and norms. But they also need to invest in the relationship, to establish a buffer of trust that will make it easier to confront problems early.

A common problem is that the understanding and trust built up between the negotiators does not quite carry over to people who have to make the partnership work. As the head of one international cosmetics group conceded: “Our approach to partnerships has evolved over time. Now, partnership negotiations might start out with the top officers, but early on we bring in a multifunctional team – those people who have to manage the alliance and live with it.”

It is a matter of gaining a better awareness of the partner’s aims, competences and internal methodologies in order to anticipate better any potential clashes – and where warnings or explanations will be necessary.


Poor dynamics between partners often take root at the outset. After that, the parties may expend a lot of time and energy trying to salvage the situation, often to little avail. We would urge partners to invest more time up front in order to set up more positive dynamics in the first place.

However compelling the strategy and guiding principles of a partnership, success still depends on the attitudes and behaviour of the human beings involved and, in particular, of these individuals’ ability to recognise and interrupt, or better yet prevent, the destructive cycle of mistrust and under-performance.

Jean-François Manzoni is a professor of leadership and organisational development at IMD

Jean-Louis Barsoux is a senior research fellow at IMD

They are the authors of “The set-up-to-fail syndrome: How good managers cause great people to fail” (Harvard Business School Press, 2002)

Further Reading

• Buerger, D. and Nuelle, F., “Strategic partnering: Roundtable,” Chief Executive (1995)

• Metiu, A., “Faraway, so close: Code ownership over innovative work in the global software industry,” unpublished dissertation, the Wharton School (2001)

• Mikolajczyk, S.J. “Historic pact: Goodyear-Michelin alliance to promote run-flat standard,” Tire Business, (2000).

• Miller, D.T., Visser, P.S. and Staub, B.D. “How surveillance begets perceptions of dishonesty: The case of the counterfactual sinner,” Journal of Personality and Social Psychology (2005)

• Reuer, J.J., Strategic alliances: Theory and evidence, Oxford University Press (2004)

• White, S. and Ishii, S., “Mitsubishi and Volvo: Managing co-operation in NedCar,” Insead (2003)

Copyright The Financial Times Limited 2023. All rights reserved.
Reuse this content (opens in new window) CommentsJump to comments section

Follow the topics in this article


Comments have not been enabled for this article.