Disagreement to Convergence: Wisdom in Diversity

By Adam Kingl

Teams typically desire harmony above all. The most common fear is that of conflict. The assumption is that conflict and effectiveness are mutually exclusive, or even the cause-and-effect reasoning that harmony contributes to effectiveness. The bigger risk, however, is too much homogeneity and agreement. The irony is that the manager seeking harmony could in fact cause his or her team to underperform.

In this article, I will make a case for embracing diversity as an imperative for team effectiveness, accuracy and innovation. I will use examples from research, thought leaders’ opinions, popular television, consulting techniques and business practice.

What a would team look and feel like, if it sought disagreement before answers, if it did not equate vigorous dialogue with ineffectiveness, if it actively added to its diversity at every opportunity? Sounds like a nightmare couched in rhetoric? Not necessarily. Diversity of opinion as well as team composition, if managed well, can be a higher performing team than the country club atmosphere of the uniform team that values its homogeneity, harmony and speed to consensus. At the same time, this article assumes but does not explore in depth, that the multicultural manager would be more effective in managing a diverse team by virtue of his or her experience and perspective.

Let us first explore if diversity equals effectiveness. A fascinating study1 has explored whether or not social similarity helped or hindered a team to be more task-effective. The results showed that, with diverse opinions at play, perception is not reality. Socially similar teams evaluated their own perceived effectiveness as very high, whereas socially dissimilar teams did not judge themselves to be particularly effective. When the teams were assessed with a group task, their actual effectiveness was almost the opposite of their self-judgement. The socially dissimilar teams performed their task almost 50% more accurately than the socially similar teams.

[ms-protect-content id=”9932″]





We must ask that if the dissimilar teams were managed under principles that included celebrating their diversity, would they not only be more effective than homogenous teams, but recognise that fact? We may assume that in this situation, the team would have enhanced morale and motivation to complement its high performance. The challenge for the manager of the socially diverse team, therefore, is to create a norm that listens for disagreement and considers the opposing view rationally with the goal of achieving the best answer as paramount. In other words, teams can be both harmonious and effective if they do not equate disagreement with discord.

When we discuss team harmony above, we may picture a team that works together in the same space. To remove that assumption may reveal further advantages to manage a diverse team that is also virtual. There is a consulting company called How Might (www.howmight.com) that relies on voluminous, diverse and virtual opinions in order to add value to its clients.

How Might has very few full-time employees. When a client presents a business issue, How Might translates that issue into a compelling question, which may even begin with ‘How might…,’ hence the company name. The company then creates a digital platform to work toward a solution; it solicits its virtual and global network of experts in all manner of fields to contribute to answering the question. Some will present a few words, others more evolved thinking. In this initial iterative stage, the answer is completely open to discussion. Like many social media platforms, the network may disagree with one another, and the best possible answer will evolve as members add to and vote for its favourite ideas, similar to the manner in which Facebook encourages us to ‘like’ something. How Might’s value is predicated on the idea that diversity matters, that harmony is less effective than diverse people iterating toward an answer.

But we all recognise that the world of consulting can be very different from the world of business practice. Would the principle of diversity as a virtue still apply in industry? London Business School Professor Gary Hamel, in his brilliant book The Future of Management2, describes the example of Best Buy in soliciting diverse opinion to achieve better insight than the so-called experts. Best Buy’s Vice President of Consumer and Brand Marketing, Jeff Severts, was irked with the criticism he had received that his spend on advertising was not achieving the forecasted revenue. His theory was that the forecasts were too aggressive rather than the advertising not being effective. He set out to conduct an experiment to validate or disprove his hypothesis.

As a controlled experiment, he wanted to see if an average group of Best Buy employees could predict gift card sales more accurately than the financial experts in the organisation whose job was to forecast. Severts asked several hundred employees to participate in guessing gift card sales for the following month, having only very little historical data. The experts – the professional forecast team – were historically accurate within 5%. A tough test for the amateurs…

The several hundred employees were as diverse as the face of Best Buy: stockists, customer service representatives, store managers, back office functions. As it happens, their collective, average guess was only 0.5% off the actual sales figure. In other words, the employee sample was ten times more accurate than the ‘experts.’

To ensure that this result was not a fluke, Severts conducted another, larger experiment. This time he asked for forecasts of the company’s total sales, rather than a single product’s, over the holiday season. He [Severts] asked Best Buy’s experts, whose job is to make such forecasts, and he asked hundreds of random employees, incentivising them to make an estimate with a small prize for the most accurate guess. This time the experts’ forecast was off by 7%, and the average collective guess of the employees was off by only 0.1%, in other words 99.9% accurate!

The experts’ forecast was off by 7%, and the average collective guess of the employees was off by only 0.1%, in other words 99.9% accurate!

Now, intuitively, this example may not make sense. An amateur in any field would not typically outperform an expert. The difference lies in the diversity of the large crowd of employees. While no one of them would be expected to outperform the professional forecasters, as a group, they possess geographically and functionally nuanced information that the experts could not possibly retain. For example, the store manager in Minnesota may recognise that she is having a particularly harsh winter and so will discount her estimate, thinking that there will be fewer shoppers this year. The stockist in Miami will have noticed that the new store was placed on a busy shopping street, and so will be more optimistic in his guess. All of those tiny, local adjustments become part of the average estimate. At the same time, those individual forecasts that are widely off are pushed to the radical ends of the bell curve, and so influence the average less than the more widely held view. And there you have it – a concrete argument for leveraging both diverse and collective opinion.

If you’re still sceptical, I shall cite an example with which we are probably all familiar: Who Wants to Be a Millionaire®. This TV quiz show has been broadcast all over the world for well over a decade, in dozens of languages and local hosts. What always remains the same are the famous three Life Lines that the contestants are able to access to help them answer the questions. The hitch is that a contestant can only use each Life Line once, and so must choose when and how to use them very judiciously.

If you are one of the few who is unfamiliar with the Life Lines, here they are:

• Ask the Audience: Everyone in the audience votes for their preferred of the four possible answers on a keypad, and the contestant sees what percent of the audience have voted for each answer.
• 50:50: The contestant can reduce the four answers to two remaining possible answers.
• Phone a Friend: The contestant may call someone they know, and solicit his or her guess on the right answer within 30 seconds.

If you have watched this programme even a few times, you would have recognised that most contestants select Ask the Audience the first time they use a Life Line. The average contestant perhaps discounts the general ignorance of the average audience member, and so squanders this Life Line on the easiest question on which he or she may only require a little help. Conversely, the contestant typically saves his or her final Life Line on Phone a Friend. Once again, it is assumed that the ‘expert’ will be more effective than a diverse sample.

Just like the Best Buy story, the wisdom of the diverse crowd is proven to be more accurate than the single mind by a long margin. If the contestant selects the most popular answer from Ask the Audience, there is an error probability of only 6.2%, while Phone a Friend yields an error probability of 46%.3,4

Another study has demonstrated with a mathematical model that, if one wishes to maximise the chance of answering all fifteen questions correctly and so achieve the goal of winning the whole prize pot, one should use Phone a Friend at Question 11 and Ask the Audience at Question 14.5 Aspiring millionaires take heed!

As I write this, I am mindful that we conduct a team exercise on several London Business School Executive Education programmes that illustrates that the view of the expert in the team must still be given its due. I would argue that the smaller the team, the more this lesson is true. With larger teams, I recommend a balance between considering the expert view and leveraging the collective wisdom inherent in diversity and numbers. The other team dynamic that could occur is that the expert is used to being right, and may win a debate through the power of their own confidence and of having more data to hand.




The How Might business model partially uses the technology of social media to disintermediate any undue influence of a strong personality. Another solution may be to consider the views of the ‘crowd’ and the expert separately, so that one view does not influence or anchor the other, and then share for discussion. A final recommendation is to find ways in which the emerging collective view is not a victim of groupthink. The Best Buy example also demonstrates how technology prevents this risk.

Instead of following the cliché of ‘agreeing to disagree,’ perhaps embracing multiple and diverse points of view could lead to the stronger paradigm of ‘disagreeing to converge.

A diverse group’s mosaic of different views, some opinions perhaps ignorant or radical, balances the important but maybe narrow view of the expert. The challenge for the manager of such a team or organisation is to find incentives and create cultural norms that make the soliciting of views a regular occurrence, while identifying patterns, averages and trends in those views (as any individual opinion is not likely to be better than that of the expert). There can be no doubt that the multicultural manager would be an asset in his or her being open to working in the ways recommended here. Instead of following the cliché of ‘agreeing to disagree,’ perhaps embracing multiple and diverse points of view could lead to the stronger paradigm of ‘disagreeing to converge.’

About the Author
Adam Kingl is Director of Learning Solutions for Executive Education at London Business School, responsible for overall quality and design of its programmes. He is also an Associate of the Management Lab (MLAB). His areas of specialism include organisational learning, creativity and innovation, virtual teamwork, and high performing teams.

Adam has spoken at events including the CIPD, the Harvard Business School Alumni Club, and TiE – the global entrepreneurs’ network. He has contributed to publications such as The London Times, Business Strategy Review, Employee Engagement Today and South China Morning Post.

Adam has consulted to companies such as the BBC, Burson-Marsteller, BP, De Beers, Disney, LVMH, Oman Oil, Pixar, Tesco and Zurich. Adam holds degrees from London Business School, UCLA and Yale.

1. Phillips, K.W., Lillenquist, K.A., & Neale, M.A. 2008. Is the Pain Worth the Gain? The Advantages and Liabilities of Agreeing with Socially Distinct Newcomers.
2. Hamel, Gary. The Future of Management. Harvard Business School Press. 2007.
3. Franzen A. & Pointner S. Social Networks 2010. Calling social capital: An analysis of the determinants of success on the TV quiz show “Who Wants to Be a Millionaire?”
4. The 50:50 Life Line achieves an error probability of 15.7%.
5. Perea, F. & Puerto, J. Management Mathematics for European Schools 2001. A Simple Analysis of the TV Game ‘Who Wants to Be a Millionaire?’



Please enter your comment!
Please enter your name here