What can give the wisdom of the crowd a boost? And with AI now in the picture in many aspects of decision-making, is it enabling us to truly make better choices? 

This month, INSEAD faculty offer insights across the fields of leadership and organisations, marketing, as well as finance. These include how group discussions improve team performance and what happens to the quality of decisions when AI isn’t well-designed into the process.

The researchers also uncover how allowing trade-ins could affect sales and profitability in the retail business, why studying consumers’ food choices over time reveals much more than what they eat, and how leveraged buyouts can narrow the wage gap.

Group discussions help better harness the wisdom of crowds

How can we obtain more knowledge from a group of people? Group discussions can help individuals better calibrate their self-perceived competence, which, in turn, can facilitate harnessing wisdom out of the group. This is according to research by Enrico Diecidue and his co-authors.

They show that discussions reveal a truer picture of the actual competence of individuals and can be a simple exercise to help calibrate self-perceived competence (which tends to be poor in accuracy). The improvement in performance post-discussion suggests that it can be an easy-to-implement approach to effectively exploit crowd wisdom.

Read the full paper

Group discussions help harness the wisdom of crowds

Human-AI collaboration is often deployed on the premise that algorithms provide precision while human decision makers provide contextual judgement. But this complementarity has its limits, say Michael Freeman and Liang Xinyu, whose study focused on radiologists evaluating chest X-rays with varied access to AI predictions and clinical history. 

Their findings suggest that AI changes attention rather than effort. With AI, radiologists remain active but engage less with clinical history. As a result, their assessments stay closer to the AI prediction. The findings show that AI advice can crowd out human contextual knowledge rather than simply complement it. Therefore, effective AI deployment depends not only on model accuracy, but also on when and how AI advice enters the decision process.

Read the full paper

Retail trade-ins: Impact on sales, returns and profitability

Some retailers give customers the option to trade-in a used product to offset payment for a new one, and also to return a new product purchased through a trade-in. To understand how such schemes affect sales and profitability, Atalay Atasuand his co-authors studied the trade-in programme of a national jewellery retailer. 

As the retailer increased the number of stores offering trade-ins, sales increase by 11.4% while return rates increased by 4.3 percentage points for eligible products. For ineligible products, sales increase by 2.7% – with no returns. Offering the option to trade-in products introduces a trade-off for retailers: It leads to higher sales, but also higher product returns, which can hurt profitability. Retailers must therefore carefully assess the trade-in eligibility of different products to ensure profitability. 

Read the full paper

The effect of leveraged buyouts on pay gaps

Following a leveraged buyout, target firms tend to see within-firm pay gaps narrow and profitability increase, relative to firms that didn’t undergo a leveraged acquisition. This insight is revealed by Lily Fang, Alexandra Roulet and their co-author, having analysed two decades of French administrative data on leveraged buyouts.

Generally, employee turnover tends to narrow the pay gap, with the effect being more pronounced at the top of the wage distribution because employees who are let go are paid more (and new joiners less) than similar employees, especially among skilled workers. Leveraged buyouts amplify this effect due to the increased turnover among managers.

Read the full paper

A new framework for studying how people make food decisions

Consumers make a multitude of decisions about what to eat every day, influenced by internal motivations and external factors. Importantly, these choices are highly connected over time. To better identify consumption patterns that reflect consumers’ everyday choices, Pierre Chandon and his co-authors propose a continuous-time approach to studying consumption, rather than approaching eating as individual episodes. 

Under the Start-Stop Continuous-Time Food Framework, each eating episode is characterised through three key interrelated dimensions: when to start eating, what to eat and when to stop eating. The relationships between these eating episodes over time could reveal a fuller picture of consumption patterns. The framework could pave the way for consumers to meet their goals and enable the design of more effective policies.

Read the full paper