On this page, you can find an overview of projects of INSEAD faculty and PhD students using randomized field experiments. These projects are grouped according to their development stage.

Completed projects


Journal of the Academy of Marketing Science (forthcoming)

To examine whether four pre-selected front-of-pack nutrition labels improve food purchases in real-life grocery shopping settings, we put 1.9 million labels on 1,266 food products in four categories in 60 supermarkets and analyzed the nutritional quality of 1,668,301 purchases using the FSA nutrient profiling score. Effect sizes were 17 times smaller on average than those found in comparable laboratory studies. The most effective nutrition label, Nutri-Score, increased the purchases of foods in the top third of their category nutrition-wise by 14%, but had no impact on the purchases of foods with medium, low, or unlabeled nutrition quality. Therefore, Nutri-Score only improved the nutritional quality of the basket of labeled foods purchased by 2.5% (-0.142 FSA points). Nutri-Score’s performance improved with the variance (but not the mean) of the nutritional quality of the category. In-store surveys suggest that Nutri-Score’s ability to attract attention and help shoppers rank products by nutritional quality may explain its performance.
Click here to download the paper.

AEA Papers and Proceedings (forthcoming)

This paper explores the relationship between individual identity and organizational attachment. Using individual data from employees at a large employer in the services sector we show that making individual values salient (through a value affirmation) on average reduces organizational attachment. However, this effect is heterogenous across individuals: those initially attached to the organization increase their attachment while, those who started off less identified with the organization reduce their attachment. Overall, the results illustrate the importance of heterogeneity and how individual identity/values and organizational attachment can conflict.


Strategic Management Journal, 2020

Winner, Best Proposal Award for Corporate Social Responsibility and Sustainability track, Strategic Management Society 2018 India Conference.

The emerging practice of “impact investing” optimizes both financial and social outcomes, and thus promises to support hybrid organizations that simultaneously pursue financial and social goals. We argue, however, that impact investing decisions may be prone to behavioral factors that limit their outcome efficiency. In a portfolio allocation task designed to reflect the essential features of an impact investing decision, we find across a range of scenarios that individuals systematically fail to choose investment portfolios that achieve financial and social outcomes efficiently and thereby waste opportunities for value creation. We further show in online and in‐person experiments that outcome inefficiency is related to “categorical cognition”: suppression of categorical labels on investment options increases efficiency.
Click here to download the paper.

Manufacturing & Service Operations Management (2018)

A large proportion of the world’s population has no access to electricity and so relies on noxious kerosene for their lighting needs. Solar-based solutions require a large up-front investment and are often unaffordable in this market owing to consumers’ tight liquidity constraints. As an alternative, there are business models relying on rechargeable light bulbs that are sold at a subsidized price (which renders them affordable) and require regular micropayments for recharges (which eases liquidity constraints). These bulbs provide a cheaper and healthier light source than kerosene, yet their adoption is lower than expected, and some consumers continue to use kerosene. This paper explores the potential drivers of such preferences and proposes strategies to alter them.
Click here to download the paper

Management Science, 2018

Firms commonly undertake philanthropic campaigns as a means of attracting and retaining customers. Such campaigns often take the form of charity-linked promotions, whereby a firm donates a specific amount to a charitable cause when a customer takes up the promotion through a related purchase. We carried out three field experiments to study such promotions in the context of an online taxi-booking platform. Customers were randomly assigned to different treatment groups, which received either a charity-linked or a discount-based promotion from a range of monetary amounts. Take-up rates for charity-linked promotions were not only much smaller than for discount-based promotions but also less sensitive to the exact amount involved, consistent with a view that the decision to take up a charity-linked promotion was driven in part by a “warm glow” from mere association with giving. We also find a selection effect in promotion take-up: charity-linked promotions were disproportionately taken up by people who had already been more active customers. Although a promotion take-up does seem to represent new demand rather than mere substitution of a booking that would have occurred anyway, longitudinal data analysis reveals little evidence of a lasting treatment effect on long-term demand beyond the promotion period for either kind of promotion. Given the high cost relative to benefit for the promotional bookings themselves, this finding raises concerns regarding the prevalent practice of firms devoting significant funds for short-term promotions without rigorously examining their exact impact.
Click here to download the paper.
See also: INSEAD Knowledge Article.

Journal of Applied Psychology, 2016

Two field experiments examined if and how values affirmations can ameliorate stereotype threat-induced gender performance gaps in an international competitive business environment. Based on self-affirmation theory (Steele, 1988), we predicted that writing about personal values unrelated to the perceived threat would attenuate the gender performance gap. Study 1 found that an online assignment to write about one’s personal values (but not a similar writing assignment including organizational values) closed the gender gap in course grades by 89.0% among 423 Masters of Business Administration students (MBAs) at an international business school. Study 2 replicated this effect among 396 MBAs in a different cohort with random assignment and tested 3 related mediators (self-efficacy, self-doubt, and self-criticism). Personal values reflection (but not reflecting on values including those of the organization or writing about others’ values) reduced the gender gap by 66.5%, and there was a significant indirect effect through reduced self-doubt. These findings show that a brief personal values writing exercise can dramatically improve women’s performance in competitive environments where they are negatively stereotyped. The results also demonstrate that stereotype threat (Steele & Aronson, 1995) can occur within a largely non-American population with work experience and that affirming one’s core personal values (without organizational values) can ameliorate the threat.
Click here to download the paper.

Management Science, 2015

The feasibility of using field experiments to optimize marketing decisions remains relatively unstudied. We investigate category pricing decisions that require estimating a large matrix of cross-product demand elasticities and ask the following question: How many experiments are required as the number of products in the category grows? Our main result demonstrates that if the categories have a favorable structure, we can learn faster and reduce the number of experiments that are required: the number of experiments required may grow just logarithmically with the number of products. These findings potentially have important implications for the application of field experiments. Firms may be able to obtain meaningful estimates using a practically feasible number of experiments, even in categories with a large number of products. We also provide a relatively simple mechanism that firms can use to evaluate whether a category has a structure that makes it feasible to use field experiments to set prices. We illustrate how to accomplish this using either a sample of historical data or a pilot set of experiments. We also discuss how to evaluate whether field experiments can help optimize other marketing decisions, such as selecting which products to advertise or promote.
Click here to download the paper.

Journal of Consumer Research, 2011

Does asking people about their future behavior increase or decrease the likelihood that they will repeat their past behavior? In two laboratory and two field experiments, we find that behavior prediction strengthens behavior repetition, making people more likely to do what they normally do, when personal norms regarding engaging in a behavior are weak or not easily accessible. However, when personal norms are strong or made accessible at the time of the prediction request, behavior prediction weakens behavior repetition and increases the likelihood that people do what they think they should do—even if it’s not what they normally would do. These findings provide new tools for influencing behavior repetition, reconcile some seemingly contradictory past findings, and contribute to the debate regarding the relative importance of habits and intentions in guiding behavior.
Click here to download the paper.

Personality and Social Psychology Bulletin, 2007

Self-affirmation theory proposes that people can respond to threats to the self by affirming alternative sources of self-integrity, resulting in greater openness to self-threatening information. The present research examines this at a group level by investigating whether a group affirmation (affirming an important group value) increases acceptance of threatening group information among sports teams and fans. In Study 1, athletes exhibited a group-serving attributional bias, which was eliminated by the group affirmation. In Study 2, the most highly identified fans exhibited the most bias in terms of their attributions, and this bias was eliminated by the group affirmation. These studies suggest that groups can serve as resources from which people can draw in response to threatening group events.
Click here to download the paper.

Advanced projects with working papers


Does growth training help entrepreneurs to scale-up new ventures? The field experiment answering this question uses a sample of 181 startup founders from the population of Singapore-based entrepreneurs in 2017.
The treatment consisted of classroom sessions conducted in workshop and lecture formats that provided content in growth-catalyst tools comprising of effective business model design, building effective venture management teams and leveraging personal networks, that help in entrepreneurial resource mobilization. Also, participants received individualized business coaching addressing their venture’s issues and challenges in these domains.
Results show that entrepreneurs that received training in the three growth-catalyst tools achieved higher sales and employee growth for their ventures. In addition, entrepreneurs with higher educational attainment, higher prior work experience and higher growth goals benefited much more from the training intervention.
Click here to download the paper


24 April 2018
The authors examine the effectiveness in field settings of seven healthy eating nudges, classified according to whether they manipulate 1) cognition—via “descriptive nutritional labeling,” “evaluative nutritional labeling,” or “salience enhancements”—2) affect—via “hedonic or sensory cues” or “healthy eating prods” —or 3) behavior—via “convenience enhancements” or “plate and portion size changes.” Their multivariate three-level meta-analysis of 286 effect sizes, controlling for eating behavior, population, and study characteristics, yields a standardized mean difference (Cohen’s d) of .23. Effect sizes increase as the focus of the nudges shifts from cognition (d=.09, equivalent to -48 kcal/day) to affect (d=.24, -118 kcal) to behavior (d=.37, -199 kcal). Interventions are more effective at reducing unhealthy eating than increasing healthy eating or reducing total eating. Effect sizes are larger in the US than in other countries; in restaurants or cafeterias than in grocery stores; and in studies including a control group. Effect sizes are similar for food selection vs. consumption, for children vs. adults, and are independent of study duration. Compared to the typical nudge study, one implementing the best nudge scenario can expect a fourfold increase in effectiveness, with half due to switching from cognitive to behavioral nudges.
Click here to download the paper

1 November 2017
This paper investigates whether social identity considerations and norms may be driving occupational choices by women. We implement a randomized field experiment to analyse how the self-selection of women into the technology sector changes when we randomly vary the recruitment message to potential applicants to a 5-month software coding program offered only to low income women in Peru and Mexico. In addition to a control message with generic information, in a treatment message we correct misperceptions about expected returns for women and their ability to pursue a career in technology. This de-biasing message doubles the probability of applying (from 7% to 15%). We then analyse the stark differential self-selection patterns for the treatment and the control groups to infer the potential barriers that may explain occupational segregation. We find evidence that both expectations about monetary returns in the sector and a perceived “identity” cost (as reflected by an IAT test and survey measures) of a career in technology operate as barriers. We interpret our results in the light of a Roy model where women are endowed with returns to skill in the technology and services sector, and bear an identity cost of working in technology (à la Akerlof Kranton, 2000). Through a follow up experiment in Mexico DF we are able to rule out alternative explanations for our results and point to what dimensions of the initial treatment matter most. Our results suggest social identity can explain persistent occupational segregation in this setting and point towards policy interventions that may alleviate it.

Click here to download the paper.

Innovation Growth Lab: Featured trial of the month in September 2018

18 September 2017
Firms must often decide how to target households that did not respond to past promotions. For example, when prospecting for new customers, households that purchase are no longer eligible, and so the remaining households are an increasingly pure pool of non-responders. Past response data reflects the behaviour of households that responded, and these households generally differ from the rest in unobservable ways. We show that despite this, the decisions of the past responders in a group can help firms target the remaining non-responders. In particular, the timing of past responses can help to reveal whether a geographic region is (a) exhausted of future responders, or (b) the future responders just require additional exposures. We use this insight to develop three different timing measures. The measures are calculated and validated using a sequence of mailings in a large-scale field experiment. The measures perform well, even when some regions have only a handful of past responses to calibrate them. They confirm that the decisions of past responders can help firms target non-responders.
Click here to download the paper.


16 August 2017
Firms often want to target different customers with different actions, and the marketing literature contains many models designed to optimize targeting policies. While these models are sometimes validated using simulations or historical data, the gold-standard approach to comparing alternative targeting policies is to implement the alternative policies on randomly selected groups of customers and then compare the aggregate outcomes. We show that we can improve the efficiency of these comparisons by using the same experiments but changing the analysis. Instead of comparing the aggregate outcomes, it is more efficient to compare the outcomes within each customer segment, where the segments are defined using the candidate policies. Differencing within segment reduces variation introduced by between-segment differences, and we can then aggregate the within-segment differences across segments. Our key contribution is to show that the choice of the segments is important. By segmenting using the policies’ recommended actions, we know that for some segments the difference in outcomes is zero by construction. This can greatly improve efficiency by eliminating random variation in the comparison of these segments. We illustrate these benefits theoretically and empirically, using data from a recent field experiment. We also show how this segmentation approach extends to an alternative experimental design and identify limitations that may hinder implementation of the proposed approach.
Click here to download the paper.


5 August 2017
Recent laboratory findings suggest that highlighting the multi-sensory experience of eating leads people to eat less food while increasing eating enjoyment. But can pleasure-based interventions be an effective and business-friendly alternative to nutrition labeling in an actual restaurant setting? To find out, we gave 98 paying restaurant customers either a regular “all-you-can-order” lunch menu with simple food descriptions, the same menu with calorie and fat information, or a menu with more vivid multi-sensory food descriptions. We then measured how sensory and nutrition labeling influenced food intake, pleasure expectations, and eating pace, and the mediating effects of these factors on customer value (fair value perception). The sensory menu decreased food intake by 159 kcal (-16%) yet increased customer value by €2.79 because it increased pleasure expectations and slowed down eating. In contrast, the nutrition menu decreased customer value by €0.82 because it led to a 298 kcal (-31%) reduction in food intake uncompensated by changes in pleasure expectations and eating pace. Furthermore, the effects of menu interventions increased with party size but were independent of gender. Compared with nutrition menu labeling, a pleasure-based intervention can lead to a triple win for public health, the restaurant trade, and the joy of eating.
Click here to download the paper

26 January 2017
Many countries regulate the quality of food and drugs, yet it remains unclear whether markets can be relied upon to deliver high quality in the absence of regulation, notably where companies can advertise the superior quality of their products. The authors present evidence from two field experiments in China’s infant formula industry, which has seen a trust crisis after several safety scandals. They show that the disclosure of information about product quality have a non-positive or even significantly negative impact on consumers’ purchase decisions and self-reported trust in the industry, as information reminds consumers of past scandals and draws their attention to potential safety risks.
Click here to download the paper


3 October 2016
We investigate how firms can use the results of field experiments to optimize the targeting of promotions. We evaluate seven widely-used segmentation methods using a series of two large scale field experiments. The first field experiment is used to generate a common pool of training data for each of the seven methods. We then validate the seven optimized policies provided by each method together with uniform benchmark policies in a second field experiment. The findings reveal that model-driven methods (Lasso regression and Finite Mixture Models) performed the best. Some distance-driven methods also performed well (particularly k-Nearest Neighbors). However, the classification methods we tested performed relatively poorly. The precision of the data varied with the level of aggregation, and the model-driven methods made the best use of the information when it was more precise. The model-driven methods also performed best in parts of the parameter space that are well represented in the training data. The relative performance of the methods is robust to modest differences in the settings used to create the training and validation data. We explain the poor performance of the classification methods in our setting and describe how it could be improved.
Click here to download the paper.


1 October 2016
Free market advocates consider consumer choice unambiguously welfare-enhancing, but critics argue that availability of certain products can be detrimental for society. Contributing to this debate, the authors study the case of controversial skin whitening products sold widely in emerging markets. Although positioned as empowering female consumers by providing more choice, these have been scrutinized for perpetuating women’s disempowerment by reinforcing sociocultural biases.
To test these claims, the authors experimentally examine a possible relationship between women’s disempowerment and preference for skin whitening products in India, and find some evidence of a positive relationship. Participants primed temporarily to feel more disempowered show greater preference for the stronger (and medically risky) products, but not for the milder ones. Implications from our findings for corporate social responsibility and policy are discussed.
Click here to download the paper.


16 April 2014
I study the role of norms on tax compliance through a field experiment on property taxes in Peru. Randomly chosen subsets of residents in two municipalities in the Lima province were informed, through an official letter from the municipality, about the average rate of compliance, the average level of municipal enforcement, or both. A last group was only reminded of the payment deadline. The results of the experiment reveal a more complex response to information on norms than has previously been documented. They also show that simple nudges can have large and long-lasting effects. Analysis of the administrative data reveals that disclosing information on the level of compliance had a large positive impact on compliance (20% relative to the control group). The payment reminder also raised compliance by 10%, however, an effect that persisted even after the municipality initiated legal proceeding against delinquents. The enforcement treatment did not have a significant effect on compliance net of the reminder effect. The study design also included surveys, conducted both before and after the intervention, in which a subsample of taxpayers was asked about their beliefs concerning the levels of compliance and enforcement. Both the norms and the enforcement treatments raised beliefs about compliance as well as about enforcement. Interestingly, the reminder letter also raised beliefs about compliance. To assess quantitatively the impact of norms through different channels, I fit a model in which residents take into account expected monetary penalties from noncompliance, the disutility of tax evasion rises with the fraction of residents who comply, and individuals hold subjective beliefs about the probabilities of both detection and compliance. The estimated model shows that the norm intervention acts by changing beliefs about both compliance and enforcement. There is also a large residual effect that I interpret as a strengthening of the intrinsic motivation to comply.
Click here to download the paper.


Scholarship on business education to date has emphasized the benefits of using social media technology in the classroom; however, discussion of potential risks remains largely absent. We address this gap in the literature by suggesting that the use of social media could trigger concerns among students about their social status within the learning community of the classroom. We predict that raising the expectation of evaluation by the learning community and using a formal grading system reduces learning effectiveness, and that these effects are mediated by reduced cognitive absorption and reflection. We design a randomized classroom experiment in an undergraduate course to test our predictions, and we find support for most of our predictions. We close with the discussion of the theoretical contribution to the business education literature by advancing the understanding of status dynamics in the learning community of a classroom, practical implications for instructors to be thoughtful when using evaluations and social media technology in the learning community of a classroom, and limitations and directions for future research.

Early-stage projects


There is a growing interest for in-store interventions that can nudge people to purchase healthier foods, but a lack of data about their real-world effectiveness, especially over the medium term. The study aims to assess the ability of three point-of-sale interventions in grocery stores to improve the nutritional quality of food purchases.

We shed light on the micro-processes of entrepreneurial resource mobilization by investigating how entrepreneurs elicit referrals from their pre-existing network contacts to gain access to seed-round investors. We theorize network activation strategies by which entrepreneurs persuade their network contact to initiate referrals. Our conceptual model distinguishes between activation strategies that elicit contact’s cooperation by appeals to reciprocity versus offers of monetary incentive. We test our predictions in a field experiment at an Indian entrepreneurship accelerator and two within-subject vignette experiments with executives. Results across the three studies suggest that network-activation strategies that emphasize reciprocity rather than monetary incentive are more effective in securing referrals to seed-round investors, particularly when the entrepreneur has strong ties to the network contact.

A large part of the success of micro-finance models in emerging economies is attributed to organization of disadvantaged women into self-help groups (SHGs) that supposedly lead to increased demand for productive loans as well as timely repayments. While self-help group formation encourages regular meetings and financial savings amongst members, it is not clear whether it creates mechanisms that might facilitate micro-entrepreneurship among poor women.  For instance, do repeated interactions in self-help-groups lead to any social capital gains for low-income women? Further, does individual social capital lead to entrepreneurial activity in low income contexts? This study aims to address these questions using an experimental design at a micro-finance organization based in rural India.

Modern slavery - the compulsion of labour through the use of slavery-like tactics such as human trafficking and debt bondage (Crane, 2012) - is deeply affected by the policies and practices used in international supply chains. Voluntary reporting by corporations is a major element of efforts to fight modern slavery. Our field experiment tests for the existence of peer effects on modern slavery reporting, using as our setting the population of medium-to-large corporations in the U.K. In collaboration with a U.K.-based NGO, we are using a broad mailing of peer information as a basis for tracking peer effects on the incidence and quality of reporting behaviour. By illuminating the forces that shape corporate reporting on sensitive human rights issues, we expect our study to inform future policy design and strategies of civil society organizations.


Smart payment and mobile technologies have created interesting opportunities for Mobility as a Service (MaaS) to become a promising urban transportation solution for future generations. Such solutions are particularly important to dense urban areas such as Singapore. MaaS is typically presented as fixed-priced packages of transport options that are designed to suit various commuting needs, with a public transit component forming its core, which is sometimes complemented by other point-to-point transportations. For commuters, MaaS can provide the convenience of a flat-rate pass and potential savings through a bundling discount, as well as significant psychological benefits such as reduced decision conflict and pain of paying. From a social perspective, MaaS can encourage less driving and greater reliance on public transit. We propose to examine the acceptance of MaaS products and study commuters’ post-subscription behavioural changes, with a focus on whether MaaS subscription promotes shifts from private to public transit. We will use behavioural experiments to study the design and public acceptance of MaaS products and MaaS products’ impact on promoting public transit.

Patients use the internet to exchange information with other patients. The availability of such information may lead to a change in health care that is comparable to the change the telescope brought to astronomy – suddenly information is accessible that seemed unreachable before. In this research project, we the antecedents of patients’ behaviour. Specifically, we examine what drives patients’ tendency to visit patient communities, to share their information, to interact with other patients, and to adjust their behaviour based on information they have learnt about in the community.



If you have any questions related to the Lab, we are here to help you.


. Contact us

INSEAD Personalised Experience




Save & Manage



It is easy, simply log in:

Via Social

  • icons
  • icons
  • icons


Use your email address