Consent is a critical part of communication between users and providers of Digital Financial Services (DFS). However, the consent and notice model for DFS has many deficiencies, especially for constrained users, i.e., users with limited literacy and first-time smartphone users. This project sets out specific recommendations on how providers such as Account Aggregators can make consent customer friendly. These recommendations emerge from a qualitative primary study which examines why customers currently do not engage meaningfully with consent artefacts and how a behaviourally informed design could improve understanding and engagement

We find that among low-income borrowers, the borrowing journey is shaped by financial stress, time pressure and fear of being denied formal credit, which in turn influences consent decisions. This macro-context shapes individuals’ micro-decision of consent. Furthermore, the shrink-wrap design of consent artefacts offers little scope for negotiation or challenge, affording customers little real choice, leading customers to passively comply.

These findings have been translated into actionable recommendations for redesigning consent artefacts to ease the customer’s loan application journey. They aim to make the journey easier and empower customers by acquainting them with the concepts of consent, personal data, explaining to them the role of their persona data in the lenders' decision, reducing the barriers to engagement by using local languages, visual cues live chats and finally offering an accessible redress and enquiry mechanism. The recommendations are available here, the findings from the behavioural study are available here, and the first principles research on the need for informed consent in financial services is available here.

Social engineering ploys, where customers are manipulated into authorising fraudulent transactions, are a serious customer protection concern. Regulators and financial institutions have responded with awareness campaigns, like TV commercials (TVCs), to raise awareness and reduce customers’ tendency to engage with fraudsters. This project deploys an outcome-based survey (OBS) BS to evaluate the effectiveness of UPI-fraud-awareness campaigns in reducing individuals’ propensity to engage with fraudulent communication.

Leaning on behavioural science and market research literature, this OBS measures campaign effectiveness along 4 dimensions: recall, appeal, comprehension, and impact. To recreate the hot state that individuals encounter when faced with fraudulent messages, the OBS simulates common frauds in a lab-in-the-field setting and supplemented with focused group discussions. As proof of concept, this OBS was piloted with 80 low-income and new-to-UPI users from four Tier I and Tier II cities to test the effectiveness of three TVCs.

We discovered that TVCs with relatable characters and simple messages in a storytelling format fare better on recall, appeal, and comprehension. However, awareness campaigns primarily influence cognition and are limited in their ability to shift attitudes or behaviour, which are shaped by habits, emotions and vulnerability to biases. This is the reason that fraudsters continue to use ‘hot state’ scenarios to make customers feel panicked or elated so that they do not think clearly. These findings reiterate the importance of designing awareness campaigns that remind customers that nothing is ever as urgent as some unknown caller seems to claim.

This project studies how borrowers place trust in digital lenders in the rapidly expanding digital credit ecosystem in India. In 2021, India had nearly 1,100 unique loan apps across various platforms. However, more than half of these were found to be illegal - they operated without proper licensing, promised quick loans without paperwork at questionable terms such as exorbitant interest rates, hidden fees, and used unethical, coercive recovery methods. These practices are in stark contrast with the customer protection guidance prescribed by the regulator to regulated lenders.

The project enquires into two questions: what are the expectations that prospective borrowers have of trustworthy lenders and, how borrowers gauge the trustworthiness of a lender. We design a primary study to understand how customers trust and contract with a digital lender. Our study of trust relies on exploring the proximate grounds that people rely on to trust someone or something. We commenced a mixed-methods study of digital financial services’ users in early 2024. The study is supported by qualitative interviews spanning a range of topics from what trust meant in personal relations and financial transactions to the obligations that come with being considered trustworthy. This is complemented by an experiment where respondents assess the user interfaces of digital lending apps for trustworthiness. 

Based on our discussions with users, we develop a theory to explain the making of trust in digital lending. A short video summarising the findings is available here; the foundational research on the psychological and affective dimensions of trust is available here and here.

This project developed a survey-based monitoring tool to enable self-regulatory organisations (SROs) to monitor the activities of microfinance institutions (MFIs), with a particular focus on risks to customers. The project was initiated in the context of significant regulatory changes that resulted in major revisions to how MFIs underwrite. The impact of these regulations on borrowers, therefore, remained to be understood. A systematic tool was thus needed to capture their impact, both as a point-in-time exercise and through continuous assessment. The tool was piloted with six members of Sa-Dhan, one of the two SROs for microfinance institutions. Given the efficacy of the tool, Sa-Dhan subsequently scaled it to six additional members.

The project developed a market monitoring tool for Unified Payments Interface (UPI) providers that leverages publicly available social media and application review data to surface, in near real time, the hurdles customers face. While UPI has rapidly become India’s dominant payment infrastructure, traditional complaint channels capture only a small fraction of user experiences, even though users frequently encounter onboarding challenges, failed transactions, and delays in grievance resolution. 

To address this visibility gap, data was collected from Twitter and the Google Play Store for major UPI providers that together account for over 90 percent of transaction volumes. The data were modelled using natural language processing methods to identify an exhaustive set of customer hurdles, calculate their prevalence rates, and generate indicators of provider responsiveness and temporal trends. The tool, Dvara PARSE, is currently running, and it captures grievances daily and categorises them into themes like onboarding hurdles, inadequate grievance redress, unauthorised charges, etc. Further, the tool identifies overlaps across these grievance categories and quantifies patterns of customer sentiment and periodic traffic related to providers. The overall accuracy of the tool stands at 94 percent. 

By automating the identification and measurement of customer-facing issues, the monitoring tool yields actionable insights for providers and regulators, enhancing the detection of emerging problems in the UPI ecosystem.

The project developed a machine–learning–based debt distress detection tool for microfinance borrowers that makes distress visible and measurable using administrative data already held by lenders. In current practice, lenders observe only repayment outcomes (delinquencies), even though distress often sets in much earlier, with borrowers adopting different repayment strategies and coping mechanisms, such as cutting essential consumption, withdrawing children from school, or foregoing medical care, to continue repaying. 

To address this blind spot, we trained and field-tested the model on data from over 250,000 borrowers in partnership with the Robert Bosch Centre for Data Science and AI at IIT Madras and a leading NBFC. We were able to analyse loan characteristics, repayment behaviour, and account-level patterns. The model identifies latent distress and the future probability of delinquency, enabling lenders to detect potential borrower harm early. Field validation showed delinquency prediction accuracy exceeding 95 percent. Distress prediction accuracy of approximately 75 percent also represents a significant advance over prevailing practices where distress remains entirely unmeasured. 

As a next step, the detection tool will feed into a broader set of Debt Distress Protocols aimed at testing and institutionalising post-detection interventions, shifting the system from passive observation of harm to structured, evidence-based response strategies.

The project undertook a systematic assessment of the monetary and non-monetary costs faced by Indian consumers using Buy Now, Pay Later (BNPL) products, a category of short-term digital credit that had grown rapidly without commensurate regulatory oversight. 

The research methodology involved the usage of several BNPL products to reviews their terms and conditions, analyse product features, examine fee structures and implicit interest rates. Legal experts were also consulted to gauge applications’ compliance with existing consumer protection and credit regulations. This work culminated in the analytical report, “The Costs of Using Buy Now, Pay Later (BNPL) Products”. The report documented the true costs associated with BNPL credit, identified gaps in transparency and disclosure, highlighted risks arising from the then business models and weak grievance redress systems. It also offered recommendations for reform, including treating BNPL as regulated credit, mandating clear cost disclosures through Key Fact Statements, strengthening credit reporting, and establishing stronger consumer protection protocols. 

Following dissemination, these findings were subsequently reflected in regulatory developments led by the Reserve Bank of India. After the publication of the report, the RBI’s Digital Lending Guidelines (2022, refined through FAQs and updated Directions in 2025) incorporated several of the report’s recommendations: BNPL was formally brought within the ambit of regulated credit when extended through banks and NBFCs; lenders were required to provide transparent, standardised disclosure of all fees; BNPL credit flows were mandated to be reported to credit bureaus; grievance redress and collections were strengthened; and the use of unregulated intermediaries for credit disbursal and repayment was curtailed.

The project was initiated with the objective of developing an integrated GRM for digital lenders, in partnership with an industry association. It was premised on the fact that digital lenders often serve as intermediaries and, therefore, many grievances are not necessarily directly addressable by them. Further, due to the varied scale of operations across digital lenders, maintaining an omnichannel GRM is often difficult for smaller entities. Accordingly, the project set out to create a blueprint for a responsive omnichannel GRM that could be housed at the industry association, with necessary integration with other implicated actors, primarily banks and NBFCs. The outputs from the project, however, did not materialise, as the partnership fell apart due to developments in the industry.

Information asymmetry is a situation where one party to a transaction has more or better information than the other. In financial markets, it is often the case that the provider has more information than the poor customer about a product but does not disclose this information accurately in order to make an unsuitable sale. Life insurance is a good use-case to study in this regard. If poor customers buy life insurance at all, they usually buy endowment plans, but these are not typically suitable for them. 

This project proposed to solve this problem by focusing on the disclosure aspect. Accordingly, we employed behavioural science to first understand the drivers of life insurance purchases among low-income households, and then experimental methods to test the effectiveness of diverse disclosure formats in influencing the purchase decision. We discovered that disclosures that merely provide accurate information about endowment plans are not enough to change habits. Rather, a superior alternative option should also be offered, and accurate information about its advantages over endowment plans needs to be clearly presented. Then switching behaviour follows. 

Our recommendations based on this study call for simple, easy-to-understand disclosures that aid product comparison and comprehension. Specifically, our recommendations called for the introduction of a simplified Customer Information Sheet (CIS) with features explained in simple language, provision of an illustration on customized benefits and compulsory signing of both. The project has gained broader relevance, as the IRDAI has now implemented all of these recommendations and provided an illustrative CIS for insurance companies.

Women nano entrepreneurs lead complex financial lives, managing both business finances and handling money matters at home. The complexity arises from the fact that financial decision-making is largely intuitive, undocumented, and frequent. Due to the informal nature of their enterprises, they typically lack access to tools that support effective money management. In a world where women from low-income families are inundated with one-size-fits-all financial products, we ask: What if women nano entrepreneurs could manage their daily money lives with the same ease as they send a WhatsApp message? Could a digital money management tool make this possible? And who among the entrepreneurs would find such a tool most relevant and why? 

We are exploring innovations in this space, in collaboration with MeraBills, a Technology Service Provider that offers book-keeping services and generates business intelligence for its users, typically small informal businesses. We believe that digital tools, such as the ones offered by MeraBills, can meaningfully advance women's financial inclusion by helping them save better, access suitable credit, and gain confidence in using digital financial services. Tools that women can adapt to their own needs and context can prove to be a useful way for women to manage their financial responsibilities, both at home and work.

To improve access to healthcare, most states in India offer health insurance schemes with fully or partially discounted premiums, primarily targeting low-income households. One such scheme is the Mukhyamantri Ayushman Arogya Yojana (MAAY), launched by the government of Rajasthan in 2021 which aimed to provide cashless, affordable healthcare to all families in the state. Yet early evidence shows that many people still struggle to benefit from the scheme. The project examines why these exclusions persist, despite MAAY’s promise of universal access. 

Using a citizen-centric framework, we examine exclusion at five critical stages: becoming aware of the scheme, enrolling successfully, using benefits at hospitals, raising grievances, and renewing membership. Breakdowns at any of these touchpoints, such as lack of information, documentation mismatches, denial of care, technical issues, or inaccessible grievance channels can leave families without timely treatment. These obstacles prevent citizens from obtaining timely treatment and undermine the scheme’s core objective of protecting households from health-related financial shocks. 

Building on our earlier work on exclusion in social protection, this project explores how broader economic, social and political exclusions shape people’s ability to access health insurance. Economic constraints, weak social networks, and limited political voice can interact to heighten vulnerability and reduce a citizen’s ability to navigate the system. By documenting these mechanisms of exclusion, the project aims to inform policymakers and support the design of more inclusive health insurance systems that ensure every individual can obtain care when they need it without friction.

The concept of financial health is rapidly emerging as a crucial metric for financial inclusion. Over the last few years, there has emerged a call to look beyond the straightforward and easily measured access and usage metrics of financial products to gauge their impact on users’ lives as measured by their financial health or resilience. To measure the financial health of low-income households, this survey tool applies an input–output–outcome framework which can identify gaps that remain unaddressed by existing financial products. 

In partnership with PricewaterhouseCoopers, we have deployed the tool across a sample of 4,000 low-income households in seven Indian states. The 20-minute survey captures ownership of a broad suite of financial products (inputs), along with the frequency and depth of their usage (outputs). It also provides a product-agnostic measure of household financial health (outcomes), defined here as the ability of low-income households to manage cashflows effectively and withstand contingencies. This framework enables the identification of gaps in both access and usage across financial product categories, differences in uptake and engagement across demographic and livelihood cohorts, and aspects of financial resilience that remain underserved by formal financial services.

The survey’s insights can be applied by multiple stakeholders: (i) financial service providers can use it for market reconnaissance and opportunity discovery, (ii) regulators can use it to monitor the effectiveness and impact of financial inclusion efforts, and (iii) implementation agencies can use it as a targeted needs-assessment tool.

The project was conceived with the recognition that women from low-income households are offered meaningful opportunities to add liabilities to their portfolios through well-serviced microloans, while comparable asset-building products, such as savings instruments, are not similarly designed or delivered. 

In the first stage of the project, a conceptual understanding of the various savings products tested on this cohort was built. Fieldwork was then undertaken to capture women’s perspectives on savings and day-to-day money management. Suggestions and feedback were subsequently sought from a range of stakeholders, including Small Finance Banks, Banking Correspondents, and Chit Fund companies, so that supply-side constraints and motivations could be mapped. In the second stage, insights from these streams of work were synthesised, and core challenges affecting savings behaviour among women from low-income households were identified. Based on these insights, a novel savings product was conceptualised. 

This product was envisaged not only as a mechanism for accumulating savings but also as a practical tool for managing household cash flows in ways that align with women’s financial realities.

This project was conceived with the goal of revealing customer protection issues that may be going unreported and to collect better insights into underlying problems faced by customers in the Banking, Financial Services and Insurance sector. Seven grassroot level organizations were recruited to surface and harvest over 100 lived experiences of constrained Digital Financial Services’ customers. The customer stories capture the customer’s experience at each touchpoint of their customer journey - from onboarding to the point of grievance or exit - and document the process as well as their personal reflections and emotions. 

Our findings from this project revealed typical low-income, constrained customers’ pain points. It also shed light on the respondents’ inhibitions at each stage of their customer journey, their constraints that go beyond literacy and digital familiarity, their expectations from the service provider, which were markedly different from the parameters which are checked for in customer satisfaction surveys. 

The aggregate data from this exercise as well as the literature on service and its associated constructs point to gaps in service quality measurement, that warrant further investigation, specifically, the lack of consensus on what constitutes minimal service standards or adequate service quality. To conclude, we have developed a framework and tool that is meant to complement the existing set of service quality tools and metrics by assessing service quality adequacy from a constrained customer’s point of view.