Apurv Mehra bio photo

Apurv Mehra

me.transforming('developer','researcher')

IIITD Gmail Twitter LinkedIn Github

Background

There is a lot of excitement about how technology might support financial inclusion and build financial capability for communities who are currently largely financially excluded. This project is a continuation of an earlier project (covered by Auto-drivers, Work and Technology IRB), where we examined, amongst other things, auto-rickshaw drivers loan payment practices and the work of the cash collectors tasked with collecting money from them, as well as the TWU back office. During this ethnographic study, we identified a number of opportunities for intervention – from streamlining TWU’s back office processes, helping them make more informed loan allocations and providing the collectors and drivers with tools to build financial capability and better manage their loans through a combination of information, motivators, nudges and social competition.

Research Questions

  1. How to designing technology for improving loan management and improving financial capability in a resource constraint setting ? The technology should be able to convey complex financial information to population with limited technical and print literacy.
  2. Designing for augmentation rather than automation.
  3. Designing for agency that enhances users that enhances users capacity to act, rather than controlling and constraining practice.

Setting

We are undertaking this research with the support of Three Wheels United (TWU) – a social enterprise , that aims to make the lives of auto-rickshaw drivers better by enabling them to get loans from mainstream banks to buy their autos – by standing guarantor at mainstream banks. There are several stakeholders in this scenario such as drivers, collectors, community organizers, NGO’s and staff from TWU back office. We have developed interventions which have been deployed with TWU’s collectors and back-office staff.

Ethnographic Study

We conducted observations of the back office, conducted rounds of observations and interviews with drivers and in situ interviews with the collector(s) in each NGO. We accompanied two collectors from NGO1 (Zoharin and Haniya) on their collection rounds and in the back office; attended loan collection meetings with NGO3’s collector Ramesh and hung around NGO2’s office, to observe drivers making payments to their collector Priya. We observed community organizers on the streets and in community meetings. Observations were recorded through extensive field-notes, often by two researchers, including a fluent Kannada (Karnataka’s local language) speaker. As well as observations of the drivers encountered during collections, we conducted 33 semi-structured interviews with drivers. We asked questions on education; family; technology use; financial situation, and their experiences with the loan.
Our analysis took a broadly ethnomethodological perspective. Ethnomethodological ethnographies explicate the knowledgeable, artful ways in which participants organize their practice and reveal the ways in which technologies and other artefacts are used. The authors read through and discussed all the observations and interviews in various analytic sessions. They organized them into themes as interesting topics began to emerge. The findings outlined here were emergent, coming from the data. More details of the ethnographic study which led to design of the following interventions can be found in our paper - Working Digital Money into a Cash Economy.

Interventions

We have till date designed and deployed 2 major interventions. The first one was for streamlining back office workflow and the second one is an intermediated application used by collectors and community organizers.

Streamlining the Back Office Work Flow

Findings from an ethnographic study show that whilst providing a vital service, TWU managed its processes through an ad hoc mixture of paper and technology. Their processes enable the flexibility crucial for supporting a financially-vulnerable community, but unfortunately important information was often not to-hand, leading to sub-optimal decisions being made about loan allocation. We developed and deployed an integrated technical solution to support TWU’s work.

The figure give a before after picture and how the intervention helped streamline their workflow. At a high-level, our system integrated with the legacy applications wherever possible to avoid any loss or mismatch of data. The design process was bottom-up and iterative. To build the mobile app, information on payments and allocations needed to be digitized. We took this as an opportunity to help TWU with their manual time-consuming allocation process. We held multiple design sessions with TWU management and the loan officer, as system design progressed. Once in operation, we also added new features where appropriate. The system and the intervention is described in more details in our paper - Prayana: A Journey Towards Financial Inclusion.

Designing the Intermediated Application

In the next phase of our project we designed a novel financial management application, Prayana (‘Journey’ in Kannada). Whilst there are many financial management technologies available they are not typically designed for low income users in resource-constrained settings. Resources in our setting are constrained on many levels. Both TWU’s employees and the auto-drivers have a range of technical, print, and financial literacies. For example, half the drivers in our study had finished primary school or had no schooling, with those without schooling being unable to read. Most drivers use feature phones without data, and earn largely in cash. TWU operates on slim margins, with no in-house technical expertise, and consequently was using an ad-hoc mixture of paper and technology, built up over time, to manage their services. Whilst this enables the flexibility that is crucial for auto-rickshaw drivers to pay off their loans, it means that both drivers and collectors had limited information on loan status and progress. Furthermore, we found that loan payment is a collaborative achievement between collectors and drivers. This led us to design an intermediated smartphone application to be used by collectors with drivers, overcoming the practical barrier of drivers’ limited smartphone ownership, but with the additional advantage of supporting the collaborative work of loan payment.

Our initial motivation was to provide more resources for the drivers to help them better understand and manage their ongoing loan. From the literature, information, motivators and nudges seemed promising. We describe how we adapted these techniques for users with a range of print, technical, and financial literacies. Although not widely field tested yet, the design of Prayana is deeply grounded in the findings of an ethnographic study and an iterative design process. At each stage of the design and user testing, we found improved understanding and ease of use by participants. Furthermore, with our field deployment where TWU has transferred its entire collections (currently consisting of four collectors, eight community organizers and around 200 drivers) onto Prayana, gives a clear indication of perceived benefit. This, and the fact that these methods have proved successful elsewhere, gives us confidence in the validity of our design.

More details about the various design sessions and discussions around individual screen design can be found in our paper - Prayana: Intermediated Financial Management in Resource-Constrained Settings. We also have video walk-through of our application and training videos which can be found here.

Analysis

We are collecting a lot of qualitative and quantitative data regarding app usage from the field. We want to use them to understand how technology is being and the main take aways from the intervention.

Ongoing Work

  1. Drivers Application: With release of Jio[] there has been a rapid change in the smart phone and data use statistics of India. Our early explorations with a limited drivers shows uptake of smart phone usage along with data pack. A lot of drivers also want loan information available to them directly on their phones. We intend to provide this by designing driver specific application to push relevant information to them directly.
  2. Social Competition: Drivers are a close knit community. A lot of research has shown how peer pressure and social competition can help driver behavioral change[link]. We want to explore if we can make loan repayment a community exercise (it’s already done partly by the community organizers we want to augment their work further) and improve repayments by designing a social competition with incentives.
  3. Financial Literacy: We want to explore if we can encourage drivers and their family to improve financial literacy by providing them education content through drivers app and piggy backing on success of techniques like pay to learn.

Publications and Resources

  1. Working Digital Money into a Cash Economy: The Collaborative Work of Loan Payment. O’Neill, J., Dhareshwar, A. & Muralidhar, S.H. CSCW (2017) 26: 733. https://doi.org/10.1007/s10606-017-9289-6. link
  2. Apurv Mehra, Sambhav Satija, and Jacki O’ Neill. Prayana: A Journey Towards Financial Inclusion. Proceedings of the Ninth International Conference on Information and Communication Technologies and Development. ACM, 2017. pdf
  3. Apurv Mehra, Srihari Muralidhar, Sambhav Satija, Anupama Dhareshwar and Jacki O’ Neill. Prayana: Intermediated Financial Management in Resource-Constrained Settings. Accepted in 2018 CHI Conference on Human Factors in Computing Systems. ACM, 2018.