Social innovations and the fight against poverty: An analysis of India's first prosocial P2P lending platform

Prosocial P2P lending platforms are a novel and powerful example of a digital social innovation (DSI) in which the operating model relies primarily on digital technologies and the overarching focus is on the “social” aspect of the innovation. These platforms establish a virtual connection between low‐income individuals and lenders, helping the former access loans at low rates of interest. In realising their mission of fighting poverty, prosocial P2P lending platforms maintain a challenging hybrid – online and offline – focus. This paper explores how prosocial P2P lending platforms enact their hybrid orientation. It draws on an inductive qualitative study of Rang De, India's first prosocial P2P lending platform. The analysis highlights five clusters of actions: digital attention‐building, digital credibility‐building, digital empathy‐building, intermediary relationship‐building, and borrower relationship‐building. The paper argues that significant strengths on the online side help establish a sustainable business model. A willingness and commitment to maintain a high degree of engagement with the complex offline world of low‐income borrowers helps develop the model as an impactful social innovation.

Most P2P lending platforms are pure commercial ventures and do not have an explicit focus on grander social challenges. In contrast to this distinct commercial orientation, a small number of prosocial P2P lending platforms have emerged. These platforms aim to use the Internet as a catalyst for achieving the United Nations' grand sustainable development goal (SDG 1) of ending poverty (Carr, Dickinson, McKinnon, & Chávez, 2016;McKinnon, Dickinson, Carr, & Chávez, 2013). The lending platform Kiva is probably the most famous example in this context (see Berns, Figueroa-Armijos, da Motta Veiga, & Dunne, 2020;Burtch, Ghose, & Wattal, 2014;Johansen & Nielsen, 2016). Prosocial P2P lending platforms identify prospective borrowers from low-income households, build an online profile for them and make it possible to lend them money through a digital transaction on the Internet (Carr et al., 2016). The focus here is on helping the poor access relatively small loans by matching them with individuals willing to lend (Coates & Saloner, 2009). The grand idea is to tackle the wicked intractable problem of poverty through innovative use of digital technologies (cf. Reinecke & Ansari, 2016;Sandeep & Ravishankar, 2018;Schwittay, 2019).
Prosocial P2P lending platforms are thus an exemplary instantiation of a digital social innovation (DSI) in which the operating model relies primarily on digital technologies and the overarching focus is on the "social" aspect of the innovation (Qureshi, Pan, & Zheng, 2019). Given the relative newness of prosocial P2P lending platforms, scarce research exists on how they expand the reach of technological solutions and develop innovative services to tackle poverty. The small body of empirical work that does feature prosocial P2P lending platforms only employ them as a setting to undertake broader analyses, such as the impact of culture and geography on lender preferences (eg, Burtch et al., 2014) and the management of lenders' information asymmetry (eg, Riggins & Weber, 2017). This paper is one of the first in-depth exploratory studies showcasing prosocial P2P platforms as an important emerging DSI phenomenon. Specifically, it responds to this special issue's encouragement to report on novel instances of DSI.

| Hybrid orientation of prosocial P2P lending platforms
In realising their mission of fighting poverty, prosocial P2P lending platforms maintain a challenging hybridonline and offlinefocus. This hybrid orientation is fundamentally different from typical commercial ventures that also maintain both online and offline retail operations to maximise financial outcomes (cf., Huang, Pan, & Liu, 2017). For prosocial P2P lending platforms hybridity is a core facet of their social mission: to harness the immense reach of the Internet and to fight against poverty (Coates & Saloner, 2009). The main purpose of the online channel is the generation and maintenance of a steady supply of lenders (ie, it is the supply side of operations). On the other hand, the main purpose of the offline activity is to engage intensively with low-income borrowers (ie, it is the demand side of operations) and generate social impact. On the supply side, a strong online presence and digital capabilities are required to attract prospective lenders and retain their commitment. The key challenge here is to generate awareness of the prosocial platform's existence and make it stand out amongst other P2P lenders claiming to offer more lucrative investment opportunities. A purposeful set of actions are also needed to establish trust in the social orientation of the platform (cf. Chen, Lou, & Van Slyke, 2015;Xu & Chau, 2018) so that lenders are confident their investment will be properly managed and result in positive impact for a low-income recipient. Simultaneously, on the demand side, the focus is on offline rather than online activity. Significant offline work is required to reach out to borrowers from poor backgrounds, given that in many cases they (ie, low-income groups and financially excluded communities) have neither affordable access to high-speed Internet nor are digitally literate enough to engage with online platforms. Thus, the only way to stimulate demand and generate social impact is by maintaining close contact with borrowers. Platforms need to reach out to low-income communities, assess their financial needs, and mediate access to prosocial lenders (Coates & Saloner, 2009). The success of the "social" innovation dimension of prosocial P2P lending platforms is therefore squarely linked to their engagement and work with digitally excluded low-income borrowers. In this sense, the hybrid online-offline focus of prosocial P2P lenders stems from a compelling need to build contrasting modes of engagement and operations with key stakeholders (cf. Sandeep & Ravishankar, 2016). This paper reports on an in-depth qualitative case study of Rang De, India's first prosocial P2P lending platform, and addresses the question: how do prosocial P2P lending platforms enact their hybrid orientation? Drawing on an inductive analysis, the findings illustrate five clusters of actions underpinning the enactment of hybridity in prosocial P2P lending platforms. Three clustersdigital attention-building, digital credibility-building, and digital empathybuildingrelate to the platform's online focus and two clustersintermediary relationship-building and borrower relationship-buildingrelate to its offline focus. The rest of the paper is organised as follows. The next section discusses the notion of hybridity and its applicability to prosocial P2P lending platforms. Next, the research methods and the Rang De case are presented and analysed. The final section is the discussion and conclusions.

| HYBRIDITY
Hybridity broadly refers to a composition of disparate elements. 1 In the business and management literature "hybridity" speaks to tensions between multiple, competing forces that are manifest in some form of combination. Hybridity is typically rooted in the demands of non-trivial internal and external audiences (Haveman & Rao, 2006) such as powerful leaders, regulators, shareholders, and civil society groups. There are several classic illustrations of hybridity in the literature such as exploitation-exploration (Benner & Tushman, 2015), network-hierarchical forms of organising (Adler, 2001), and subsidiary-headquarters cultures (Schein, 2009). Hybrid actions are visible in an organisation's internal ordering of activities and can play a crucial role in the development of innovative business models. A good illustration is the structural separation of business units into those that deliver specialised IT application development and those that support routine business operations in IT services firms (Ravishankar & Pan, 2008). Another example is the synchronisation and combination of online retail stores and offline retail stores (Huang et al., 2017). Hybridity is a conundrum because it calls for the implementation of strategies comprising of conflicting dimensions (Smith & Besharov, 2019). In other words, despite its exciting possibilities the notion of hybridity implies some form of contrasting resource allocations, investments, and activities that may be difficult to sustain (Battilana & Lee, 2014). Some organisations experience hybridity as an inevitable condition requiring a calculated response from key actors. For others, a hybrid orientation could be a strategic and deliberate choice. An example of the former is the market pressure to maintain both cost-leadership and product differentiation, which demands creative hybridization approaches from managers (Heracleous & Witz, 2010). An example of the latter is social enterprises whose upfront mission encapsulates a hybrid -charity and businessfocus (Ebrahim, Battilana, & Mair, 2014;Sandeep & Ravishankar, 2016;Shepherd, Williams, & Zhao, 2019;Smith, Gonin, & Besharov, 2013). Battilana and Lee (2014) drew on insights from three strands of literature and conceptualised hybridity as combinations of (a) organisational identities, (b) organisational forms, and (c) institutional logics. In recent years, much of the conceptual development around the notion of hybridity has drawn on social enterprises, which confront a vivid hybridity challengeto effectively combine their business-orientation and their charity-orientation ( Battilana and Lee's (2014) review speaks largely to this body of work and focuses on hybridization that combines business-and charity-oriented actions. 2 In terms of the three strands of literature and combinations described above, social enterprises are hybrids of (a) a strong business identity and a charity identity; (b) a for-profit organisational form and a not-for-profit organisational form; and (c) a commerce logic and a charity logic. Prosocial P2P lending platforms, with their focus on sustaining a profitable business venture with a "social" core, can be viewed as typical social enterprises from all three perspectives (ie, identity, form, and logic).
A different way to think about the notion of hybridity in social enterprises would be to avoid discrete categorizations and view it, instead, as a continuum anchored by strong social and economic objectives at opposite ends (Shepherd et al., 2019). As the pendulum swings excessively towards one of the ends, it could trigger a goal-balancing problem and a concomitant struggle to reconcile inherently contrasting objectives. In other words, there exists the risk of mission drift -a gradual drift towards one of their identities or forms or logics at the cost of others (Mersland & Strøm, 2010). For example, in balancing the twin objectives of serving the poor and maintaining profit margins, microfinance providers have been known to prioritise financial objectives by charging high-interest rates to the poor, thereby drifting from their original social mission (Armendariz, D'Espallier, Hudon, & Szafarz, 2012;Battilana & Dorado, 2010;Hudon & Sandberg, 2013).
The literature on hybridity suggests different approaches to its management (Battilana & Lee, 2014), which connect to the broader notion of hybridity and not just to its manifestation in social enterprises. The first approach involves integrating contrasting aspects such as multiple conflicting identities (Pratt & Foreman, 2000). In the case of microfinance providers, for example, such an approach is evident in how activities such as credit scoring and loan collection are intrinsic to the business model, and thus integrated into the very social mission of the company (Battilana & Lee, 2014, p. 413). Similarly, MNCs are known to integrate the disparate identities of collaborating global units through manager intermediaries, who work to create a strong unitary identity (Eisenberg & Mattarelli, 2017). A second approach is to compartmentalise the contrasting elements and keep them separate, rather than seek synergies between them (Pratt & Foreman, 2000;Sandberg & Werr, 2003). For instance, impact sourcing companies are known to keep their social orientation (ie, a provider of employment to marginalised individuals) separate from their business orientation (ie, a provider of IT outsourcing services), presenting mainly the former to local communities and the latter to clients (Sandeep & Ravishankar, 2016). Similarly, organisations develop distinct business units for exploratory activities and keep them separate from those focused on exploitative activities (Benner & Tushman, 2015;McDonough III & Leifer, 1983). Compartmentalization could also help present different images to different audiences and manage impressions (Sandeep & Ravishankar, 2018). A third approach is what may be termed as the "creative" approach (Battilana & Lee, 2014) featuring active creation of innovations that transcend institutional limits and known blending strategies to define new paradigms. For example, social accounting can be viewed as a novel tool that presents a new way of evaluating the contribution of business to society (Ramus & Vaccaro, 2017).

| Prosocial P2P platforms and hybridity
The hybrid (charity and commercial) orientation of prosocial P2P lending platforms are evident in their mission to address the problem of poverty through internet-led business venturing (Coates & Saloner, 2009). The founders' prosocial motivations create the broad and overarching challenge of generating social impact whilst keeping the business afloat (cf. Sandeep & Ravishankar, 2015). The charity-commercial hybrid orientation can therefore be conceptualised as a meta form of hybridity that defines prosocial P2P platforms. Indeed, this meta hybrid orientation underpins many social enterprises and it has been the focus of several studies (eg, Battilana & Dorado, 2010;Ebrahim et al., 2014;Smith & Besharov, 2019). However, in practice, for prosocial P2P lending platforms the most problematic aspect of their meta charity-commercial hybridity is not so much the fact that they possess it by design.
Rather, it is the meta hybridity's direct effect and tense manifestation in real terms: the everyday enactment of their hybrid online-offline focus, which speaks to their quest to deliver both economic and social value.
Their online focus is apparent in the self-presentation of their identity as a digital platform that interfaces with low-income borrowers. The digital form is central to interactions with lenders, who mostly see the prosocial P2P lending platforms' online manifestation (Carr et al., 2016). The digital image is thus crucial to the firm's relationships on the supply side (ie, with current and prospective lenders) and it enables lenders to "connect" with borrowers (McKinnon et al., 2013). In other words, the online focus helps recruit and retain lenders and is key to the financial viability of the business. The digital core implies that prosocial P2P lending platforms need to make particular types of investments in order to create and maintain advanced digital capabilities. For instance, they may need to hire technology professionals, ensure they stay motivated, and pay them salaries comparable with the market. Overall, P2P lending platforms' economic value creation hinges on their ability to maintain a strong online focus. In sharp contrast to the online focus, they also need to maintain an offline focus in order to reach out to low-income borrowers and develop a trust-based relationship with them. The offline focus requires prosocial P2P lender platforms to almost morph into locally embedded microfinance providers and engage closely with marginalised and vulnerable borrower communities who typically have limited or no access to the internet. The offline focus requires a completely different skillsetit requires the expertise and the contextual intelligence to recruit and support low-income individuals; it needs local knowledge to assess problems of loan recipients and address them sensitively and with compassion. In order to participate in, and contribute to, the world of low-income borrowers P2P lending platforms need to make investments that are different from those needed to sustain their digital capabilities. For instance, they may need to hire professionals in the third sector work and motivate them and pay them salaries comparable with the market. Overall, P2P lending platforms' social value creation hinges on their ability to maintain a strong offline focus. The next sections of the paper develop these arguments through an empirical investigation of a Rang De, a leading Indian prosocial P2P lending platform. It addresses the question: how do prosocial P2P lending platforms enact their hybrid orientation?

| RESEARCH METHODS
This research focused on Rang De (www.rangde.in), India's first prosocial P2P lending platform. Rang De facilitates the delivery of unsecured loans to low-income individuals and owners of micro-businesses through its online platform. I adopted an interpretive case study method (Walsham, 2006). An interpretive approach is particularly wellsuited to better understand the lived reality of actors in their particular settings (Ravishankar, 2013). I did not adopt a pre-conceived framework to analyse the data. But a careful initial reading of Rang De's work and the inductive data analysis process led to the presentation of "hybridity" as a key theoretical idea of interest in the paper.

| Research setting
Rang De was co-founded in 2008 by Smita, who had a professional background in social work and Ramakrishna, who had worked until then in the information technology (IT) industry. The Rang De platform provides a virtual connection between lenders and low-income borrowers, with a particular focus on rural India. The founders started Rang De on a shoestring budget of GBP 6000. They convinced a group of IT engineers who had previously worked with the IT services company Infosys to create the Rang De platform. The ex-Infosys engineers admired and shared the prosocial motivations of Rang De and developed the platform. They were so impressed with the founders' vision for Rang De and their mission of fighting poverty that they not only charged a relatively small fee for building the platform, but also agreed to receive this fee in small instalments over an extended period. In addition, India's ICICI Group's Foundation for Inclusive Growth agreed to pay Rang De's initial operational costs. Until 2017 Rang De operated as a public charitable trust (ie, a not-for-profit organisation; www.rangde.org). Following major regulatory changes in 2017 introduced by the Reserve Bank of India, which classified P2P platforms as providers of financial services and stopped viewing them as technological platforms that facilitated financial activity (see Minupuri, 2019), Rang De changed form and re-started operations as a P2P Non-Banking Finance Company. In other words, it is now registered as a private limited company (ie, a for-profit organisation; www.rangde.in). The Rang De platform features a set of borrowers' profiles, each of which presents the borrower's demographic data, the purpose of the loan, and the type of activity for which the loan is requested. Lenders register on the website www.rangde.in, choose borrowers and transfer the money, which is routed through Rang De. The loans are unsecured and interest rates range from 4.5% for education loans to 10% for micro-businesses. Rang De's loans are offered at significantly lower interest rates than the available alternatives in rural areas, which make them affordable to low-income borrowers (Thomas, 2018). Rang De's 15 000 lenders have given away loans worth around INR 77 crore (approximately GBP 8 million) to more than 65 000 borrowers.
In identifying and managing the relationship with borrowers, Rang De works with several intermediaries (typically non-governmental organisations referred to as "impact partners" by Rang De). Borrowers repay monthly and Rang De pays its lenders an interest rate of 2%-5% plus the original amount invested on each loan. In this respect, Rang De is different from the well-known prosocial P2P lending platform Kiva, whose lenders receive no interest on their investment. Rang De typically retains 2% of the interest rate paid by borrowers to cover its operational costs and pays the intermediaries a fee for each loan arrangement. Operating from its Bangalore office, Rang De has so far worked with 39 intermediaries in 18 Indian states. Intermediaries identify potential borrowers and facilitate their application for a Rang De loan. Prospective borrowers apply either in their individual capacity or as part of collective self-help groups (SHGs). Intermediaries also collect and manage borrowers' loan repayments. The repayment rate is remarkably high (94.2%). Borrowers, intermediaries, and lenders are at the heart of the Rang De model, which has gained significant recognition as an exemplary case of a prosocial P2P lending platform.

| Data collection
We began our data collection by meeting the Rang De founders in November 2016. We followed this meeting with four rounds of fieldwork. We conducted interviews through a combination of snowball and purposeful sampling techniques with informants at Rang De, "impact partner" intermediaries, lenders, and borrowers. In total, we conducted 48 in-depth interviews (Table 1). Throughout the fieldwork, we combined semi-structured interviews (see Appendix A) with numerous conversations and informal gatherings with informants at the Rang De office, which offered further insights and helps us triangulate the interview data. We visited two "impact partner" intermediaries, during which we also met borrowers linked to a waste pickers organisation in Bangalore and a social work organisation in the Mysore district of Karnataka. In the early stages of the work, we also interviewed four professionals with T A B L E 1 Summary of interviews  4 | ANALYSIS

| Digital work and fieldwork
Rang De employees engage in two types of activities, one corresponding to their online focus, which can be termed digital work and the other corresponding to their offline work, which can be termed fieldwork. We inductively identified three clusters of actions comprising digital work: digital attention-building, digital credibility-building and digital empathy-building, and two clusters of actions comprising fieldwork: intermediary relationship-building and borrower relationship-building.

| Digital attention-building
Commercially oriented platforms dominate the P2P lending field. This spurt in online investment opportunities poses a key problem for Rang De: being a relatively small entity makes it hard to "stand out". This leaves Rang De with the challenge of capturing the attention of potential lenders and highlighting the positive differences between its work and that of other loan providers. Digital attention-buildingfocused Internet-led efforts to spread awareness and to capture the attention of prospective lendershappens through social media and the Rang De webpages. F I G U R E 1 Enactment of hybrid focus: First-order codes, theoretical categories, and aggregate theoretical dimensions feed. Rang De uses social media to share "impact stories" as noted above, and to attract the attention of prospective lenders by showcasing small loans as a powerful tool in the global fight against poverty. The impact of digital attention-building resonated heavily in interviews with lenders, who reported having become aware of Rang De through reposts, retweets, or messages received by personal acquaintances reporting a positive lending experience.
According to an informant who started lending via Rang De in 2015: I could tell from the videos, pictures and the way things were presented that they were doing good work. The Rang De website certainly got me interested and it compelled me to know more about their operations. And then I started to lend on Rang De.

| Digital credibility-building
Digital credibility-building refers to web-based actions that develop Rang De's image as an honest and ethical prosocial P2P lending platform. Building credibility is a key challenge for Rang De given many well-documented negative stories of "prosocial" microfinance initiatives that ended up not fulfilling their original purposes. Interviews with lenders revealed that initially, it was hard for them to distinguish Rang De from many projects that, while depicted as having a social goal, resulted in fraud or systematic misuse of investors' money. According to several lenders suspicion tends to be the first reaction. As two lenders put it: I came across Rang De through a post on The Better India website and my reaction was "scam, fraud!" After all, what seems too good to be true is, more often than not, just thattoo-good-to-betrue. After what seemed like a lifetime of hearing about such initiatives that went wrong, I had low trust to start off with. and At the very beginning you don't know what is going to happen. I mean, you are not familiar with the founders, you are not familiar with the organisation at all. All they have is a webpage, so you don't know if it is going to help people or it is a fraud. You don't know how it's going to turn out, but I'm happy it did turn out well, and happy it is not a fraud…because, you know, it is kind of hard to say.
The majority of the things that go on [in the microfinance sector] are not so transparent.
The other problem for Rang De is the perception that low-income borrowers are unlikely to repay. According to a regular Rang De lender: Some individuals think that poor people do not want to pay back money. This thought is ingrained in some people's minds because of societal factors…so I can say that this idea about poor people not being willing to pay back is an idea that stays in the minds of the relatively well-off.
Rang De uses technology innovatively to build credibility in the lender community. Presence and high visibility on social media is considered particularly important, as it helps the transparent and truthful documentation of changes brought to people's lives. Rang De employs its website and social media channels as a trust-building tool, as detailed by informants in the communications team: In the early days we would only make videos rarely and sometimes just to explain the impact that we've had saying "This is an example of someone who benefited from a Rang De loan. This is in her own words. This is what she says." But that is not enough. Now if you go on the Rang De website, every day, every month there are new needs and instead of telling people "This is now available. Please go and invest," we make these videos so that they're easy to understand. and You're reading so many posts in a day, you're reading so many tweets in a day, how do you know what is right and what is worth following? You can only know through constantly generating more content and then sharing that and you stay true to what you're doing, ensuring that every time there's a new partner we send out a post; every time there is someone whose life was impacted and there's a good story to tell, we put that up.  That empathy-building efforts appeal to both pan-Indian and sub-Indian identities of lenders is validated by the fact that many lenders choose to invest in their state of origin, or in a business operating from a location in India they have prior experience with.

| Intermediary relationship-building
Intermediary relationship-building refers to Rang De's efforts to develop and sustain strong collaborative links with intermediary organisations. The relationship that Rang De establishes with borrowers is not mediation-free as in many P2P lending platforms, but mediated by "impact partner" intermediaries who have direct knowledge of borrowers. The Rang De model depends on these intermediaries who work closely with low-income individuals and enrol them as borrowers. The intermediaries (typically NGOs) play the crucial role of identifying borrowers, screening, and accepting their applications and making sure the loans are used for the right purpose (eg, to support the growth of a micro-business). Intermediaries also work with local community self-help groups (SHGs), which enable their members -typically low-income owners of micro-businesses -access Rang De loans. Thus, through the "impact partner" intermediaries a relationship of trust is built in which borrowers have a familiar local organisation to turn to at all points in the lending and repayment process. Intermediaries play a critical role in the Rang De model -they connect borrowers with lenders, thereby making it possible for a relatively small P2P platform like Rang De to expand its network to 18 Indian states. As observed by the Head of Rang De's impact team: Our expertise lies in technology, processes and fundraising. The partners are the ones who have the social capital. They enjoy trust with the community; they are able to do that kind of needs assessment, gather information about the borrowers' repayment ability or follow up with them in case there are delays. So we partner with these organisations to access borrowers.
The "impact partner" intermediaries also collect repayments from borrowers and transfer the monies to Rang De, which then transfers the agreed rate of interest to lenders. The intermediaries' proximity to borrowers and SHGs also helps them keep track of emergency situations that may lead to monthly instalments not being repaid as per schedule. Rang De's conversations with intermediaries revolve around the practicalities of managing loans on the ground. For example, if any of the borrowing terms and conditions are infringed, members of the Rang De's impact team bring it up with intermediaries and discuss how the problem could be solved. There are regular feedback meetings, wherein loan administration and the ways in which it may be improved are discussed. A Rang De informant elaborated on the relationship with intermediaries: With intermediaries, we only discuss borrowing and lending, not what we do on the platform. For them, we are a microfinance organisation (…) we speak with them about managing loans and keeping an eye out for lenders, who charge very high interest rates.
Intermediaries also emphasised their commitment to the welfare of low-income borrowers. As the head of an intermediary organisation in Bangalore explained: We don't refer to borrowers as beneficiaries. For me it's a very loaded term -I don't use it. That implies a setting in which I give and you takethat's not our relationship. Our relationship is membership based. You and me are equal. So we don't use the word beneficiaries at all.
We also found that intermediary relationship-building did not always ensure that rules were followed. For example, when visiting borrowers in rural Mysore, we encountered a few cases in which the loan appeared to be misallocated: two women interviewed during our visit, candidly admitted having used their loan not for intended purposes, but to support their respective husbands' businesses. One of our Rang De informants explained that redirection of loans was breaking Rang De's loan rules, although such violations did not seem to preoccupy lenders who were inclined to take the view that wherever loan money goes, it helps a poor family.

| Borrower relationship-building
Borrower relationship-building refers to Rang De's efforts to develop and sustain strong direct links with borrowers.
Rang De conducts frequent "field visits", during which Rang De's impact team (typically accompanied by an impact partner intermediary) interacts closely with borrowers. This interaction takes the shape of semi-structured interviews which are, in the first place, appraisals of how the borrowers spent their loans, and of the extent to which the loan has satisfied their needs. Interviews then move to other urgent needs, and to the actions that Rang De can take to best serve them. Rang De also invites lenders to join these visits and meet borrowers. Many lenders take up this opportunity and report feeling happy and satisfied, seeing the "real" impact of their "social investment". During these field visits, Rang De does both intermediary relationship-building and borrower relationship-building. Going out to talk to borrowers improves Rang De's visibility in its target communities and helps attract both new borrowers as well as lenders: To attract lenders, we need to create an attractive digital interface. But to offer valid content on the interface, we spend long hours in the field: interviewing borrowers, getting to know the people, becoming aware of their struggles. The digital interface has no value if it is not supported by fieldwork with borrowers in rural areas.
In our observation of field visits, we found that borrowers know the intermediary well but are less familiar with Rang De as the source of their loan. In this sense, field visits help bring Rang De to the attention of borrowers, strengthening its overall credibility and trustworthiness. Rang De's relationship-building efforts with borrowers, thus, helps retain existing borrowers and enrol new ones. Field visit interviews with borrowers take place in private or public village spaces with no traces of the digital identity that Rang De projects to lenders. In the discussions that take place with borrowers, Rang De employees do not refer to the digital aspect of the platforms, but only discuss the loans made, the ways these are managed by the individuals and SHGs, and possible issues with repayments: [With borrowers] we do not talk about our digital job. We talk about loans, how they are using them, what they have gained from them, and how we can make them work better. In fact there is no digital work when we are on the field, very often we are isolated and completely disconnected from the Internet for many days.
Field visits also serve another important purpose. Rang De improves its awareness of the conditions borrowers face, their emerging needs, and its own ability to update and refine loan products. This kind of intensive engagement with borrowers has further enhanced lenders' appreciation for Rang De. In the words of a Rang De informant we shadowed during a field visit: It is important to meet borrowers directly, to know exactly what they need from the loans and if they are receiving exactly what they need. But it is also important to do so to make sure that the community knows about us, and is willing to engage with what we bring to them.
Drawing on the experience of its field visits, Rang De has enacted a new project called Swabhimaan (meaning "self-respect" in several languages), which has recently been piloted in a village of rural Karnataka. The project is based on installation of bioscopes (digital kiosks) in villages, which enables potential borrowers to complete a financial literacy module and take a test of basic financial literacy. In the locations where the Swabhimaan model has been piloted, residents who pass the test can apply for a loan without any mediation. The plan is to create Swabhimaan selfservice digital kiosks in many villages, which will allow digital applications and direct digital disbursals of loans. A Rang De technology team informant noted: The (SHG) model for loans works great in getting people, especially women, to come together but we are not blind to its problems (…) there may be a coercive element to the loans being made through SHGs. We don't know. What we are trying to achieve with Rang De Swabhimaan is to allow individuals to directly seek loans at low interest rates. The money will directly be transferred to their bank accounts.
Many people from low-income backgrounds in India do not have access to intermediaries or SHGs and, therefore, remain financially excluded. The "big vision" for Rang De, as one of its co-founders explained to us, is to become the "public distribution system for loans in India" that can be accessed by poor people in all states. This vision cannot be enacted by the current dependence on a mediated system of loans. The belief is that Swabhimaan could be scaled up, with India's Unified Payments Interface (UPI) payments system and the Aadhar biometric unique identity system as its foundation. Rang De's senior managers feel Swabhimaan could create the possibility of disbursing loans directly and help Rang De establish a more direct relationship with its borrowers.

| DISCUSSION
The above analysis of Rang De shows how prosocial P2P lending platforms enact their hybrid focus through two types of activities: digital work comprising Internet-based (ie, online) clusters of actions and fieldwork comprising field-based (ie, offline) clusters of actions. Rang De's enactment of its hybrid online-offline focus resonates with the different approaches to managing hybridity highlighted in the literature (eg, Battilana & Lee, 2014;Kraatz & Block, 2008;Pratt & Foreman, 2000).
Rang De's digital work -digital attention-building, digital credibility-building, and digital empathy-building -is mainly targeted at the supply-side audience (ie, lenders). It is noteworthy that Rang De's fieldwork activities are frequently referenced in its digital work. For instance, video narratives of borrowers' fledgling micro-businesses and loan requests produced during field visits are made available on the Rang De website and form an important part of empathy-building. Similarly, recordings of interviews conducted with borrowers in their villages, during which they describe the positive impact of a Rang De loan on their respective micro-businesses, are uploaded on the website and contribute towards credibility-building. Rang De's digital presentations feature many such carefully curated fieldwork moments, and Rang De impact teams' many face-to-face interactions with borrowers are captured and represented in digital format. The sensitive visual portrayals of borrowers' lived experiences add an important layer of authenticity to lenders' online experience. In other words, Rang De's digital work articulates an investment opportunity for prospective lenders, whilst also providing vivid evidence to support claims of positive social impact. There is a clear emphasis on representing and integrating offline actions into its digital work. Rang De's enactment of a hybrid focus thus contains shades of the "integration" (Battilana & Lee, 2014) approach highlighted in the hybridity literature.
Rang De's fieldwork actionsintermediary relationship-building and borrower-relationship buildingreflect its prosocial orientation. They are also different from other prosocial P2P lending platforms, where positive social impact is assumed, with the platforms themselves staying removed from borrowers' lives and possessing little contextual awareness of the realities of rural life (cf., Mckinnon et al., 2013). Rang De invests significant time and resources in its engagement with intermediaries and borrowers. The involvement often goes beyond disbursing loans and collecting payments. In practice, Rang De's "impact team" performs the role of a socially responsible and locallyembedded microfinance provider, immersing itself into the social world of rural India, negotiating complex cultural norms, and hierarchical systems and having a much more hands on approach to its engagement with intermediaries and borrowers. Rang De's fieldwork is also significantly removed from its digital identity. It rarely refers to the P2P lending platform in its interactions with intermediaries and borrowers. The focus is simply on explaining the process of accessing Rang De loans, which from a low-income borrowers' point of view, come at a significantly lower rate of interest than the alternatives. Thus, to the extent that Rang De keeps the digital identity hidden in its fieldwork actions presenting itself as just another lender, it follows what the literature refers to as the "compartmentalization" or "separation" approach (Kraatz & Block, 2008;Pratt & Foreman, 2000) to managing hybridity. This separation also has a structural component. Rang De's impact team does fieldwork and the technology team maintain the web operations. As noted in the analysis, Rang De's senior managers are also experimenting with Swabhimaan, a novel and creative project that has the ambitious aim to disintermediate the entire borrowing process for low-income individuals. Rang De's plans for overcoming the limits of the current model through the Swabhimaan project highlights a new and uncharted paradigm for prosocial P2P lending and thus, echoes with the "creative" approach (Battilana & Lee, 2014) highlighted in the hybridity literature.

| THEORETICAL CONTRIBUTIONS
This paper has explored the enactment of a hybrid online-offline focus in prosocial P2P lending platforms. It makes multiple contributions to scholarship. First, it shows how the enactment of a hybrid focus in prosocial P2P lending takes the form of a medley of hybrid practices rather than any one particular hybrid approach. Platforms' obligation to produce vivid depictions of social impact for their online lender audience leads to an "integration" approach to hybridity. Similarly, low-income borrowers' obvious disconnect with the digital core of the model leads to the implementation of a "compartmentalization" approach in managing demand-side relationships. There is also a temporal component to this medley of hybrid practices. Over time, actors may identify shortcomings of their strategies leading them to plan for paradigm-changing "creative" solutions such as total disintermediation of the borrowing process.
Second, the Rang De experience suggests that prosocial P2P lending platforms sustain and expand their lender base when they pack both emotion and transparency into their digital work. Whilst gripping emotional content can help prospective lenders identify closely with prospective borrowers and trigger feelings of compassion and empathy, transparent detailing of operations is central to lenders' investment decisions. In contrast to many P2P lenders whose operations are shrouded in secrecy with sophisticated but opaque algorithms at the heart of the business model, Rang De is a model of exemplary transparency. It specifies in detail where every lender rupee is going. This detailing is not only in terms of spelling out the interest rates for different loan products, but also in providing lenders real time updates of the flow of money from their account to the borrowers' (and in the opposite direction, for repayments). I would suggest that for many prosocial P2P lending platforms emotional content may be easier to sustain than transparency. However, presenting audited accounts of operational expenses, maintaining high disclosure standards about the processing of loans, and showcasing the track record of intermediary organisations could help prosocial P2P platforms underline their social focus and scale up operations.
Third, the social innovation aspirations of prosocial P2P lending platforms are both enabled and constrained by intermediaries (cf. Burtch et al., 2014;Coates & Saloner, 2009;Johansen & Nielsen, 2016;Sandeep & Ravishankar, 2018). Rang De's experience indicates that whilst they can reach lenders through their online efforts, the only way to reach those living in conditions of extreme poverty is via local charities and social enterprises that work with poor communities in rural areas. The extent to which local intermediaries share the vision of prosocial P2P lending platforms is bound to vary across settings (cf. Schwittay, 2019). Intermediaries are typically resource constrained but engaged in multiple grassroots' initiatives focusing on sustainable livelihoods. The engagement with a P2P lending platform is one amongst several relationships they maintain. In this respect, they may not see the P2P lending platform as a priority partner. Rang De senior managers are meticulous in identifying and screening intermediaries for alignment in terms of social mission and this process obviously improves the probability of the loans generating a long-term positive impact on the lives of borrowers. At the same time, prosocial lending platforms cannot function without intermediaries, and it is a part of the model they are not always in control. For instance, as illustrated in our analysis, despite the platform's best intentions a proportion of loans can be misallocated when intermediaries misrepresent a borrower's need for a loan. In short, a pervasive digital divide in many societies means that a layer of complex non-digital intermediation is inevitable in the quest to address poverty through digital technologies.
Fourth, notwithstanding their dependence on intermediaries, prosocial P2P lending platforms, precisely because of their sophisticated supply-side digital infrastructure, can have a significant positive impact on financially excluded individuals trapped in vicious cycles of poverty. Rang De loans have been accessed by low-income borrowers in 18 Indian states. In Rang De's absence, many of these borrowers would have either been unable to access credit or suffered because of usurious rates of interest (cf. Carr et al., 2016;Cull, Demirguç-Kunt, & Morduch, 2007;McKinnon et al., 2013). In this respect, prosocial P2P lending can touch the lives of marginalised low-income borrowers. This capacity to transcend institutional limitations and reach the so-called bottom-of-the-pyramid (BoP) markets underpins the transformational potential of prosocial P2P lending.
Finally, prosocial P2P lending platforms face a difficult choice between committing to "depth" of social impact and investing in "breadth" of social impact. This dilemma is also a reflection of the trade-off between reaching the poorest borrowers and profitability (Cull et al., 2007). Rang De's intention is to create a deep impact on the lives of the poorest and hence, senior managers are extremely selective in their choice of intermediaries. As explained in our analysis, robust "due diligence" practices are followed to make sure that they only work with intermediaries with a shared vision. Rang De's field activities testify the platform's social commitment to all borrowers. However, this focus on building close relationships has meant that Rang De is working with a relatively small group of intermediaries and borrowers. Success, in terms of scaling-up its operations, has been modest. An alternative approach would be to focus on what may be termed as the "breadth" of social impact whereby platforms expand rapidly and reach out to as many intermediaries and borrowers as possible, without necessarily committing to intensive fieldwork activities.
The argument for focusing on "breadth" is that it allows a much larger proportion of low-income individuals to access loans at below-market interest rates, although platforms may find it hard to evaluate social impact in detail.

| PRACTICAL IMPLICATIONS
For managers of prosocial P2P lending platforms, intermediary relationship-building is an important ongoing activity.
Intermediaries help establish local presence, facilitate the borrowing process and help enter new BoP markets. The more ambitious the social orientation with respect to targeting the poorest (eg, those living on less than US$2/day), the greater the importance of ethical intermediary organisations with a good track record of working in rural areas.
Thus, it is crucial for managers to complement advanced digital technologies with the creation and development of relationships on the ground. This process could take time. But it stands the best chance of creating an impactful social innovation at scale. Thus, senior management teams with a diversity of experiences and complementary skillsets (eg, IT and development economics) may be required to manage the hybrid focus of the platform. In the absence of deep contextual intelligence about the BoP markets they wish to serve, senior management teams may struggle to meet their espoused social goals around poverty reduction, although they may still manage to survive as a P2P lending platform.
The model's overall success also depends on senior managers' ability to recruit and retain talented people. The higher salaries offered in more commercially oriented companies can be a big temptation for current employees (especially for those working on the technology side of the model). Thus, high attrition rates could become a big problem. At the recruitment stage, senior managers may need to carefully assess the "social" orientation of the applicants alongside their technical capabilities. As some recent research has shown, social encoding of employees are central to successful social innovations (eg, Sandeep & Ravishankar, 2015).
Transparency of operations could be a significant source of competitive advantage for prosocial P2P lending platforms when stacked against competitors who highlight powerful algorithmic capabilities but provide little detail about their operations. Maintaining high standards of transparency about how the model works could help enhance reputation and scale the business. For instance, as seen in Rang De, creating opportunities for lenders to visit borrowers so that they may better understand local settings and assess the impact of their loans, can develop platformlender trust.
At its core, Rang De type P2P models have a strong social focus and may find it difficult to generate double-digit return on investments for stakeholders without diversifying into non-BoP markets. This diversification could represent a dilution from the platform's explicitly stated goal of fighting poverty. Managers may need to develop a framework of ethical principles that will help them manage trade-offs between the diversification potential and the social orientation of the platform. Rang De's experience suggests that investments in (online) digital capabilities enhance the diversification potential of the model, whilst the (offline) field capabilities enhance the "social innovation" potential of the model. Platforms could also benefit from creating structured offline operations on the supply side. They may help overcome a current "digital divide". Rang De lender profiles show that few lenders are above 50 years of age, which indicates that platforms may be missing an opportunity to recruit older lenders willing to lend but unwilling to go online to do so. Although offline supply-side operations increase the cost base, they also help platforms significantly expand their lender base.

| PROSOCIAL P2P LENDING PLATFORMS AS A TOOL IN THE FIGHT AGAINST POVERTY: WHAT CAN GO WRONG?
Prosocial P2P lending platforms are a striking illustration of a DSI that helps the poor access loans at below-market interest rates. The expectation, of course, is that such loans will present opportunities for the poor to escape poverty traps and improve their lives. Rang De's story illustrates a sensitive and conscientious approach to prosocial P2P lending. At the same time, extrapolating from Rang De's case it is possible to consider some of the possible pitfalls and unintended consequences of the model in its fight against poverty.
The financial sustainability of the model when it operates as a not for-profit charity is linked to donations. When the platform operates as a for-profit enterprise its sustainability depends on maintaining both a steady supply of lenders willing to invest and borrowers with good repayment records. The strong digital capabilities of such platforms mean that senior managers may find it easier to expand their lender base than expanding their borrower base of low-income individuals. The obvious solution to the problem would be to outsource most of the demand-side activities (eg, identifying borrowers, assessing their financial circumstances, verifying the credibility of the loan request, etc.) to intermediaries. The implementation of this "solution" could escalate to serious reputational damage with accusations of "impact partner" intermediaries being nothing more than glorified loan repayment collection agents. The point here is not that intermediaries should not disburse and collect loan repayment. But it is the worrying possibility that in terms of the social mission of lending platforms, the model could leave the poor almost entirely at the mercy of intermediaries, leading to exploitation rather than the expected empowerment. Similarly, pressure to grow the business could lead platforms to diversify their product offerings in ways not directly related to their social innovation focus. They could start targeting individuals who are not really in desperate need of loans. The implication is that the platform could drift away from the "social element of the 'digital innovation'".
Another criticism of the prosocial P2P lending model it that it encourages "everyday humanitarism", offering simplistic solutions to the complex and multifaceted problem of global poverty (Schwittay, 2019). Further, rather than tackling deep rooted structural inequalities in society, the model may reproduce cultural tropes and objectify the rural poor as always in need of help from urban saviours. The argument is that projected images and videos of prospective borrowers in desperate conditions may serve to encourage "financial clicktivism" and create the naïve impression that lending money with a few clicks on the computer is a long-term solution to the problem of creating sustainable livelihoods for the poor (Schwittay, 2019). The field visits for lenders arranged by P2P platforms could similarly face accusations of promoting "poverty voyeurism" (cf. Weiner, 2009). Finally, one could argue that prosocial P2P lending platforms invade the privacy of the poor and strip them of their dignity by presenting their life stories on the internet for public consumption.
The origins of some criticisms are "critical" scholars in academic institutions and civil society organisations with particular ideological orientations. These entities (rightly) bemoan the state's negligence of the poor, but (wrongly) almost always attack private enterprise led social innovations, demanding impossible moral standards from them.
Drawing on the Rang De case, I would argue that the point for prosocial P2P lending platforms operating in challenging settings is not that criticisms are invalid, but that they seem to offer no actionable solutions to the challenging problem of maintaining a hybrid focus and generating a positive impact on society. Therefore, I suggest that misplaced ideological critiques can do a great disservice to prosocial P2P lending models, risking their future by encouraging regulatory overreach, and affecting the poor's ability to access loans at low interest rates.

| LIMITATIONS AND FUTURE RESEARCH
A comprehensive evaluation of the impact of digital interventions on global poverty requires large scale field studies incorporating a longitudinal design. This paper is limited in that it has detailed the experience of only one P2P lender in helping low-income individuals access loans. It is important to recognise that the videos narratives and images of satisfied and happy borrowers show their feelings at a given point in time. Further, in our data collection, we were not able to ascertain exactly how many video narratives of borrowers were produced by the Rang De impact team and how many were generated by intermediaries. Further longitudinal research is needed to explicate the processes through which loans generated in the prosocial P2P lending model impact families and contribute to addressing challenges of crippling poverty.
An important assumption anchoring our study is that prosocial P2P lending platforms are well-placed to contribute to the United Nations' SDG 1 goal of poverty reduction (ie, they can be impactful DSIs) because such a social mission is integral to, and not just a by-product of, their business model (cf. Sandeep & Ravishankar, 2015). This assumption somewhat limits the scope of the paper. Whilst such a premise is useful in highlighting the work of DSIs that engage with the poorest and most vulnerable communities, it may (unintentionally) create the impression that other more commercially-oriented P2P lending platforms play only a limited role in reducing global poverty. Future research needs to compare the social impact outcomes of prosocial P2P lenders with mainstream P2P lending platforms and rural microfinance organisations.
The role of regulators has been mostly ignored in this study. As noted in the Research Methods section, Rang De was a public charitable trust until 2017 when it changed to a private limited company structure. Our study has explained Rang De's lending model and its hybrid focus without considering the impact of change in legal structure on operations. Although the ramifications of this change are not central to the arguments of this paper, the pivoting of P2P lending platforms from one legal form to another in response to regulatory requirements is a worthwhile topic for research. Future empirical work also needs to evaluate the impact of a change in legal form from non-profit to for-profit on supply side behaviours. For example, it raises the interesting question of whether lenders on prosocial P2P platforms prefer "giving" over "lending" (cf. Moss, Renko, Block, & Meyskens, 2018).

| CONCLUDING REMARKS
Prosocial P2P lending platforms are a novel and powerful illustration of a DSI. They help the poor access loans at reasonable rates of interest and thus, they have the potential to be an important socio-technological instrument in the global fight against poverty. In this paper, I have explored and discussed the specific ways in which they build supply-side (lender) and demand-side (borrower) relationships. I urge scholars to build on this study and contribute to a more thorough understanding of the "social innovation" potential of P2P lending platforms and the challenges they face in enacting their hybrid focus.

ACKNOWLEDGEMENTS
I am indebted to Silvia Masiero (silvima@ifi.uio.no) for her significant contributions to the data collection and data management process. I am grateful to Rangan Varadan, Abhay Kulkarni, and everyone at Rang De for their time and generosity.

ENDNOTES
1 In biology, hybrids refer to the offspring of two different species.
2 The business-charity combination is only one example of hybridity. It does not define the notion of hybridity.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.