1 Prana’s thesis emphasises that financial returns and development impact can be mutually reinforcing. What early experiences or market observations shaped this conviction? In the Indian context, what specific gaps were you aiming to address that mainstream venture capital was overlooking?

Our conviction that financial returns and social impact can be mutually reinforcing is based on our learnings from 15 years of investing at Omidyar Network India, where the team has backed over 90+ companies serving India’s lower-middle-income (LMI) population. We saw first-hand that businesses solving genuine problems for this population such as affordable healthcare, accessible credit, better livelihoods with a good understanding of their customers and frugal innovation (product or business model innovation) built extraordinarily loyal customer bases that translated into strong unit economics and durable growth. Technology has made access to the LMI very cost-efficient and companies can create impact at scale if they deliver a consistent value proposition, and generate financial returns for their shareholders in the process.

Over this journey, we saw a structural gap in early stage funding for solutions aimed at India’s LMI population. Impact investors focused on low-income households using traditional brick-and-mortar models, while mainstream tech VCs chased the top 10% of consumers, often through technology-led solutions. Very few were building a dedicated practice around tech-enabled solutions for the 270 million “Aspirant” and “Seeker” households, families earning ₹1.25–₹15 lakh annually who are highly digitally savvy, value-conscious, and increasingly aspirational. Prana aims to pair our deep sectoral knowledge and understanding of our target audience, with a right-sized fund, to back capital-efficient, technology-driven companies purpose-built for this massive, underserved segment.

2 Prana’s focus on Aspirers and Seekers, households earning below ₹15 lakh annually requires solutions that go beyond affordability to true usability and relevance. What indicators does Prana use to assess whether a product is genuinely designed for these users? Where do new platforms most often misjudge behavioural patterns, trust dynamics, or last-mile realities?

We look beyond the pitch deck to real usage patterns. Our due diligence involves interacting with the end consumers to examine how useful or impactful products are and whether they are built around vernacular-first interfaces, sachet pricing (small-unit, pay-per-use models), and workflows that accommodate intermittent connectivity, for example. We constantly combine field visits and third party studies/ interviews to measure not just satisfaction, but genuine quality-of-life improvement.

Where platforms most commonly stumble is in transplanting urban, English-first product assumptions onto LMI users and communicating with them with straightforward translations of English messages instead of custom-building communications for LMI. Companies often get pricing and business models (e.g. upfront vs. pay-per-use) and unit economics (e.g. average order values, logistics cost, customer acquisition cost) wrong in trying to adapt from upper income/ global models, again. Wiom (home broadband solutions for LMI) and DrinkPrime (clean water at home) are successful examples of our portfolio companies that have developed unique models specific to the LMI at scale. Trust dynamics are different—these customers rely heavily on word-of-mouth and local networks, so referral loops matter more than digital marketing. Last-mile realities like shared devices, limited data packs, and preference for YouTube and local-language content are often underestimated. Our portfolio companies such as FarMart in agritech, Vedantu in education and IntrCity in mobility have succeeded because they designed around these realities from day one, rather than retrofitting.

3 India’s digital public infrastructure, including UPI and ONDC, has significantly altered the operating environment for technology-led models. How are Prana’s portfolio companies leveraging these public rails? Beyond enablement, how does DPI reshape unit economics, distribution, or customer acquisition?

Digital Public Infrastructure has been transformational for the segment we invest in and startups embedding DPI into their workflows have benefited disproportionately. UPI alone processes over 680 million daily transactions at an average ticket size of ~$15, precisely the high-frequency, low-value transactions that define LMI commerce for companies in our portfolio such as Dealshare and Peppertap. DPI has made viable a whole category of business models that simply could not exist ten years ago.

Some examples from our portfolio:

• Kiwi, a credit on UPI platform, is making personal credit lines available to hundreds of thousands of new-to-credit customers in partnership with leading private and public sector banks at affordable rates for the first time.
• M2P, a fintech infrastructure enables access to financial services for the LMI through its banking as a service stack, bringing down the cost of lending services, KYC and digital origination.
• Wiom is riding on the PM WANI public wi-fi framework to proliferate access to broadband at a fraction of the cost that it would take to make private broadband available at every home.
• Hexa Health is using the Unified Health Interface to approve hundreds of patients who enrol every month under the Ayushman Bharat Digital Mission for ophthalmic surgeries, barely within 5 months of launching the programme.
• Teal is building digital access to asset datasets available from the ULI, Vaahan and other public databases, enabling better and faster underwriting by large financial institutions.

Looking ahead, we are excited by what the ONDC, National Health Stack and Agri-Stack have in store for Prana’s focus sectors. At its core, DPI is reshaping unit economics by reducing costs to acquire/ onboard and costs-to-serve customers, both of which are sensitive levers to make business models targeting the LMI audience viable.

The Aadhaar + Account Aggregator framework allows consent-based data sharing that dramatically reduces KYC + credit underwriting costs for lending companies, as an example. The Unified Lending Interface launched by RBI is transforming small-business lending, particularly for managing risks through common data sharing, which could translate into lower systemic costs and interest rates in the long run. ONDC has the potential to significantly reduce last-mile logistics costs for products being delivered to the LMI, and along with UPI, which has already reduced payment costs to a fraction, can make rural e-commerce in India amongst the cheapest in the world.

Going beyond costs, the National Health Stack and the Agri-Stack, if implemented well, have the power to radically transform sectors, where lack of standardised patient health data and information on farmers have historically been huge inhibitors, thereby allowing newer and highly innovative, customised solutions to emerge.

4 Prana emphasises capital-efficient, high-volume, low-ARPU models while operating in sectors that may require longer gestation. How does the firm evaluate trade-offs between near-term capital efficiency and longer-term value creation? What role do fund size and stage focus play in balancing returns with impact?

The tension between capital efficiency and a longer gestation period is real, but we have found it is often overstated. After the correction of 2022–23, when global cost of capital rose sharply, the market has firmly moved away from burn-for-scale towards sustainable unit economics. This actually suits LMI-focused businesses, which were never designed around heavy subsidisation. High-volume, low-ARPU models work when you achieve operational frugality: high sales-team productivity, low-cost distribution, and product designs that are inherently lean.

We believe that capital efficiency has a direct correlation to long-term value creation, and is not actually a trade-off, for a few reasons. For starters, it allows young companies to survive through funding winters, while high burn models have a higher likelihood of perishing soon. Second, companies tend to be asset-light, prioritise high margin lines of business, and maintain low costs, and thereby facilitating shorter path to profitability. Third, capital efficiency leads to lower dilution for founders and investors alike, thereby increasing the probability of higher return multiples. Lastly, public markets tend to reward high RoCE businesses with valuation premiums.

Our target size of $100–200 million for every fund vehicle has been deliberately calibrated for the stage we invest in and likelihood of success, while many of our peers have gravitated to larger (>$500 million) fund sizes – as a function of large teams, economics, ambitions – and also investing across stages as a result. Our fund size ensures we do not feel pressurized to deploy funds within the investment period, especially if it means wrong market timing, while being sizeable enough to ensure portfolio diversification and relevance to each portfolio company. Importantly, a smaller fund size significantly increases the chances of the classic ‘fund-returners’, critical to make the early-stage VC model work, as validated empirically. This is a structural advantage, particularly for the sectors and business models we invest in.

We also believe our consistency in stage of investing meaningfully improves our chances of success. For example, we see a right-to-win and our best success rate when most VCs are not yet ‘hot’ on a particular sector, while our nuanced understanding of how a sector is evolving helps us form an early view and invest ahead of the curve. We have seen this play out in MSME lending within financial services in 2012, live online learning in education in 2016, agri-output marketplaces in 2019 and so on. Clarity on stage allows us to design our firm around a high-risk, high-return framework – bringing alignment on fund construct, number of companies in a fund, what fits the bill (stage and returns profile), entry valuation and ownership, due diligence processes commensurate with the stage, and target hold periods for each of our investments, all of which are critical for financial success.

Early stage investing best allows us to also shape the impact that our portfolio companies can have on the LMI through our ‘impact acceleration playbook’, learning from our engagement with the likes of Wiom, VerSe, DrinkPrime, Hexa Health and many others from our portfolio. This hypothesis was validated by an independent impact assessment study conducted by Dalberg, which interviewed a number of our portfolio founders to determine how impact investors can be different from commercial VCs.

5 Assessing the depth and durability of development impact, across financial inclusion, livelihoods, and climate resilience remains complex. How does Prana operationalise impact measurement in practice? Additionally, as technologies such as AI reshape these sectors, how is the firm ensuring that innovation remains inclusive rather than widening existing gaps?

We measure impact along three dimensions, Reach (number of lives touched), Depth (quality-of-life improvement), and Inclusion (percentage of customers from LMI segments). In addition to data from the portfolio companies themselves, we have commissioned periodic third-party surveys through impact measurement agencies such as 60 Decibels, interviewing thousands of end-users. We further map investments to SDG targets and align metrics with IRIS+ and ESG frameworks, while we work on adding climate and sustainability metrics into our framework. We use a proprietary scoring system, which integrates four dimensions namely impact depth, scale, vulnerable population reach, and poverty alleviation strength, into a composite score. This lets us compare impact across very different sectors in a consolidated and consistent manner.

6 As AI reshapes sectors from MSME credit to climate risk and education, it brings both opportunity and the risk of widening inequities. Looking ahead, how is Prana evolving its investment approach to ensure these technologies drive more inclusive outcomes rather than deepen existing gaps?

AI is the most significant enabler we have seen since UPI for the segments we serve. We see enormous potential for inclusion if deployed thoughtfully.

Consider MSME credit: traditional underwriting excluded most small businesses because data was scarce. DPI-enabled access to GST filings, property data, KYC (MCA, IT), UPI transaction flows, account aggregator models created over the last few years improved access brought about by digital trails. AI models trained on these will make access and decision making faster and cheaper. In agriculture, AI-powered advisory tools can deliver crop-specific, location-specific guidance to smallholder farmers through a basic smartphone, something that once required an army of analysts. In education, we are seeing AI models that personalise the learning experience of every student: replicating a master teacher experience at a fraction of the cost, remedial learning for a child left behind, doubt resolution at an individual step level, all unimaginable in a regular classroom learning experience. We are therefore excited by the potential that the technology has for accelerating impact and our ‘inclusion’ measurement (% of user base from the LMI segment) applied to each portfolio company over a period of time will inform us and hold us (and our founders) to account on inclusive outcomes.

At the same time, we are sensitive to exclusion risks: does the model work for low-literacy users? Does it function on basic smartphones? Work in low-bandwidth environments? Is the training data representative of LMI users? Can the interface serve someone more comfortable in Hindi or Tamil over English? We are also very conscious of gap risks: Will this benefit only the rich and not reach the LMI owing to high costs? Will it eliminate/ replace livelihoods for the LMI?

We believe that AI will take an embedded form and shape in the Indian context, particularly for the LMI, without the user actually recognising it as AI – making it more ubiquitous. Prohibitive costs today will likely mean that customised LLMs or SLMs, with voice as the default interface, will emerge to make it more affordable and accessible. Our choice of investments is already geared towards companies that are making AI accessible for LMI consumers and small businesses. Our investment decision-making is evolving to identify and measure the AI impact (including unintended consequences) that each portfolio company will have on the LMI. Our education and MSME investment theses will also reflect the need to re-skill millions of Indians for jobs that are being replaced by AI en masse. Our investment in companies like Masai and Entri (AI-ready skilling) and i-Merit (AI data annotation employing trained LMI workers) reflects our belief that AI should create opportunities for this population, not just extract value from them.

Our approach and positioning is evolving from “tech-for-impact” to “AI-for-impact,” but the core question remains the same: does this genuinely improve the quality of lives of the bottom 93%?