Based on insights from the Financial Inclusion Lab engagements, this playbook helps FinTech founders navigate partnerships with banks, nonbanking financial companies (NBFCs), microfinance institutions (MFIs), and other financial service providers. The playbook outlines key stages and emphasizes stakeholder mapping, early engagement on compliance, clear documentation, and organizational buy-in to help startups turn promising pilots into scalable, long-term partnerships.
Blog
Unlocking SHG potential through Account Aggregator: A white paper on data-driven financial inclusion
This white paper examines how integration of the National Rural Livelihoods Mission’s (NRLM) LokOS platform with India’s Account Aggregator framework can strengthen financial inclusion for self-help group (SHG) members. Secure, consent-based data sharing could improve financial visibility and access to formal credit, insurance, and other services. The paper presents a roadmap for data standards, regulatory approvals, consent frameworks, and field implementation. It also outlines safeguards and stakeholder responsibilities to ensure an inclusive, secure, and community-centered integration.
Adaptation finance needs a delivery system, not another fund
Why does so much adaptation finance still fail to reach the last mile?
The adaptation finance debate is dominated, even obsessed, by how much more money we need. That matters. But an equally important question gets far less attention: why does so much of the finance we already have fail to reach the last mile?
Why the adaptation finance gap is also a delivery gap
The United Nations Environment Programme estimates that developing countries will need about USD 310-365 billion a year for adaptation by 2035, while international public adaptation finance was USD 26 billion in 2023. That enormous financing gap is alarming. Yet a narrow focus on its size can obscure an equally important problem: we still lack an effective delivery system to move the right type of capital to the firms and households that need to adapt.
While we have been agonizing about the withdrawal of donor agencies, a very significant part of the adaptation finance challenge is operational. Capital and capability must move through a chain from policy to development finance institutions (DFIs), financial institutions, solution providers, and end users. Each handoff has different binding constraints. As SNV notes, viable businesses need different types of capital as they move from validation to scale, each with distinct risk profiles. Gireesh Shrimali of the Oxford Sustainable Finance Group at the Smith School notes in his LinkedIn post that actors should allocate each risk to the party best able to bear it. Given current funding constraints, we must also identify the minimum public funding required to make each transaction financeable.
The channels exist, and infrastructure gaps matter. Yet the architecture that connects needs, investable activities, appropriate products, risk-bearing instruments, and wholesale capital is missing.
Which two questions should determine the financing tools we use?
1. What capital fits the stage of development or maturity?
Grants and technical assistance are most valuable when they validate technologies, business models, measurement systems, and distribution channels that private investors will not yet underwrite. As the business model becomes clearer, concessional debt, guarantees, or impact-first capital can absorb or share the residual risks. Commercial capital becomes more plausible and feasible when cash flows, governance, and risk data justify it. This progression does not follow an automatic “graduation” path or fixed rule. Some adaptation investments generate substantial public benefits but little or no private cash flow and yet provide the foundations for other adaptation activities or enterprises. These investments will continue to need public finance. The north-star test is whether scarce public capital buys additional resilience or mobilizes finance that would not otherwise occur.
GCF FP252 project page; SNV, Building the Impact Capital Continuum
2. Which actor and instrument can best address each binding risk?
The practical question is not a singular one “Can the donor community or public sector de-risk this?” but rather a trinity of questions “Which specific risk prevents a viable investment from proceeding? Who can manage it most efficiently? And what is the least-cost transfer or sharing mechanism?” This reflects Gireesh Shrimali’s argument that we need targeted risk allocation rather than blanket guarantees, and often a combination of instruments where several risks constrain the investment. Similarly, the Organisation for Economic Co-operation and Development (OECD) recommends a portfolio of approaches to overcome the small transaction sizes associated with adaptation and financiers’ unfamiliarity with the asset class.
Typical matches depend on the binding constraint. Policy and regulatory risks usually require policy reform. Those reforms may then create opportunities for blended finance. Portfolio guarantees, contingent liquidity facilities, or meso-level insurance can address credit risk and covariant physical risks. As we have seen, and struggled with, in inclusive financial services, foreign exchange and tenor mismatches may require local-currency funding or hedging instruments. Where small transaction sizes make lending uneconomic, aggregation mechanisms and origination incentives could reduce transaction costs. Yet the global network of inclusive financial institutions built over the past 50 years remain a hopelessly underused channel for solving precisely this problem. MSC’s work shows that banks, microfinance institutions, cooperatives and other providers already have the clients, distribution infrastructure and transaction histories needed to originate adaptation-linked finance at scale. The challenge is to repurpose existing products, adjust terms and risk management to climate realities, and develop new products only where necessary.
What should a country-level adaptation finance architecture do?
Put the two frameworks together and six practical rules emerge. Each shift attention away from creating yet another financing window and toward repairing the hand-offs that stop adaptation investment from reaching the last mile.
- Start at the last mile and work upward.
Use participatory hazard maps, vulnerability and supply-chain analyses to identify the adaptations people and firms value, need and will use; their costs and cash-flow effects; and the products required to finance them. MSC’s locally led adaptation toolkit provides a seven-step route from supply-side assessment to a costed adaptation plan. Applications in Bangladesh, Uganda and India show how an analysis of hazards, livelihood impacts and cash flows can reveal specific adaptation investments and the product changes needed to finance them. MSC’s 3Rs (repurpose, rejig, and reinvent) framework offers a complementary approach to adapting financial institutions’ product suites. MSC’s recent work also distinguishes between the unresolved adaptation funding gap and an incomplete “financeability architecture.”
- Define what an investable adaptation transaction looks like.
Move beyond broad sector labels and establish a practical chain from adaptation need to activity, expected resilience and cash-flow effects, financeability, and appropriate instruments. Define eligible activities and outcomes, the evidence required at origination, safeguards against maladaptation, and how teams will measure adaptation benefits. This must be simple enough for a loan officer to recognize an eligible investment and standardized enough for DFIs to aggregate thousands of small transactions. MSC’s work on climate-smart agriculture in India shows how borrower segments, technologies, and provider risks can inform product design, operational processes, and standardized financing approaches.
- Separate capability gaps from risk gaps.
Diagnose the problem before selecting an instrument. Technical assistance (TA) can address weak financial models, data, governance, credit assessment, or product design, whereas a guarantee cannot. By contrast, another dose of training or TA cannot solve a genuine loss allocation or liquidity problem – these require blended finance. However, better climate-risk information does not automatically lead to better adaptation finance. If lenders identify vulnerable clients more accurately but do not change their products, credit assessment, or risk allocation, improved risk assessment may lead to less lending to those who most need to invest in adaptation.
- Layer risk management at both institution and client level.
CGAP recommends a layered approach combining reserves for routine volatility, contingent finance for medium shocks and meso-level insurance for severe shocks. Together, these layers provide a useful way to think about products that help clients invest, hedge, cope, and recover. Insurance can address defined residual risks, but it cannot substitute for grants, guarantees, suitable credit structures, or adaptive social protection where the underlying investment is not financeable.
- Move from bespoke deals to standardized portfolios.
Small transactions will not scale if every guarantee, loan, or adaptation assessment is custom-built for each specific hand-off or challenge. Standard product templates, pooled receivables, portfolio guarantees, and shared data can cut origination costs. The World Bank Group’s guarantee platform illustrates the institutional logic of simplifying and standardizing guarantee products. MSC’s climate-smart agriculture work in India translated borrower segments, needs, and technologies into standardized financing products and operational processes that regulated financial institutions can integrate.
- Measure additionality and market development alongside leverage.
Track who gains access to finance they would not otherwise have received, changes in price and tenor, portfolio performance, resilience outcomes, and the public risk-bearing required for each additional dollar mobilized. Aceli Africa provides an example already in progress. Its 2020-2025 program reports 55 lenders, 5,576 loans, and USD 440 million in agri-SME lending, with USD 10 in private lending per USD 1 of incentives, and explicitly accepts lower leverage to achieve higher additionality through smaller loan sizes.
So, the first question we ask should not be “How much concessional capital can we deploy?” It should be, “Which broken hand-off(s) prevent a viable adaptation investment from moving, and what is the minimum capability, risk-sharing, and capital package needed to repair it?”
The next frontier for GeM: From digitization to inclusion
How can the Government e-Marketplace (GeM) remove the operational and financial barriers that hold back small sellers?
The first blog in this series examined how GeM has transformed access to public procurement. Yet the journey extends far beyond registration. As that blog showed, three broad challenges still limit small enterprises: friction around deposits and platform charges, uncertainty in seller payments, and wider barriers to inclusive procurement. This blog examines those challenges and the digital solutions that could address them. GeM’s next phase of inclusion requires a stronger operational and financial architecture to underpin public procurement.
How can GeM simplify seller deposits and platform charges?
Micro-enterprises encounter several financial touchpoints when they participate in public procurement. These include earnest money deposits (EMDs), performance bank guarantees (PBGs), transaction charges, and annual milestone charges (AMCs).
Many of these transactions still extend beyond the GeM ecosystem. A transaction may begin on GeM, while external systems or physical channels complete, verify, or track the financial process. For example, sellers can initiate PBGs through GeM, but banks and buyers manage issuance, submission, verification, and release. Most related documentation remains manual and outside the portal. Buyers and sellers therefore have limited visibility into the status of deposits, guarantees, and refunds. The fragmented process creates avoidable manual work and makes end-to-end tracking difficult.
What would an integrated financial architecture look like?
GeM can replace fragmented financial processes with an integrated architecture. The government can bring these seller-side transactions onto GeM through deeper integration with banks and payment aggregators. Each journey could begin on GeM, route digitally to the relevant financial institution, and return execution and status updates to the portal. GeM could, for example, work with financial institutions to create an escrow-based mechanism for EMDs, similar to the one that existed for caution money. Sellers could apply through GeM. A partner institution would issue the EMD and hold it in escrow until the procurement outcome is known. The system could then automatically release or refund the amount, as applicable, and update the status on GeM.
A unified dashboard with real-time status updates and automated notifications could reduce manual follow-ups and improve transparency for buyers and sellers. It could also shorten resolution times, reduce administrative costs, and give sellers greater certainty about cash flow.
How can GeM make seller payments more predictable?
Timely payments are critical for small businesses, particularly those with limited working capital. GeM has defined payment timelines, but the end-to-end journey varies by buyer type and may span multiple systems, stakeholders, and approval processes. Payments by public sector undertaking (PSU) buyers are more time-bound because they use the GeM Pool Account (GPA). This escrow-like account parks buyer funds and includes an 80%/20% auto-release fallback. If a buyer does not initiate payment within 10 days of the issuance of the Consignee Receipt and Acceptance Certificate (CRAC), the system automatically processes and pays 80% of the consignment value.
For government departments, however, the system blocks the budget at the time of order placement without setting aside funds. Once the CRAC is generated, the Drawing and Disbursing Officer (DDO) must process the payment, followed by the Pay and Accounts Officer (PAO). Although the rules set a 10-day payment timeline, no equivalent automatic release mechanism exists when processing is delayed. Budget blocking without fund parking, combined with manual processing, makes payment timing less predictable for sellers. The data reflect this uncertainty. According to the MSME Samadhaan portal’s Age-Category Amount Wise Pendency Report, as of September 2026, micro and small enterprises (MSEs) had filed 28,503 grievance applications involving USD 1.6 billion in delayed payments. Of those applications, 4,083 remained pending, including 3,899 for more than one year.
Which levers can improve payment predictability?
No single fix can make payments predictable. Three levers must work together: Budget and fund availability, rules-based execution, and liquidity bridging. Each requires a distinct digital design choice.
The first lever strengthens fund flows by ensuring that procurement has an available budget and that funds are set aside when payment becomes due.
Across procurement modes, every order should define payment conditions, delivery milestones, and compliance requirements. GeM can translate these conditions into machine-readable rules that govern fund release.
The second lever uses rules-based execution to enforce those conditions. A rules-as-code (RaC) approach could embed them in smart payment systems and shift workflows from manual discretion to automated, milestone-based execution. “If-then-else” logic could define each payment condition, the responsible approver, the verification method, and the turnaround time. The result would be a live, transparent checklist for all stakeholders, with better visibility and traceability throughout the payment lifecycle.
GeM can codify the conditions as standardized rules to ensure consistent compliance and reduce discretion. Integration across GeM, PFMS, and state IFMS platforms could support fund blocking, payment pre-authorization, milestone-based releases, and real-time payment tracking.
The third lever bridges the seller’s liquidity gap by providing working capital against receivables. GeM’s ongoing integration with the Trade Receivables Discounting System (TReDS), which allows sellers to access invoice financing against their orders, moves in this direction. Deeper integration with FinTech firms and private financial platforms could expand financing options. Sellers could then access working capital based on their GeM orders and transaction history.
How can GeM support the wider seller ecosystem?
Participation barriers extend beyond payments. For many small enterprises, challenges begin long before buyers place an order and continue after buyers fulfill it. Sellers often struggle to classify products and prepare catalogs, particularly when they must navigate complex category structures and technical specifications. Logistics and reverse pickups remain expensive and difficult to manage, especially for enterprises in rural and semi-urban areas. Limited working capital further restricts their ability to fulfill larger orders and expand. Women entrepreneurs and first-time sellers often face additional barriers because of information asymmetry and limited digital capabilities.
What would an ecosystem platform provide?
The next generation of procurement platforms can evolve from transaction platforms into ecosystem platforms. Rule-based catalog tools, multilingual interfaces, and assisted onboarding can simplify product listings and lower digital literacy barriers. These features can help first-time sellers navigate categories and procurement pathways and reduce participation costs. Partnerships with logistics providers such as India Post can support affordable last-mile delivery and reverse logistics, which eases the operational burden on small sellers. GeM can also embed access to finance within the procurement ecosystem. Invoice discounting platforms and partnerships with financial institutions can provide working capital, which allows enterprises to finance production without waiting for government payments. Over time, and with user consent, GeM can use transaction histories to build alternative credit profiles for small businesses and expand access to formal finance.
What comes next?
How can GeM turn digital access into sustained participation?
Many of the necessary building blocks already exist. The priority is to connect them through a transparent, interoperable architecture that reduces friction across the seller journey. This integration can make deposits easier to track, payments more predictable, liquidity more accessible, and participation more practical for small enterprises. The third blog in this series will examine the longer-term shifts that global public procurement practices suggest for India.
India’s Gender Budget needs to count the women it reaches
Twenty years after the Gender Budget Statement (GBS) was introduced, India has built a gender budgeting architecture few others have attempted: A mandated annual statement, a classification system grouping schemes by the share of spending estimated to benefit women, and gender budget cells across ministries. The 2026-27 statement reports ₹5.01 lakh crore for women and girls, 9.37% of the Union Budget.
India has institutionalised gender-responsive budgeting (GRB). The next opportunity is to strengthen the data systems beneath it, so gender budgeting captures reach and informs planning. Take G-RAM-G (formerly MGNREGA). The budget share tells us how much is allocated for women. But to understand reach and impact on women, we need to know how many women get work, how many days of work they receive, and what changes for them. The systems largely exist; departments already maintain these records. What is needed is a field that captures sex disaggregation in the data collection format, not a new system.
Nearly eight out of ten rupees in the 2026–27 Gender Budget go to schemes that benefit both women and men. But how much of that money reaches women? For many such schemes, this is not directly known; the share benefiting women is estimated using available information. Think of a ₹10 lakh family car: We know its cost, but not each person’s share of the benefit unless we know who uses it and how often. The government faces a similar problem when schemes serve women and men.
This challenge is not unique to India, several countries implementing gender budgeting at scale face similar challenges. In India, line departments use programme information and prescribed approaches to estimate the share attributable to women when records do not routinely capture sex-disaggregated information. Reliable, sex-disaggregated administrative data is foundational to evidence-based GRB: It can help governments assess who allocations reach, test the basis for attribution, and use the evidence to shape future spending.
This is why the next phase has to move outward from the centre. Gender budgeting may be designed at 30,000 feet, but its ultimate test is on the ground: whether a panchayat can build a girls’ toilet, or a health programme can reach adolescent girls needing screening. Reform happens where the underlying records are created.
And this is not a new idea for government. A JICA-supported project with the Himachal Pradesh government on forest ecosystems and livelihoods, for instance, collects sex-disaggregated data on participation in project activities and other project outcomes. An ADB-supported human capital project with the Meghalaya government similarly collected and analysed sex-disaggregated data to monitor women’s participation and identify implementation gaps. These examples show that the challenge is not collecting gender data from scratch; it is making this kind of data capture routine across government programmes.
Three practical reforms can turn principle into practice.
First, sex-disaggregated data belongs on the source document. Small changes to existing formats, adding fields to muster rolls, recording beneficiaries by sex, or capturing person-days by sex can make a large difference. Linking this data to expenditure can show who programmes reach and how spending benefits women and men.
Second, every estimate should show its basis. A simple field indicating whether a figure is enumerated or estimated would make this visible. It would help departments identify where estimates are weakest, while nodal authorities could target technical support accordingly.
Third, provide technical support where data is generated. GRB training has often focused on gender focal points; the next step is to support the technical, MIS, accounts and programme staff who create and reconcile these records. The support model used in externally aided projects can be adapted for nodal departments and administrative academies.
Better source data would make the GBS more useful for planning. Verified beneficiary data could inform scheme appraisal, identify gaps in reach, and shape future allocations, turning the statement from annual reporting into an input to budget decisions. For policymakers, this means treating routine records as gender-budgeting infrastructure. For departments, it means manageable changes: revised forms, defined indicators, named officers, and sustained support.
The first 20 years of gender budgeting established that women belong in the Union Budget. The next should make the numbers answerable: who benefits, what changes, what impact, and how we know.
This was first published in Hindustan Times on 2nd November 2026.
From access to inclusion: How GeM is reshaping public procurement in India
Can the Government e-Marketplace (GeM) turn digital access into sustained participation for Indian small sellers?
India’s annual public procurement market is estimated at around USD 570 billion. The system has evolved from a manual, centralized model under the Directorate General of Supplies and Disposals (DGS&D) in the 1950s to the Central Public Procurement Portal (CPPP) in 2012, which increased transparency. Yet established vendors who understood complex tendering procedures, paperwork, and bureaucratic processes dominated procurement for decades. The launch of GeM in 2016 marked a further shift. It created a national digital marketplace where government buyers and suppliers could transact directly. Figure 1 traces this journey from DGS&D to GeM 4.0:
Figure 1: Evolution of India’s public procurement landscape
Since its launch, GeM has expanded in transaction value and seller diversity. Its participants now include micro and small enterprises (MSEs), women-led enterprises, and startups.
These numbers show that GeM has done more than digitize procurement: it has widened access to government markets for MSEs, women-led enterprises, startups, self-help groups (SHGs), artisans, and micro-enterprises. Yet access alone does not deliver inclusion. Smaller sellers must also manage deposits and guarantees, list products, fulfill orders, track approvals, and receive payments predictably. These less visible financial and operational frictions determine whether digital access leads to sustained participation. This blog examines them.
How GeM has expanded access
GeM reimagined public procurement as a single, end-to-end digital journey. The process has five stages. Buyers and sellers first register and complete verification. Sellers then list products, and buyers search the “marketplace”. A buyer defines its requirements, chooses a procurement route, and places an order. “Contracting” establishes payment milestones and other financial requirements. After the seller “fulfills” the order under the agreed terms and service levels, the buyer accepts it and proceeds to “payment processing”. Figure 2 provides a high-level overview of these stages, along with the key stakeholders and interactions involved.
Figure 2: Overview of the GeM system and functionalities
A parallel layer of financial flow supports this journey. At various stages, sellers may pay charges or provide financial instruments. These include business cockpit charges for advanced analytics, earnest money deposits (EMDs) to participate in bids, performance bank guarantees (PBGs) after winning a contract, and annual milestone charges linked to sales. Buyers’ main financial obligation is to pay the seller’s invoice after order fulfillment and acceptance.
GeM has addressed several initial barriers. Its New Revenue Policy has reduced transaction costs. The platform has also discontinued caution money, set time-bound payment requirements linked to the Consignee Receipt and Acceptance Certificate (CRAC), and launched inclusion-focused initiatives such as SWAYATT, Startup Runway, and #VocalForLocal outlet stores.
But what happens when small sellers join the platform? A set of less visible financial and operational processes then shapes their experience. Deposits, guarantees, product categorization, fulfillment requirements, buyer approvals, and payment flows determine whether access translates into sustained participation. They also define the next phase of inclusion on GeM.
Where participation still breaks down
MSC (MicroSave Consulting) conducted a dipstick diagnostic study in 2024-25 to assess GeM’s end-to-end digital procurement system. It found challenges beyond onboarding that run throughout the procurement journey. A report by SEWA Bharat identified similar constraints, including onboarding difficulties, government-mandated pricing, and process design that remains inaccessible to small sellers. Building on these findings, MSC’s assessment highlights three areas where GeM can strengthen participation and build seller trust, as depicted in Figure 3:
Figure 3: Three identified areas of improvement
First, sellers continue to face friction in providing deposits and platform charges on GeM.
Small businesses encounter several financial processes on GeM, including transaction charges, annual milestone charges, EMDs, and PBGs. Many still rely on manual workflows, offer limited traceability, and operate across disconnected systems. For an EMD, for example, the seller must upload a scanned copy to GeM and submit a hard copy to the buyer. The financial interaction therefore extends beyond the portal.
Second, payment to sellers still involves multiple approval points.
Timely payments are critical for small businesses. GeM has set payment timelines, but the end-to-end journey still spans several systems, stakeholders, and approval processes. After delivery, payment depends on generating CRACs, obtaining buyer approvals, completing verification, and coordinating across GeM, the Public Financial Management System (PFMS), and state-level integrated financial management system (IFMS) platforms. These sequential steps require substantial manual intervention. Sellers therefore have limited visibility into approval status, who is responsible for the next action, and expected payment timelines. A delay at any stage can create a bottleneck that sellers struggle to track and resolve. The process does not earmark funds when government buyers place orders. Even after successful delivery, sellers must wait for downstream approvals and available funds before payment. For MSEs, these delays increase participation costs and weaken trust in the ecosystem.
Third, broader barriers to inclusive procurement remain.
Digital literacy, internet access, language barriers, and complex GeM onboarding continue to affect entrepreneurs in rural and semi-urban areas. According to the MSME Samadhaan portal’s Age-Category Amount Wise Pendency Report, as of 20th October 2026, MSEs have filed 28,503 delayed-payment applications worth USD 1.65 billion. Of these, 54% involve central government buyers. Unclear product categorization and inconsistent enforcement of procurement quotas can also reduce participation by women-owned and micro-enterprises. Inclusion therefore requires more than opportunities. Enterprises also need the capabilities and support to participate effectively.
Has GeM translated digital access into sustained participation? It has widened the marketplace, but the answer depends on how effectively it removes the frictions that small sellers face across participation and fulfillment through payment.
GeM can take a two-step approach. First, it can make near-term improvements within existing systems and processes to simplify participation and make payments more predictable. Blog 2 in this series examines these practical improvements and the frictions that GeM can address within its existing architecture. The second step is a longer-term shift that extends beyond system-level improvements to address the institutional changes needed to strengthen the broader procurement ecosystem. Blog 3 in this series explores this direction by drawing on global best practices in public procurement and considering future pathways for GeM’s evolution.
