Follow the roles, not only the firm.
FromFinancing firms→ToFinancing network roles
A brand’s journey on an open network passes through several participants, each doing a different job. Finance usually starts with the firm. This shift asks whether it could also start from the roles, so that each participant is understood for the work it actually does.
Five participants, one pack
Meera has designed a millet snack and wants to sell it illustrative. By the time a pack reaches a retailer’s shelf, five participants have touched it. Meera shapes the product. A women’s collective in the hills makes it. Arjun, a stockist in a district town, carries inventory. Farida’s neighbourhood node stores and dispatches. The retailer orders through a buyer app on the network.
When Meera looks for finance, the natural place to start is her company’s accounts: a business two years old, with modest turnover. They describe her firm accurately. They weren’t designed to describe the five roles that move her product.
From a firm to a set of roles
One firm, one balance sheet
Several roles, each carrying part of the journey
| Today | On an open network |
|---|---|
| The firm | The role each participant performs |
| Consolidated accounts | Activity within each role |
| One balance sheet | Several participants, each carrying part of the journey |
An open network unbundles the value chain into roles. Finance can follow the unbundling.
How it would work
A role registry. The network defines the finance-relevant roles at protocol level, and participants can be found by role. Each role carries a different kind of exposure:
| Role | The exposure it carries |
|---|---|
| Opportunity creation | Capability |
| Experimentation | Product and market development |
| Inventory selection and carrying | Inventory |
| Productive asset operation | Productive assets |
| Trade execution | Verified transactions |
| Collection | Receivables and settlement |
| Portfolio formation | Aggregated commercial risk |
A participant can hold more than one role. Arjun can carry inventory and also collect from retailers.
Go deeper
What already exists
Open commerce networks already define roles. The Government’s description of ONDC lists buyer applications, seller applications, logistics providers and technology enablers[1]. Finance already recognises that the relationship between a buyer and a supplier can itself be financed: on TReDS, an MSME seller can be paid early against an invoice its buyer has accepted[2]. This shift carries role definitions into finance.
Scope. We haven’t checked whether the roles ONDC defines are the right ones for finance, or how a registry of finance-relevant roles would sit alongside ONDC’s own. We found no evidence yet that role-level underwriting beats firm-level underwriting. That is the pilot’s question.
Rules and privacy
Roles describe the work a participant does, not their permission to lend. The Reserve Bank’s digital-lending rules apply to banks, NBFCs and other regulated entities, and a lender stays fully responsible for any agent acting for it[3]. A registry would record roles, and regulated entities would still do the lending.
For a sole proprietor, a record of what they do is personal data, held with consent and removable[4].
Getting it right
Keep roles few and plain. Seven are enough to start, and a participant can hold more than one.
Describe, don’t exclude. A role is something a participant does, not something it is. Newcomers enter by doing the work.
Take roles from what the network already records, so that no one has to fill in another form.
Who it helps. Meera is seen for the work her brand actually does, Arjun for the stock he carries and the collective for the production it runs. A lender can see where each exposure sits before deciding whether it fits.
What could go wrong
- Roles that don’t match reality.Real participants wear several hats, so roles describe activity and one participant can hold several.
- A registry that becomes a gate.If only registered roles can be financed, newcomers are shut out, so entry stays open and the registry records what already exists.
- Role-level underwriting no better than firm-level.The pilot compares the two on the same transactions.
- More labels, more cost.If recording roles is a burden, small participants won’t do it, so roles are taken from what the network already records.
How we’d test it
Take a sample of network transactions in one category and region, and tag each by the roles involved.
- The test
- Role-level view
- What we would measure
- Compare performance and losses by role with what a firm-level view would have shown for the same transactions. Keep the roles that make a difference and drop the rest.
- What would end the idea
- Performance and losses look the same across roles, so roles add nothing to the firm-level view.
Sources
- [1]Ministry of Commerce & Industry, Revolutionizing Digital Commerce: The ONDC Initiative (4 January 2025)Checked; a government document
- [2]Reserve Bank of India, Reserve Bank of India (Trade Receivables Discounting System) Directions, 2026 (RBI/DPSS/2026-27/406, 23 June 2026), paragraphs 3 and 15Checked; the Reserve Bank’s text
- [3]Reserve Bank of India, Reserve Bank of India (Digital Lending) Directions, 2025 (RBI/2025-26/36, 8 May 2025), paragraphs 3 and 5Checked; the Reserve Bank’s text
- [4]PRS Legislative Research, The Digital Personal Data Protection Bill, 2023Checked; a summary of the Bill as passed by Parliament in August 2023
Spotted a source we missed or a rule we misread? Tell us what we got wrong.
If each role is different, is the risk inside each role different too, and does it stay the same as the transaction moves?