Kestrel
All writing
Engineering28 Mar 2026 · 8 min read

Data licensing for NBFCs and banks: what to ask for

A practical checklist for an NBFC negotiating a B2B data license. Coverage, freshness, DPA, indemnity, audit rights.

NBFCs and banks are the largest buyers of bulk India company data, mostly for MSME underwriting. The market for this is opaque: vendors quote a single line-item price for "India company database", customers do not know what they are paying for. A practical checklist follows.

Coverage: ask for a coverage statement (% of MCA active companies, % of GST-active GSTINs, % of IEC holders, % of Udyam MSMEs). Vendors who cannot answer in numbers are reselling a partial dataset they did not source themselves.

Freshness: ask for the median age, in days, of the records, by source. A median above 90 days is stale. A median above 30 days is workable. A median below 14 days is excellent.

Provenance: ask for per-field provenance in the schema. If the schema is a flat firmographic file with no source attribution, you cannot defend the data in an audit.

DPA: insist on a Data Processing Agreement that specifies sub-processors, retention, breach notification, and DPDP rights handling. Many vendors will not have one.

Indemnity: ask for indemnity against DPDP-derived liability where the data sourcing was non-compliant. If the vendor refuses, that is a signal about sourcing.

Audit rights: ask for the right to audit the vendor's sourcing once a year. If the vendor is sourcing lawfully, this is a non-issue.

Refresh model: ask whether refreshes are pushed (delta files) or pulled (re-download). Push is cheaper to operate and easier to integrate.

Pricing model: ask for per-company-month pricing, not flat-fee. A flat fee disincentivizes the vendor from improving coverage. Per-company-month gives them an incentive to grow.

A vendor who answers these eight questions clearly is doing the job. A vendor who waves the questions away is not.

Kestrel is the India-first GTM data engine. Search 1.89 million active companies, track 15 buying-signal types, and call the public enrichment API.

Try Kestrel free

More writing.