FeedWise

Verdict

No as described. Go small, as an India business-day API.

A generic "niche data-feed API" (IP location, exchange rates, public holidays or company data) has a ~2% chance of reaching $10k MRR in 24 months, even full-time. Each of those datasets is either free from an official or open-source source, or owned by an incumbent with a decade of data. Narrowed to Indian bank, payment-rail and exchange business days, sold on freshness and audit history, the odds rise to ~5% at ~15 hrs/week and ~10% full-time for $10k, and ~35% / ~50% for $1k. It isn't fundable, and it doesn't need to be. It's a small, durable data business with a low ceiling, and the ladder in WHAT-TO-DO-AND-AVOID.md says so plainly.

The five things that matter

1. The catalog's evidence is real but doesn't transfer. IPinfo went from a weekend project in 2014 to ~$1M ARR in 2017 and multi-million, profitable, 15 staff and 100,000+ customers later (10). But it won with hard-to-build data, a global market, and Stack Overflow traffic in 2014–16. None of those three is available to a new entrant in 2026.
2. Most data feeds are now free. python-holidays (MIT licence) covers 250 countries with states. Nager.Date covers 210, free. ECB and RBI exchange rates are free. MaxMind has a free IP database. Indian company master data is on data.gov.in. A dataset that someone publishes free, in a machine-readable form, is not a business.
3. India bank holidays are the exception we found. RBI publishes its Negotiable Instruments Act holidays as an HTML matrix for 34 office centres, with no CSV or API. python-holidays covers India's states but not RBI bank holidays. Dates change at short notice: markets and money markets closed on 22 Jan 2024 after a holiday declared days before, and RBI kept banks open on 31 Mar 2025 despite Eid.
4. The pain is real but moderate. NPCI shifts a NACH debit that falls on a holiday to the next working day automatically, so the worst failure is handled for lenders already. What's left is: wrong due-date messages, settlement and treasury dates, SLA timers, collections schedules and audit questions. Pain score: 5–6/10. Buyers will compare you to "an intern copies the RBI page each January".
5. The ceiling is the buyer pool, not the margin. Gross margin is ~90% past $2k MRR. But $20k MRR needs ~345 Indian firms paying ~$58 a month. The number of firms with the need and an engineering team is estimated at 2,000–5,000 (not verified). $1k–3k is realistic. $5k is a good outcome. $10k is a stretch. $20k is not a plan.

Feature sort (headline)

• Table stakes: a holiday list by date range, REST + JSON, a free tier, docs with copy-paste examples, national holidays.
• Real differentiators: RBI centre-level bank holidays, NSE/BSE/MCX settlement days, business-day arithmetic, change webhooks within 4 hours, source links on every date, "as known on" history.
• Distractions: more countries, IP location, FX, company data, district-level holidays, dashboards.
• Liabilities: labour-law holiday entitlements for payroll, "compliance" claims, copying other compilers' lists, portal-scraped company or GST data.

Most likely cause of death

Nobody pays. Developers copy the RBI page once a year into a config file and accept the risk. Early warning: fewer than 4 of 30 buyer calls describe a real incident caused by a wrong holiday in the last 12 months.

Next 90 days

Do: build the 2026–27 calendar for all 34 RBI centres plus NSE, BSE and MCX by hand, with a source on every date. Sell it as a ₹1,999-a-month file plus change alerts to 5 design partners by day 45. If they won't pay, stop. Stop: considering IP location, exchange rates, global holidays or company data. They are someone else's free product.