9. Assumptions and Open Inputs
The catalog row gave a category ("niche data-feed API"), evidence from other companies, and a price range. It didn't give hours per week, location, team, buyer access or a chosen dataset. Here's what I assumed, and how the verdict changes if each assumption is wrong.
9.1 Assumptions made
| Input | Assumption used | If actually different… |
|---|---|---|
| Dataset | India bank, payment-rail and exchange business days (chosen in 01) | If you have privileged access to a dataset (a licence, a data partnership, or proprietary collection no one else has), re-score it. That's the only way a different dataset beats this one. |
| Location | Based in India, can read English government notices and at least one of Hindi or a regional language | Outside India: sourcing and weekend cover get harder, and Indian buyers trust a local vendor more. Multiply the odds by ~0.7. |
| Hours per week | Two columns: ~15 hrs/week and full-time | Under 10 hrs/week: the checking routine alone eats half the time. $1k is still possible; $5k+ roughly halves. |
| Team | Solo developer, comfortable with scraping, APIs and webhooks | A co-founder from lending ops, treasury or a broker's back office is the biggest single boost to sales. |
| Buyer access | None stated | Knowing 30 engineering or ops leads at Indian NBFCs, fintechs or brokers is condition 2. Without it, multiply by ~0.6. |
| Buyer pool | ~2,000–5,000 Indian firms whose code depends on bank or exchange business days | Not verified. If it's under 1,000, $5k+ is out of reach. If it's over 10,000, the ceiling roughly doubles. |
| Price acceptance | Blended ~$58 a month (₹1,999 / ₹4,999 / ₹14,999, plus $149 USD) | If buyers anchor on Calendarific's $12, blended ARPA may fall to ~$30. Customer counts double for every rung. |
| Change frequency | Several material mid-year changes a year across RBI centres, exchanges and states | Not measured. If changes are rare (1–2 a year), the case for a monthly subscription weakens sharply. Count them from the RBI press-release archive before building. |
| Exchange rate | ₹88 = $1 | Small effect on customer counts. |
| Churn | ≤2% a month for regulated buyers, 4–5% for startups | Based on Recurly's 3.15% for $50–100 ARPA, adjusted for embedding. Not a published benchmark for data APIs. |
| Legal position | Holiday dates are facts, and India has no separate database right | Reasoned, not checked by a lawyer. Doesn't change the verdict if you use primary sources only. |
9.2 Inputs that would change the verdict most (please supply)
- Hours per week you can hold for 24 months, and who covers weekends and festival days.
- Location, and the languages you can read government notices in.
- Buyer access: names or companies of engineering or ops leads you can reach at Indian lenders, fintechs, brokers or payroll platforms.
- Any dataset you already have privileged access to. It might beat this pick.
- Your target: is $3–5k MRR, durable and low-upkeep, a good outcome for you? If you need $20k, pick a different idea from the portfolio.
- Any existing product or customers that already consume Indian holiday data.
9.3 Definitions used throughout
- Business day: a day when the named system is open. For banks: not a Sunday, not a 2nd or 4th Saturday, not an RBI NI Act holiday for that centre. For exchanges: per the exchange's trading or settlement calendar for that segment.
- RBI centre: one of the 34 RBI office locations in the NI Act holiday matrix.
- Change: any addition, move or cancellation of a date after the annual list is published.
- Notice-to-publish time: from the official notice appearing online to the change being live in the API and webhooks.
- Tier A / Tier B: as defined in ../../README.md.
- ARPA: average revenue per account per month, after annual discounts.