8. If I Were You: The Next 90 Days
The single most important thing
Get 5 Indian firms paying ₹1,999 a month for a hand-checked business-day file and change alerts by day 45, before building the API.
- Segment: engineering, treasury-ops or collections leads at NBFCs with in-house tech, lending platforms, payment firms and brokers.
- Job: "Tell me every bank and market holiday for every centre and exchange I care about, and tell me within hours when one changes."
- How: a concierge version. A JSON and CSV file per customer, refreshed daily from the sources you check by hand, plus an email or WhatsApp alert for every change. Charge from week 3. Automate only what you've done by hand at least 20 times.
Week by week
Hours are shown for the ~15 hrs/week path. Full-time, compress weeks 1–6 into 3 and use the extra time for sales calls.
| Weeks | Goal | Exit criteria |
|---|---|---|
| 1–2 | Discovery and the dataset. Book 30 calls with target buyers. Ask: "Where does your holiday calendar come from? When did it last break? What did it cost?" In parallel, build the 2026–27 calendar for all 34 RBI centres plus NSE, BSE and MCX by hand, with a source link on every date. Count the mid-year changes in the last 3 years from the RBI press-release archive and exchange circulars. | 15+ calls done. At least 4 can name a real incident. The dataset is complete, and you know how many changes a year there really are. |
| 3–4 | Concierge offer. Sell the file plus alerts at ₹1,999 a month (founding price locked for 12 months). Set up source-watching scripts that email you when a watched page changes. | First 2–3 paying. The watching scripts cover every source. |
| 5–6 | Kill test. Keep selling. Deliver alerts by hand. | Day 45: 5 paying (full-time) or 4 (~15 hrs/week). Miss: try brokers and wealth apps instead of lenders for 2 weeks. Miss again: stop. |
| 7–9 | Build the thin v1: versioned date table, is-business-day, next/previous, add-business-days, webhooks, API keys, Razorpay billing, a status page and a public changelog. | Design partners moved to the API. Webhooks set up by ≥60% of them. |
| 10–11 | Free hook. Publish the open-source package (Python and Node first, Java next) with the confirmed current year. List on PyPI, npm and one API marketplace. Write 3 changelog posts about past changes (22 Jan 2024 markets closure, 31 Mar 2025 banks kept open). | Package live. 20+ free keys. |
| 12–13 | Measure and decide. | 15 paying (full-time) or 10 (~15 hrs/week), zero missed changes, median notice-to-publish ≤4 hours. Hit: continue to the roadmap. Miss: go back to discovery with a different segment, once. |
Kill criteria (write these down now)
- Day 45: fewer than 4 paying design partners, after trying two segments.
- Day 90: fewer than 8 paying customers, or any customer finding a change before you did twice.
- Before building anything: if the 3-year count shows fewer than 3 mid-year changes a year across RBI centres and the exchanges, the subscription case is weak. Sell a yearly file at a one-off price instead, or pick another idea.
The response to a miss is a different buyer, not more datasets.
The one thing to stop doing
Stop weighing IP location, exchange rates, company data and global holidays.
They are someone else's free product. IP location needs a probe network and a decade of corrections. Exchange rates are free from central banks. Global holidays are in a free, MIT-licensed library covering 250 countries. Indian company data is served by funded KYC vendors and $2.50-per-1,000 scrapers. Each hour spent comparing them is an hour not spent finding out whether 5 Indian firms will pay for the one dataset that isn't free.
A close second: don't build a portal before a payment. The design partners need a file and an alert.
What stays on the long-term roadmap (only after proof)
- Settlement dates and special sessions for brokers.
- Bank cut-off times (RTGS/NEFT, NACH windows), after $3k MRR.
- SFTP file drop and audit export for the first Enterprise customer.
- A USD plan for global firms with Indian operations.
- Then maybe a second calendar with the same shape: tax and filing due dates that move by notification. Test demand first.