These are the features buyers of calendar and business-day data typically want, each classified with the same honesty as doc 03. Demand doesn't mean you should build it. Most of these demands are already met, or they multiply your upkeep.
Demand signals are inferred from what competitors sell (Calendarific, Holiday API, TradingHours.com, Copp Clark) and from how Indian lending, payments and broking work. They were not measured with buyer interviews. That's the first job in plans/08.
Legend: TS = Table stakes, D = Differentiator, X = Distraction, L = Liability. Priority: P0 = v1, P1 = next, P2 = later, — = don't build.
4.1 Coverage (what dates)
| Feature | Demand signal | Category | Priority | Notes |
|---|
| RBI NI Act holidays for all 34 office centres, mapped to states | High among lenders and payments | D | P0 | The core. RBI publishes it as HTML only. |
| 2nd and 4th Saturday and Sunday rules | High | TS | P0 | Easy, but a classic bug source in home-grown code. |
| RTGS holidays and bank account-closing days | Medium | D | P0 | RBI lists three holiday types: NI Act, RTGS, and account closing. |
| NSE, BSE, MCX trading holidays by segment | High among brokers | TS → D | P0 | python-holidays has NSE. Segment-level clearing and settlement holidays, and MCX evening sessions, are the edge. |
| Settlement holidays and T+1 settlement dates | High among brokers and wealth apps | D | P1 | Where exchange circulars matter. |
| Special sessions (Muhurat trading, Saturday sessions, early closes) | Medium | D | P1 | Announced by circular each year. |
| State public holidays (for payroll and logistics) | Medium | TS | P1 | python-holidays covers these free. Include them so buyers don't need two sources. |
| Tentative dates for next year before official lists land (lunar festivals) | Medium | D / L | P1 | Mark clearly as tentative. Never let a tentative date look confirmed. |
| District-level and election-day holidays | Low-medium, loud when it happens | L / X | P2, on request | Thousands of local notices, many in regional languages. Only for a named paying customer. |
| Other countries | Loud from global buyers | X | — before $3k MRR | New rulebook, new sources, new weekend cover each time. |
4.2 Logic (what the API computes)
| Feature | Category | Priority | Notes |
|---|
| is-business-day for a centre, rail or exchange | D | P0 | The one-line job. |
| next / previous business day | D | P0 | The most-used call, if the design partners are typical. |
| add-business-days (e.g. "5 business days after this date in Chennai") | D | P0 | SLA timers, cure periods, notice periods. |
| Combined calendars ("business day for both Mumbai banks and NSE") | D | P1 | Settlement logic often needs both. |
| Bank cut-off times (RTGS/NEFT windows, NACH presentation windows) | D | P2, after $3k MRR | The natural upsell. Timings change by RBI circular. |
| Pincode → state → RBI centre mapping | TS | P1 | India Post data is free. A helper, never the product. |
| Working-hours / half-day logic | TS | P2 | Rarely needed by lenders. |
4.3 Freshness and trust (why they pay monthly)
| Feature | Category | Priority | Notes |
|---|
| Change webhooks and email alerts | D | P0 | The reason the product isn't a yearly copy-paste. |
Status per date (tentative, confirmed, changed, cancelled) | D | P0 | Lets customers automate safely. |
| Source link per date (RBI press release, exchange circular, gazette) | D | P0 | Trust, and a fast answer when someone asks "says who?". |
| "As known on" history | D | P0 | Audit answer. Cheap to build if every write is versioned from day 1. |
| Public changelog page | D | P0 | Marketing, trust and SEO at once: "22 Jan 2024: markets closed, declared 3 days before". |
| Slack / Teams / WhatsApp alerts | TS | P1 | Ops teams want alerts where they already work. |
| Uptime SLA and status page | TS | P1 | Expected by Enterprise buyers. |
| Accuracy guarantee with financial penalties | L | — | Offer a service-credit SLA on uptime, never on data being right. |
| Feature | Category | Priority | Notes |
|---|
| REST + JSON | TS | P0 | |
| CSV, Excel and iCal export | TS | P0 | Ops and treasury teams live in spreadsheets and calendars. |
| Open-source package (Python, Node, Java) with the confirmed current year | D (as distribution) | P0 | IPinfo's channel was Stack Overflow. Today's equivalent is a package developers install. The live data and history stay paid. |
| SFTP / S3 file drop | D | P1, first Enterprise deal | Banks often can't call external APIs from core systems. |
| Google Sheets add-on | TS | P2 | Cheap reach into ops teams. |
| MCP server for AI agents | TS / D | P2 | Cheap listing, possible discovery channel. Unproven. |
| Usage dashboard with charts | X | — | Usage emails are enough. |
4.5 Adjacent datasets buyers will ask for
| Feature | Category | Priority | Notes |
|---|
| IFSC lookup | X | — | Free open-source datasets exist (Razorpay publishes one on GitHub; not re-checked). |
| RBI reference exchange rates | X | — | Free from RBI. |
| Company / GST lookup | X / L | — | See doc 03. |
| Regulatory filing due-date calendar (GST, TDS, MCA filing deadlines, with extensions) | D? | Test in month 12 | Same "date that moves by notification" shape, different buyer (accounting and compliance SaaS). A possible second product line. Unverified demand. |
| Labour-law holiday entitlements per state | L | — | Wage-law advice. Out. |
4.6 Net effect
Of about 35 demanded features, about 12 are real differentiators, and about 11 belong in v1. The pattern is consistent: every differentiator is about freshness, proof or India depth. Every distraction is about breadth. Stay deep and narrow.