The Ekadashi demand swing
Ekadashi occurs twice a lunar month — 24 days per year. In Maharashtra, Gujarat, and parts of North India, non-veg dine-in covers drop 30–55% on Ekadashi. Veg fasting-menu demand rises 40–60%. For a mixed-format casual dining outlet, this is not a rounding error; it is the difference between a profitable day and ₹5k of waste by lunch.
| Signal | Western POS behaviour | What actually happens |
|---|---|---|
| Ekadashi cover drop | Not modelled | Non-veg covers −30 to −55% |
| Fasting-special demand | Not modelled | Sabudana/kuttu/singhara +40 to +60% |
| Post-Ekadashi rebound | Not modelled | Next-day non-veg +15 to +25% |
| Shravan Monday overlap | Not modelled | Compounding drop when Ekadashi falls in Shravan |
Why Western tools miss it
- Gregorian-only calendars — Toast, Square, and most legacy systems have no Hindu calendar taxonomy. Ekadashi is invisible to their forecasting engines.
- Trailing-average extrapolation — they predict tomorrow from the last 4–8 weeks. Ekadashi appears roughly every 15 days; the trailing window dilutes the signal.
- No regional calibration — Ekadashi intensity varies by state. Maharashtra ≠ Tamil Nadu. Western tools ship one global model.
- Dish-level blind spots — even if you manually flag Ekadashi, Western tools don't translate cover drops into dish-level prep quantities for tandoor, cold prep, and fasting specials separately.
What Indian POS tools do — and don't
Petpooja, UrbanPiper, and Posist handle Indian operations well — aggregator integration, GST, multi-outlet. But none of them forecast Ekadashi demand natively. Their analytics show what happened yesterday; they don't flag what Shravan Monday will do in two weeks.
What works instead
- Hindu calendar engine — demand signals for Ekadashi, Shravan, Navratri, and 20+ windows, calibrated per state.
- Dish-level prep forecasts — cover drop translated into SKU quantities, not just a red flag on a calendar.
- 14-day forward horizon — enough lead time to adjust vendor orders and staffing, not just same-day prep.
- Nightly brief, not dashboard — the owner reads a 7pm WhatsApp brief and acts on one or two flags. No daily login habit required.
The cost of ignoring Ekadashi
In the Hindkesari pilot, fasting-day waste dropped from ₹4–6k/day to ₹500–900/day once Ekadashi was modelled before prep. Across 24 Ekadashi days per year, that is ₹80k–1.2L recovered — from one signal Western software doesn't have.