O2O occasion baskets split into three orders

How many shopping trips does a wedding actually take now? For a household in an Indian metro the honest answer is three, and only the last one happens in ten minutes. As of November 14, 2026, that split is the most useful thing to measure in quick commerce. A wedding used to mean one large planned purchase. It now arrives as three separate orders: jewellery and furniture bought weeks ahead at a large-format store, apparel and gifting ordered online before the season, and the things someone forgot — a serum, a stabiliser, a cooler — bought in ten minutes the night before the function. Google and Redseer put the whole market at 13 billion dollars in FY26, on its way to 90 billion by FY31. The more useful number is smaller: non-grocery sits at 3 billion today and is projected at 21 to 27 billion at the end of that run.

Why the basket splits before the market does

Known	Unverified	To verify
Google and Redseer value the market at 13 billion dollars in FY26 and 90 billion by FY31	How much of the FY31 figure assumes non-grocery mix shift rather than grocery volume	The category split behind the headline number
Non-grocery is 3 billion dollars in FY26, projected at 21 to 27 billion by FY31	Whether appliances and cooling are counted inside non-grocery or reported separately	The exact category basket used by each forecaster
Infisum counts 2,525 dark stores at the end of 2025 and models 7,500 by 2030	How many of those stores can hold a large appliance on shelf	Assortment depth per store, by category
Jewellery and furniture remain planned purchases, largely offline	Whether gifting jewellery moves online in the next two seasons	Return rates and authentication rules if it does

Three clocks, three channels

Three purchases, three decision clocks. The first is measured in weeks and needs a showroom, a trial, a conversation with family. The second is measured in days and needs a delivery date that lands before the ceremony. The third is measured in minutes and needs one item, in stock, close enough to arrive tonight. Only the third belongs to quick commerce, and it is the one that decides whether a household remembers the platform or the store.

KPMG's Festive Retail Pulse 2026 makes the same point from the demand side. India's festive opportunity is no longer a single peak, it argues, because spending now moves across regional celebrations, travel, weddings and year-end occasions, each with its own expectations on product quality, value, availability, payment choice and delivery reliability. The report also flags that price sensitivity and fulfilment needs differ by city tier, which is the retail way of saying one assortment cannot serve every market.

That matters before any comparison starts. Teams that measure occasion demand usually begin with a blended number and only later discover that the components were never defined the same way; a shared glossary for gross order value, net order value and orders per store costs nothing and prevents most of the argument. Without it, two departments can report the same week as a record and a disappointment, and both be right.

What makes the split visible is the category mix moving into the last window. Cooling appliances, personal care and household essentials are no longer bought a fortnight early; they are bought after the outfit is sorted and the guests have arrived. Reporting on the shift quotes Tanuj Gupta of Thermocool saying reliable cooling plays an important role here and gives the category a chance to become a larger part of occasion-led household spending. Harshit Aggarwal of Novamax makes the adjacent point: wedding spending is extending into functional appliances families keep using after the celebrations end.

That is a different demand shape from grocery top-ups. A grocery basket is habitual and repeatable; an occasion basket is one-off, urgent and unforgiving. If the serum shade is out of stock at the store two kilometres away, the customer does not substitute — she buys from whoever has it, and the whole occasion relationship moves with her.

Three forecasters, three different markets

Before planning against any of these numbers, it helps to see how far apart they are for the same year. Four published estimates for the Indian market do not agree even on its current size, and the disagreement is definitional rather than arithmetic. One measures gross order value across all categories, another measures a narrower platform-defined market, a third starts from a different base year entirely.

Source	Current size	Long-range estimate	What the scope implies
Google and Redseer	USD 13 billion, FY26	USD 90 billion by FY31, 41 to 48 per cent CAGR	Broadest definition, all categories, gross order value
Infisum Modeling Report	Not stated in the same terms	USD 65 to 70 billion optimistic, 27 to 29 billion conservative, by 2030	Built up from dark-store counts: 2,525 today, 7,500 by 2030
Mordor Intelligence	USD 3.65 billion, 2026	USD 7.39 billion by 2032, 12.41 per cent CAGR	Narrower category coverage, slower assumed adoption
MarkNtel Advisors	USD 7.1 billion, 2026	USD 18.2 billion by 2033, 12.45 per cent CAGR	Mid-range coverage, longer horizon

A 13 billion dollar market and a 3.65 billion dollar market in the same year are not in conflict. They are answering different questions. The wide number counts every rupee of order value that moves through a ten-minute delivery promise, including grocery. The narrow number counts what the researcher defines as the organised quick-commerce category, which excludes a good part of the grocery volume and much of the informal supply.

The practical consequence is blunt: if you benchmark your own growth against a headline figure whose scope you cannot state, you will draw the wrong conclusion in either direction. A brand growing 30 per cent inside a category the researcher says is growing 45 per cent looks like it is losing share. A brand growing 30 per cent inside a segment growing 12 per cent looks like a winner. Same number, opposite reading.

Assortment is the only lever a brand controls

Nobody selling through these channels controls the delivery radius, the rider supply or the platform's fee. What a brand controls is which SKUs sit in which store, and that single decision determines whether the occasion basket is fillable at all. A store carrying eight of your forty SKUs is not a distribution point; it is a lottery ticket.

The pattern across categories is consistent. Fast-moving, single-unit, low-consideration items travel well: a 30 millilitre serum, a phone charger, a pack of balloons, a small room cooler. Items that need a size decision, a demonstration or a return path do not. That is why jewellery and furniture stay where they are, and why the appliance brands quoted in the coverage talk about functional, replaceable, low-ticket units rather than the flagship models.

Depth beats breadth inside a ten-minute window

Category growth in this channel comes from depth, not from width. Adding a tenth category to a store adds a small amount of browsing demand; adding the third shade of the serum that already sells out adds availability exactly where the demand already exists. The measurement follows: report availability per store per SKU for the twenty stores that carry most of your volume, not an average across a network of a thousand.

There is a second-order effect worth watching. When a store's assortment is cut during a reset, the SKUs removed are usually the slow movers by volume — which are often the ones an occasion basket needs. A shopper assembling a wedding-night basket is not buying your bestseller, she is buying the item she forgot. Assortment resets that optimise for velocity quietly remove occasion coverage, and no dashboard reports it.

What to count every week

Occasion demand is invisible in monthly aggregates because it lands in three days and then disappears. The only way to see it is to tag it at the point of measurement, which is a half-hour of setup and an hour a week afterwards.

Field	Example value	Why it matters
Occasion window	2026-11-20 to 2026-11-27	Fixes the denominator before anyone argues about the result
Tagged order share	18 per cent of orders in window	Shows how much of the lift is genuinely occasion-driven
Late-basket SKU set	10 SKUs, fixed for the quarter	Keeps the availability check comparable week to week
All-in-stock store count	62 of 120 stores	The single number that predicts occasion stockouts
Cohort availability gap	New stores 71 per cent, mature 94 per cent	Separates replenishment failure from demand surge

Two of those five steps matter more than the rest. The all-in-stock store count is the one metric that predicts whether an occasion order can be filled at all, because a basket needs every item, not most of them. The cohort gap is the one that tells you whether a miss was demand or plumbing.

FAQ

Is occasion demand big enough to plan around?
Not as a share of annual volume, but as a share of new-customer acquisition it is disproportionate. A household that buys once for a wedding and finds everything in stock is a household you can re-target for the next occasion.
Why do the market-size estimates differ so much?
Because they measure different things. Some count all order value across categories, others count a narrower organised segment or start from a different base year. The scope, not the arithmetic, explains the gap.
Should a brand chase every occasion window?
No. Pick the two windows where your SKUs are genuinely late-basket items and build depth there. Spreading a fixed assortment across six windows produces stockouts in all six.
How many dark stores will exist by 2030?
The Infisum model puts it near 7,500 from 2,525 at the end of 2025. Treat it as a capacity assumption, not a forecast, and check it against the store count your own distributors report.
What is the first thing to fix?
The availability of the ten late-basket SKUs in the stores that carry most of your volume. It is measurable this week and it moves the number that matters.