Quick Commerce Attach Rate: Winning the Second Item

The second item is where the economics change

A quick commerce basket in an Indian metro sits somewhere around Rs 480 to Rs 650 and carries five to eight lines. The platform pays a rider, a picker and a share of dark store rent for that trip whether the basket has three items or nine. Every additional line adds maybe ten to twenty seconds of pick time and almost nothing else, so the marginal item is where the platform’s contribution comes from. It is also where yours comes from, because your fixed costs of getting listed, forecasted and stocked in that store are already sunk.

Yet almost every brand plan we see is written as if the shopper buys exactly one unit. Keyword bids, banners, visibility spend and pack decisions all optimise the first add to cart. The second unit is left to chance. That is the cheapest growth line most quick commerce brands are sitting on, because attach does not need new customers, new stores or new cities.

The three places a second item actually gets added

Attach is not one behaviour. It happens in three distinct moments, and each needs a different asset.

• On the product page, through the similar items and frequently bought together rails. This is where a variant or a flavour gets added. The shopper is still exploring, so a second flavour at the same price works better than a bigger pack.
• In the category listing, when the shopper is scanning a grid. This is where a second brand usually wins, and where adjacency placement matters. If a shopper opens chips, the drink brand that appears there gets the attach, not the drink brand that only bids on drink keywords.
• At the cart, on the add-on rail and against the free delivery threshold. This moment is late, impatient and price sensitive. A Rs 199 to Rs 499 delivery threshold is the strongest attach mechanic on the platform, and the item that fills the gap is almost always cheap, familiar and small.

Most brands have nothing to offer in the third moment. Their smallest pack is Rs 150 and their catalog has no Rs 40 to Rs 80 unit. If a shopper is Rs 60 short of free delivery, the brand is structurally excluded from that decision. Fixing that is a catalog task, not a marketing one.

Adjacency is a bid strategy, not a wish

The reflex is to bid on your own category terms and your own brand name. Those clicks are the most expensive and the least incremental, because a shopper who typed your brand name was going to find you. Attach spend goes somewhere else: on the product pages and search terms of the categories your product travels with.

Read your own basket data first. The platform reports and your category manager can both give you a view of what else appears in orders containing your SKU. Rank the top twenty adjacent categories by co-occurrence. You will usually find three or four that are far ahead of the rest and one surprise. A cold coffee brand finds bread and eggs, not chocolate. A hand wash brand finds dishwash and floor cleaner, not personal care. Point placement spend at the top three and ignore the intuitive ones that the data does not support.

Budget guidance we use: 70 percent of visibility spend on primary category defence and 30 percent on adjacency. Adjacency clicks usually cost 20 to 40 percent less because fewer brands compete there, and they buy incremental units rather than units you would have sold anyway.

Combo SKUs are a catalog decision with a margin file

The tempting shortcut is a combo listing: two units, one code, one discounted price. It works, but only when three conditions hold. The pair must already co-occur organically, so you are accelerating a real behaviour rather than inventing one. The combo must sit at a price point the category actually shops at, usually Rs 199, Rs 299 or Rs 499. And the discount must come out of a defined budget rather than out of the base pack’s price perception.

The operational cost is real. A combo is a new code, a new forecast, a new inventory line at every dark store you serve, and one more thing to go out of stock and drag your availability score down. We would not launch a combo below an expected 300 units a month per city, and we would sunset it if it cannibalises more than 40 percent of base pack volume rather than adding to it.

The other combo trap is margin. Category teams price combos off the summed maximum retail price and then apply the platform margin to the discounted value, which quietly compresses your contribution twice. Build the combo margin file separately before you agree the price, not after the listing goes live.

Measure attach at the basket, not the SKU

Summing SKU sales tells you nothing about attach. The metric is basket level: of all orders containing your brand, what share contain two or more of your units, and what is the average number of your units per order containing your brand. Track both monthly, by city, and by whether the order came through search or through browse.

A single category brand typically starts at 6 to 12 percent multi-unit order share. With a cheap add-on pack, adjacency placement and a working combo, 18 to 22 percent is reachable in two quarters. On a Rs 550 basket that moves your revenue per order from roughly Rs 130 to Rs 165 without a single new customer, and it does so at a much better cost of acquisition than any first item bid.

A six week attach programme

Week one and two, pull the co-occurrence data and identify whether you have a sub Rs 100 unit that can fill a delivery threshold. Week three, launch adjacency placements against the top three co-occurring categories with 30 percent of visibility budget. Week four, fix the product page rails by making sure your variants are correctly linked so the platform can merchandise them together. Week five, launch one combo in one city only. Week six, review basket level attach against the pre-programme baseline and kill whatever did not move it.