Zepto ads pay back when the order they win is the first of several: a product the shopper replenishes, sticks with once it works and can find in stock next time. For impulse buys, price-led categories or patchily stocked SKUs, the same spend buys one order at a loss and the slot goes to whoever bids next week. Sort your SKUs before setting a budget, and judge the campaign on whether new buyers come back, not on dashboard ROAS.
Why the Second Order Is the One That Matters
A sponsored tile on Zepto puts your product in front of a shopper who is assembling a small basket in well under a minute, usually from a category page or a short search. The screen holds only a handful of products before the shopper scrolls or moves on, so the slot is genuinely valuable. It is very good at one thing: winning the first order from someone who was about to buy something in your aisle anyway.
The trouble is that the first order on its own rarely pays for the ad. Once the platform's commission, fees and whatever promotion is running alongside the ad are taken out, the margin left on a single quick commerce order is thin, and for many D2C brands it is thinner than they assume. If you have not done that arithmetic yet, our breakdown of working out whether quick commerce volume is actually worth it is the place to start. For most brands it means the ad is only profitable if the shopper it brings in comes back.
That return visit has a specific shape on a ten-minute app. Shoppers lean heavily on their order history and reorder shortcuts, because rebuilding the same weekly basket from scratch is the opposite of why they opened the app. Once your product sits in a shopper's past orders, it can be added again without anyone bidding for it. That is the real asset a Zepto ad buys: a place in the shopper's own reorder list. If the product never earns that place, every future order has to be bought again at auction.
Where Sponsored Placement Pays Back
The products where Zepto ads earn their money tend to share three traits. The first is a natural replenishment rhythm. Staples, personal care, baby care, pet food, breakfast and cleaning products run out on a cycle, so a shopper who tries you once has a reason to need you again within weeks. A product bought once a year has no such reason, however good the first experience was.
The second is a cost to switching once the product works. A shampoo that suits someone's scalp, a cereal a child has decided they like, or a pet food that a dog's stomach tolerates all build loyalty that has nothing to do with the next ad. When the shopper's own experience is doing the retaining, the cost of acquiring them can be spread across many orders. When any brand in the aisle would do just as well, the next order goes to whoever is in the first slot that day.
The third is that the price holds. If your product sells at its normal price while the ad is running, the buyer it brings in looks like the buyer you will see in later months. If the ad only converts when a deep discount is stacked on top, you have acquired a discount buyer, and that buyer's second order depends on the discount coming back. Before putting money behind a SKU, it is also worth knowing what a returning customer is genuinely worth to you, which is the subject of our guide to calculating customer lifetime value before you set an acquisition ceiling.
Where It Just Rents Shelf Space
Renting shelf space means paying for every order and owning nothing afterwards. It is not always a mistake, but it should be a deliberate choice, and most brands drift into it without noticing. The clearest case is impulse and occasion buying: a gift box for a festival, a party snack, a flavour bought on a hot evening. The ad can win that order, but nothing pulls the shopper back to your brand next time. They will buy whatever is sitting in the first slot, which means whoever paid for it.
| Pays back on Zepto | Just rents shelf space |
|---|---|
| Replenished every few weeks, so a second order is likely | Bought once, on impulse or for an occasion |
| Shopper sticks with a brand once it works | Shopper switches to whatever is cheapest that day |
| Price holds without a deep discount alongside | Sales only move while a discount is running |
| Stocked across most stores in the cluster | Ad serves where stock is patchy or missing |
The second case is a category where shoppers pick on price. If the decision in your aisle comes down to whichever pack is cheapest today, the ad wins a sale only while you are the cheapest option on screen. A competitor's promotion next week undoes the acquisition completely, and you are back to bidding for the same shopper.
The third case is less obvious: a brand that already leads its aisle organically. If you are the product shoppers reach for and you already appear near the top of the category without paying, a sponsored slot on your own name or aisle may mostly buy orders you would have received anyway. Some defensive spend is reasonable when competitors are bidding directly on your brand name. It protects the shelf, and it should be budgeted as protection rather than reported as growth. There is also the availability problem underneath all three cases, covered in detail in why availability rather than bidding decides whether quick commerce ad money works. An ad served in a store that does not hold your SKU rents a slot you cannot even use.
"An ad that wins a place in the reorder list is an investment. An ad that has to win the same shopper again next week is rent, and the rent is set by whoever bids against you."
- Brand Integer Quick Commerce Team
Reading a Zepto Campaign by Cohort, Not by ROAS
The return on ad spend figure in the dashboard credits the ad with every order that followed a click or a view. On a small screen where your brand already appears organically, a large share of those orders would have happened without the ad. ROAS tells you that the ad was near the sale. It does not tell you the ad caused the sale, and it says nothing at all about whether the buyer will come back, which is the question that decides whether the spend was an investment.
A more honest read is to follow the buyers rather than the clicks. Fund one city or store cluster where you are reliably stocked, leave a comparable cluster unadvertised as a baseline, and watch what happens over the following one to two months, not the following week. Ask your account manager what new-to-brand and repeat reporting is available for your account, because what the platform shares varies. Where cohort data is thin, the sales trend in the advertised cluster against the held-back one is a workable proxy.
- Set a ceiling Max cost per first order from margin and expected repeats
- Pick one cluster Fund a single city or store cluster you are stocked in
- Hold one back Leave a comparable cluster unadvertised as a baseline
- Track the cohort Watch new buyers for repeat orders over 30 to 60 days
- Decide Scale, reshape or stop based on the second order
The decision rule that follows is simple. If the advertised cluster's sales stay above the baseline after the ads pause, the spend built something: shoppers found you, kept you in their reorder list and came back. If sales fall straight back to the baseline the week the ads stop, the spend was rent. Both results are useful, because the second one tells you exactly which SKUs not to advertise next quarter.
Setting the Budget Before the Bid
Most Zepto budgets are set by asking what competitors seem to be spending. A better starting point is a ceiling you write down before the first campaign goes live: the most you can afford to pay for a first order, given the margin on each order and a conservative estimate of how many more orders a new buyer places. Be cautious with the repeat estimate. It is far easier to raise a ceiling that proved too low than to recover money spent against an optimistic one.
Then allocate by SKU rather than by brand. Put the media behind the product that has the replenishment rhythm, the stable price and the pack size that suits a top-up basket, rather than spreading it across the whole range because every product deserves visibility. One SKU that earns repeat buyers in one cluster is a stronger foundation than a thin spend across thirty products in six cities.
Finally, treat Zepto as its own test rather than a copy of what worked on Blinkit or Instamart. The ad mechanics look similar across the platforms, but the store footprint and shopper base differ from city to city, so a result on one app is a hypothesis for the next, not a prediction. Run the same cluster test on each platform and let each one earn its budget separately.
Frequently Asked Questions
Are Zepto ads worth it for a new D2C brand?
Usually only once the product is listed, consistently stocked in at least one cluster and selling some units without help. Before that, ads tend to serve in stores where you are out of stock or buy trial orders with no reorder base behind them. Start with one replenishment SKU in one cluster and expand from there.
What is a good ROAS for Zepto ads?
There is no universal number, because platform ROAS includes orders that would have happened without the ad. A more useful test is whether buyers acquired through the ads reorder, and whether sales in an advertised cluster stay above a held-back cluster after the ads pause, judged against a cost ceiling worked out from your own margin.
Should we bid on our own brand name on Zepto?
A small amount of defensive spend makes sense if competitors appear on searches for your brand. Keep it modest, test it by pausing for a period, and report it as protection rather than growth, since most of those shoppers were already looking for you.
How is advertising on Zepto different from Blinkit?
The mechanics are broadly similar: sponsored listings in search and category pages plus negotiated high-visibility placements. The difference is in store footprint and shopper base, which vary by city, so run a separate cluster test on each platform instead of copying a budget across.
