How a shopper finds your pack
Zepto's engineering team wrote about its search in June 2026. It is the clearest official look we found at how an app picks what to show. Other apps may differ, so treat it as a guide.
- The results are for one place. A search is answered from what the delivery hub near that address has in stock right now.
- The app splits a search into brand, product type, size and other details. A shopper who types all three is read as three parts.
- Ranking blends clicks, add to cart, price, availability and what this shopper did before.
- New products get a trial. Zepto says it keeps a stream for products that have too little engagement to rise on their own.
- A product that is out of stock stays visible for some days, marked as unavailable, so regular buyers know to check back.
- Ads sit on the same page. Zepto says its rules for ad relevance are looser than for ordinary results, but quality floors still apply, whatever the bid.
What this means for you: your pack name, size and photo must match the words shoppers type. Zepto's own post tells of a chocolate brand that missed shoppers who typed 'meetha' and 'mithai'. Use the words your shoppers use, in the language they use.
Step 1: Fix the pack and the listing first
Ads send shoppers to your listing. A weak listing wastes the money.
- Name: brand, product, variant and size, in the order shoppers say them.
- Photos: the front of the pack, large and clear. Then the back with the label, and one that shows the real size.
- Claims: the Food Safety and Standards Authority of India (FSSAI) told online food sellers on 3 December 2024 that product claims online must match the label on the pack. A claim that is not on the pack should not be online.
- Shelf life: the same advisory says food must have a minimum shelf life of 30% or at least 45 days before expiry at the time of delivery. Ask a food-law adviser what that means for your product.
- Sizes: put the sizes of one product together, so a shopper can compare them.
The listing steps themselves are in our selling guide.
Step 2: Be in stock where you advertise
An ad for a pack that is not on the shelf in that area is money lost. Swiggy's management said on 30 July 2026 that it had made a hard reset on availability, as one of three areas of shopper experience it was investing in. The apps care about this a lot.
- Start in the cities and areas where you can keep stock steady.
- Ask the category team how often stores restock, and share your forecast.
- Ask whether you can pause ads for a pack, or an area, when stock runs out.
- Watch expiry. Eternal's July 2026 letter puts Blinkit's losses from expiry, damage, shrinkage and theft at about 1.8% of its net order value. An app that owns the stock carries that loss, so expect questions about packs that sell slowly.
Step 3: Choose your ads
Each app sells space in more than one place. Think of three jobs.
- Sponsored search: your product shows when a shopper types a word you picked. Zepto says most ad budgets go to search ads, because a search shows intent to buy.
- Banners and brand pages: they build awareness for a launch or a new pack. Zepto calls these display ads.
- Reach outside the app: Blinkit's MediaMax page says it runs Meta and Google ads using Blinkit's first-party data. That means data Blinkit collects itself.
Zepto describes a loop. A sponsored listing wins views. Sales rise. Better sales lift the pack in the ordinary results. More shoppers then search for the brand by name. That is the app's view, so test it with your own numbers.
Start small. Pick your best-selling pack and one city. Bid on your brand name, the product name and the plain category word. Then add the regional words your buyers use.
Zepto's tool offers exact, phrase and broad match. These run from narrow to wide. Begin narrow, read the search terms report, then widen.
You can set one bid for the whole campaign, which Zepto tunes towards return on ad spend (ROAS). ROAS is the sales you get for each rupee of ads. Or you can set a bid for each keyword. Zepto's engineering post also lists two rules for its ad system. A shopper is charged only for the first click on an ad. A campaign stops when its spend limit is reached.
Ask each app how you are charged, per click or per thousand views, before you spend. Ask for it in writing.
Step 4: Work out price, margin and offers
The apps earn from your margin, from ads and from shopper fees. Swiggy's management said on 30 July 2026 that its earnings per order rose because of brand margins it negotiated, advertising and user fees. It added that many brand talks happen once a year.
So ad spend comes on top of a margin you have already given. Do the sum first.
Say a snack pack sells for ₹100 on the app. After the app's share and fees, the brand keeps ₹55. Making and packing it costs ₹40. That leaves ₹15. If ads cost ₹6 for each pack sold, ₹9 is left. If ads cost ₹15, you earn nothing on that sale. These numbers are made up. Use your own.
Plan offers you can repeat. From 1 January 2027, when a price cut is announced on an e-commerce platform, the earlier price must be shown with the new one. The earlier price means the lowest price in the 30 days before the announcement. The Press Information Bureau (PIB) explained this on 10 September 2026.
Think about cart goals too. Zepto's October 2026 post shows a nudge such as 'add ₹59 more to get ₹75 off'. Pack sizes and prices that fit such gaps are easier for a shopper to add. Ask the team which goals run in your city.
Expect company: Swiggy's management also spoke of its own brand and of packs made with other brands. Ask where your pack will sit, and win on the pack, the proof and the trust.
Step 5: Check the rules for ads and claims
- From 1 January 2027, sponsored listings must be identified by clear and prominent disclosures. Never try to dress an ad as an ordinary result.
- The same amended rules bar a platform from misleading users by manipulating search results.
- The Central Consumer Protection Authority (CCPA) has issued guidelines on misleading advertisements. The government says they cover advertisers and agencies, including on e-commerce platforms.
- Keep every claim in your ad and your listing provable. If it is not on the pack or in a paper you hold, leave it out.
This is information, not legal advice. Ask a lawyer or a food-law adviser about your pack.
What to measure
- Cost per order from ads: ad spend divided by the orders the ads brought.
- Your place in search: each week, search your top ten words in your city. Note whether you show as an ad, as an ordinary result, or not at all.
- Availability: the share of days and hours your pack was in stock in each area.
- Brand searches: ask whether the tool reports searches for your brand name. If they rise, the loop above may be working.
- Repeat orders, and any ratings or complaints the app shows you.
- Margin per pack after the app's share, fees and ads.
Zepto says its reports split ad spend by campaign, search term, category, placement and device. Use what the app gives you, but keep your own sheet by month and by city.
Mistakes we see
- Buying ads before the listing, photos and pack name are right.
- Running ads in areas where the pack is often out of stock.
- Judging the campaign by clicks instead of cost per order.
- Giving a deep discount that you cannot repeat, then losing sales when it ends.
- Never reading the search terms report, so wasted words keep running.
- Using claims on the listing that the label does not carry.
When to ask for help
Ask for help when ad spend grows beyond what you can read yourself, or when you add cities and packs. We plan the listing, the ads and the tracking as one job. See our quick commerce marketing service, and the page for fast-moving consumer goods (FMCG) brands on quick commerce. For wider reading on how apps and ads fit a direct-to-consumer brand, see what is D2C (direct to consumer).
