Blinkit Ads: Buying Visibility Inside a Ten-Minute Purchase Decision

• →Quick commerce ads buy position in a browse journey, not intent in a search journey. The playbook from Amazon does not transfer cleanly.
• →Availability is the real bid multiplier. Spend behind a SKU that is out of stock in half the relevant dark stores subsidises your competitors.
• →Concentrate the budget in one city cluster before going national. A thin national spend produces impressions everywhere and learning nowhere.
• →Most of the traffic is browsing an aisle, so category and subcategory placements usually matter more than your own brand term.
• →The high-visibility banners are largely negotiated with the category team, not bought self-serve. Both routes need to be running.

What You Are Actually Buying

A sponsored slot on Blinkit, Instamart or Zepto is not a search ad, even though the interface borrows the vocabulary. On a marketplace, somebody types a query because they want to compare, then reads a listing page before deciding. In a ten-minute delivery app the shopper is assembling a small top-up basket, usually from category tiles and reorder suggestions, and the whole decision happens on a thumbnail. What you are buying is position inside that browse, for the moment the basket is open.

That changes what the ad has to do. There is no listing page carrying the argument, no review wall, no A+ module. The pack tile has to say what the product is and which variant it is at a glance, and the price has to be legible next to whatever is sitting beside it. If the tile does not resolve instantly the impression is spent. This is why brands with strong marketplace listings sometimes see quick commerce ads underperform badly: everything that was doing the convincing on Amazon is missing here, and the ad is exposing that rather than fixing it.

The inventory itself also splits into two very different things, and brands routinely engage with only one. The self-serve ad products cover placements inside search results and category listings, and a media buyer can run those on their own. The high-visibility merchandising, the front-of-app banners, curated collections and occasion carousels, is largely controlled by the platform's category team and gets negotiated rather than bought, usually on the back of a co-funded promotion or a genuinely new format. The mechanics of that conversation, and what the platform actually wants in exchange for a slot, are covered in more detail in how visibility on Blinkit and Zepto is actually decided at the dark store level.

Think of the self-serve side as always-on and the negotiated side as campaign-led. A brand running only banners has bursts of visibility with nothing in between, and a brand running only self-serve placements never reaches the shoppers who are not already browsing its category. The two do different jobs and the budget should acknowledge that rather than treating them as alternatives.

A marketplace search ad and a quick commerce ad slot are not the same product

Amazon sponsored products	Blinkit and Zepto ad slots
Shopper arrives with a query and compares options	Shopper is topping up a basket in under a minute
A long tail of keywords to bid against	Few searches, mostly category and aisle browsing
Listing copy, images and reviews do the convincing	The pack tile and being in stock do the convincing
One national campaign can serve the whole country	The ad only works in stores that actually hold your SKU

Availability Is the Real Bid

The single largest difference between advertising here and advertising on a marketplace is that your ad is only serviceable where your stock physically is. A dark store holds a few thousand SKUs, each store's assortment is decided locally, and a shopper in a pin code served by a store that does not carry your product cannot buy it at any bid. So the first question before any campaign is not what to bid, it is which stores actually hold the SKU today, and what share of the demand in your target city that covers.

This is where a lot of quick commerce media gets quietly wasted. A campaign set to run across a whole city will happily serve impressions to shoppers whose nearest store stocked out two days ago, and the platform will charge for them. Worse, a shopper who taps a promoted tile and finds it unavailable learns something about your brand that no later ad undoes. Stock health and ad spend need to be looked at on the same screen, weekly at minimum, and the campaign should be pulled back to the stores that can actually fulfil rather than left running on hope.

"On a marketplace, being out of stock costs you the sale. In quick commerce, being out of stock costs you the sale and the ad spend that went looking for it."
- Brand Integer Quick Commerce Team

There is a second availability question underneath the first, which is whether the pack you are advertising is the right one for this basket. The formats that sell in a ten-minute delivery are usually smaller, single-occasion or top-up sizes, and putting media behind your marketplace hero pack often means paying to promote the version of the product this shopper was never going to add. That decision is made before the campaign brief, and the way to think it through is set out in why the pack that sells on Amazon is usually the wrong SKU for a dark store.

It is also worth being honest about the direction of causality. Advertising can accelerate velocity in a store that already stocks you, and sustained velocity is what earns you space in more stores. What advertising cannot do is get you listed somewhere you are not, or keep you listed somewhere you keep selling out. The spend is an accelerant on an operational base. Where the base is missing, it is just an expensive way to generate impressions.

Structuring the Budget by Cluster, Not by Country

The default instinct is to set a national daily budget and let the platform distribute it. In quick commerce that is close to the worst possible structure, because it thins the money across hundreds of stores with wildly different stock positions and produces a blended number that hides everything you need to know. The better shape is deliberately narrow: pick the city or the cluster of stores where your per-store sales are already strongest, fund it at a level that can hold a position for several weeks, and treat everywhere else as not yet advertised.

Narrow spend also makes the results readable. Inside one cluster you can see whether promoted impressions turned into repeat orders, whether the pack held its price against whatever the platform was discounting that week, and whether new buyers came back without the ad. None of that is visible in a national average. Once a cluster is working, the growth move is to add the next cluster rather than to raise the national bid, because you are buying distribution and velocity rather than auction position.

Within the cluster, weight the placements toward how people actually move through the app. Category and subcategory placements reach shoppers who are browsing an aisle with no brand in mind, which is the majority of this traffic and the only place net new buyers come from. Your own brand term is cheap and converts well, and it is largely defensive: worth holding so a competitor does not sit on top of your name, not worth mistaking for growth. If you are used to building out long keyword structures on Amazon, that habit needs adjusting, and the contrast with a properly structured marketplace account is worth reading alongside this in how to structure Amazon PPC campaigns so the data stays readable.

Finally, budget the promotion separately from the media. Much of what makes a quick commerce campaign work is a price or bundle mechanic the platform is willing to co-fund, and the margin cost of that mechanic is a real part of the campaign cost even though it never appears in the ad account. Brands that plan only the media number end up funding the discount out of contribution margin they had allocated elsewhere, then conclude the ads were unprofitable when what was unprofitable was the promotion attached to them.

1. Fix availability before you bid An ad pointing at a SKU that is out of stock in half the relevant stores is paid for and wasted
2. Concentrate on one city cluster Put the budget where you already have velocity instead of thinning it across the country
3. Buy the category, not your brand name Most of this traffic is browsing an aisle, not typing a brand into search
4. Read the numbers store by store A blended national return hides the stores where the ad had nothing to sell
5. Scale by adding stores, not bids Earn the next cluster with proven per-store sales rather than a higher national bid

Measuring It Without Fooling Yourself

Return on ad spend inside the platform dashboard is the number everyone quotes and the least useful one available. It is measured in a closed environment, it usually credits the ad for orders from shoppers who were already going to reorder, and it says nothing about whether the buyer came back. Treat it as a directional signal about whether a placement is functioning, not as the answer to whether the channel is working.

The numbers worth building a view on are per-store rather than campaign-level. Sales per store per day tells you whether the product is actually moving where it sits, which is also the metric the platform's own systems care about when they decide who gets space. Share of stores where you are in stock tells you how much of your paid reach is even fulfillable. Repeat rate on the same SKU over a few weeks tells you whether the promoted trial converted into habit, which is the only thing that makes the acquisition cost here defensible.

Then look at what the spend did to the operational picture, not just the sales line. A campaign that lifted velocity in twenty stores and got you assortment in thirty more has done something a ROAS figure will never show. A campaign that produced a spike, drained stock, and left you unavailable for a fortnight has destroyed value even if the reported return looked strong during the spike. Judge the channel on the position it leaves you in, because in quick commerce shelf space is the compounding asset and the media is just how you argue for it.

Keep one honest control in the picture too. If branded search and direct traffic move while quick commerce spend is running, that is worth noting as a halo, but it cannot be attributed cleanly and should not be used to rescue a campaign whose in-platform numbers do not work. The discipline is the same one that applies to any channel that reports on itself: measure what the channel controls, and be suspicious of every number that flatters it.

When Not to Advertise Here Yet

If you are live in a small minority of the stores in your target cities, the money belongs in assortment and supply, not media. Advertising into a thin footprint buys impressions that cannot convert and teaches the platform that your promoted listings underperform, which is the opposite of what you want on record when you next ask for space. Get stocked properly first. It is slower and less satisfying and it is the actual constraint.

Skip it too where the category is genuinely unfamiliar. Quick commerce is a repertoire channel: it works best for products a shopper already understands and can decide on in a second, and it is a poor place to introduce a new format or a product that needs explaining. Something requiring comparison or demonstration will convert better through content and long-form creator work first, and can then arrive here as a known quantity rather than a puzzle on a tile.

Hold off if the unit economics do not survive the channel's structure. Platform margin, the co-funded promotion, and the media all land on the same unit, and on smaller pack sizes there is often very little room left underneath. Work out what the pack actually contributes after all three before you commit budget, because a campaign that grows volume at a negative contribution is not a growth problem, it is a pricing problem wearing a media costume. Getting that architecture right across channels is its own piece of work, laid out in how to price the same product across marketplaces without starting a channel war.

And skip it when the tile is not ready. If the pack image does not read at thumbnail size, if the variant is ambiguous, or if the title on the platform is a copy of an Amazon title that nobody will finish reading, fix that before spending. It costs nothing and it changes the conversion rate on every impression you go on to buy, paid or organic.

None of this makes quick commerce advertising a bad channel. It is one of the few places where a small brand can sit beside a national one in the same tile and get chosen on merit, and the ten-minute basket rewards products that solve an immediate need. It just rewards operators rather than bidders. The brands doing well here are not the ones with the cleverest campaign structure, they are the ones who are in stock in the stores they are paying to appear in.