Amazon PPC does not create demand, it buys attention and points it at a listing. That makes the listing, not the bid, the thing that decides what a click is worth. Structure campaigns so that every rupee tells you something you did not already know, judge ACoS against your actual contribution margin rather than a number someone quoted on LinkedIn, and expect most of the waste to sit in broad match terms nobody has read in a month.
- →A click is worth whatever the listing converts at. Scaling spend on a weak listing does not fix the listing, it just buys the same failure more often.
- →Auto campaigns are a research budget, not a growth channel. Their job is to surface search terms you then promote into exact match and control.
- →ACoS only means something next to your break-even ACoS, which is set by contribution margin. The same number can be excellent for one SKU and loss-making for the next.
- →Brand terms belong in their own campaign. Blended into the account average, their low ACoS hides how the rest of the spend is performing.
- →Most waste is boring: broad match terms nobody reviewed, ads still running on out-of-stock SKUs, and budget spread so thin no keyword collects enough clicks to teach you anything.
The Listing Decides What Your Ad Spend Is Worth
The most expensive mistake in Amazon advertising is not a bad bid. It is spending well on a listing that was never ready to receive the traffic. A sponsored placement buys one thing, a shopper's attention for a moment, and everything that happens after that moment is decided by the images, the title, the price, the review count and whether the page answers the question that made someone search in the first place. Double the budget on a listing that converts poorly and you have not bought growth, you have bought the same disappointing result more often, at a higher absolute cost.
This matters more on Amazon than on a brand's own site, because the platform is watching. Conversion rate is not just your commercial outcome, it is an input into how the algorithm treats your listing, including where it ranks organically. Ads that convert well tend to make the underlying listing stronger over time. Ads that send traffic to a page people bounce off are, in effect, paying to teach Amazon that shoppers do not want your product.
Before a campaign goes live, the honest question is whether a stranger landing on the listing has enough to decide. That usually means clean images that show scale and use, a title that reads as a product rather than a keyword dump, and enough credible social proof that the page does not feel new. If the page still looks thin, the money is better spent there first: A+ content earns its cost by answering the objections a shopper has after the images and before the buy button, and a genuine base of reviews collected without risking the listing does more for conversion than any bid adjustment available in the console.
There is a simple sequencing rule buried in this. Fix the page, then buy the traffic, then scale what converts. Brands that invert it end up with a large ad account, a small margin, and a nagging sense that Amazon does not work for their category.
A Campaign Structure That Tells You Something
Most underperforming accounts are not badly optimised. They are badly organised, in a way that makes optimisation impossible. Everything sits in two or three sprawling campaigns, every SKU shares a budget with every other SKU, and match types are mixed inside a single ad group, so when the numbers move nobody can say which decision caused it. The point of structure is not tidiness. It is that a well-organised account answers questions, and a messy one only produces totals.
The working shape is simple enough to keep in your head. Group products that genuinely share search behaviour, not products that share a shelf in your warehouse. Give each group an auto campaign whose only job is discovery, on a small capped budget you are comfortable treating as research. Give the same group manual campaigns split by match type, so that exact match, where you have evidence, is bid and budgeted separately from phrase and broad, where you are still guessing. Keep brand terms out of all of it, in their own campaign, for reasons covered further down.
What connects those pieces is a loop rather than a setup. Auto campaigns surface real search terms, converting terms graduate into exact match where you control the bid, and those same terms become negatives in the auto campaign so your own campaigns stop competing with each other. Terms that spend without converting become negative exact and stay there. Run that loop on a fixed cadence and the account gets sharper every week without anyone doing anything clever.
- Run one auto campaign per product group A small, deliberately capped budget whose only job is to discover the terms real shoppers actually type.
- Read the search term report, not the keyword report The keyword report tells you what you bid on. The search term report tells you what people searched.
- Promote converting terms into exact match Terms with orders move into their own manual exact campaign, where you control the bid on each one.
- Add those terms as negatives in the auto campaign Without this the auto and exact campaigns bid against each other on the same term.
- Cut the spenders with no orders Terms with meaningful clicks and nothing to show become negative exact, permanently.
- Repeat on a fixed cadence Weekly at the start, fortnightly once the account settles. The loop is the strategy.
Reading ACoS Without Fooling Yourself
ACoS is the number every founder asks about and the number most easily misread. On its own it says nothing, because it is a ratio of spend to ad-attributed revenue, and revenue is not margin. The only version of the question worth asking is where your break-even ACoS sits, which follows from contribution margin after marketplace commission, fulfilment, returns, and the landed cost of the product. Two brands in the same category, with different margin structures, can look at the identical ACoS and one is comfortably profitable while the other is quietly funding Amazon.
Once break-even is known, the number becomes a decision tool rather than a scoreboard. Running above break-even is a legitimate choice on a launch SKU, where you are buying velocity and review volume you cannot get any other way, or on a defensive brand campaign. Running above break-even on a mature bestseller, month after month, is not a strategy, it is a leak with a justification attached.
Two distortions are worth naming. The first is brand terms, which convert at a very high rate and a very low ACoS, and which flatten the account average into something reassuring and useless. Separate them and look at non-brand performance on its own, because that is the part that represents new demand. The second is that ad-attributed revenue is not total revenue: PPC lifts organic rank, and organic sales that follow do not appear in the campaign report. This is why many brands also watch total advertising cost of sale, ad spend against total sales on the account, as the sanity check on whether the whole operation is getting more efficient or just busier.
Where Indian D2C Ad Budgets Actually Leak
The waste in most accounts is unglamorous. Broad match campaigns nobody has opened a search term report for in a month are the single most common source, because broad match is designed to explore and exploring without a negative keyword list is just paying for other people's products to be discovered. A short weekly pass through the search term report, cutting the terms that spend and never convert, recovers more money than any bid strategy experiment.
Stock is the second leak, and it is entirely self-inflicted. Ads running on a SKU that is out of stock, or one that has quietly lost the buy box on price, spend real money for impressions that cannot convert. Whoever manages the ad account needs visibility of inventory and buy box status, or a simple rule that campaigns pause when either breaks. The third is spreading a small budget across the whole catalogue, which feels fair and is analytically useless: no keyword accumulates enough clicks to prove anything, so every decision stays a guess. Concentration is what turns spend into evidence.
The fourth is subtler and more common in India than brands admit. Ad spend gets pointed at SKUs whose real problem is price positioning or returns, not visibility. A fashion SKU with a return rate high enough to erase its own margin does not need more traffic, and neither does a product priced against a competitor who is willing to lose money for longer than you are. The discipline that helps here is the same one that governs running Amazon, Flipkart, Myntra and Nykaa as separate playbooks rather than one blended strategy: decide what each SKU on each platform is actually for, then let the ad budget follow that decision instead of leading it.
"Doubling the budget on a listing that converts poorly does not buy growth. It buys the same disappointing result more often, at a higher price."
- Brand Integer Marketplace Team
Sponsored Brands and Display: When They Earn a Line
Sponsored Products is where nearly every Indian D2C brand should spend the first several months, because it sits closest to purchase intent and produces the cleanest data. The other formats are useful, but they are additions to a working base rather than substitutes for one. Adding them early, while Sponsored Products is still unstructured, mostly makes the account harder to read.
Sponsored Brands starts earning its place once you have a brand store worth sending people to and more than one product a shopper might reasonably buy together. Its value is not the click, it is the shift from selling a single SKU to introducing a range, which is why it works better for brands with a coherent catalogue than for brands with one hero product and a long tail of unrelated items. Sponsored Display is the most situational of the three: genuinely useful for retargeting shoppers who viewed and did not buy, and for defending your own listing pages against competitor placements, but easy to overspend on when it is used as general awareness on a marketplace where nobody is browsing for inspiration.
The sequencing that tends to hold: get Sponsored Products structured and profitable on your best SKUs, add Sponsored Brands when the catalogue and store justify it, and treat Display as a targeted tool with its own small budget and its own success criteria rather than a place to park leftover spend at the end of the month.
Frequently Asked Questions
What counts as a good ACoS on Amazon India?
There is no universal number, because ACoS is only meaningful against your break-even, which is your contribution margin after marketplace fees, fulfilment, returns and the cost of the product itself. A brand with a thin margin can be losing money at an ACoS that another brand in the same category would treat as excellent. Work out the break-even ACoS for each SKU first, then decide deliberately how far above or below it you are willing to run, and why. Launch SKUs and defensive brand-term campaigns are often run above break-even on purpose. A mature bestseller usually should not be.
Should we run ads on a brand new listing with no reviews?
Only at a small, deliberately capped budget, and with the understanding that you are paying for data rather than profit. A listing with no reviews, thin images and no A+ content converts poorly, so paid traffic mostly buys expensive clicks that confirm what the listing already tells you. The better sequence is to get the listing to a credible state first, gather a genuine base of reviews, then let PPC amplify something that already converts.
Do we need to bid on our own brand name?
Usually yes, but for a specific reason: if you do not appear on your own brand term, a competitor can, and the shopper searching your name is the highest-intent traffic you will ever see. What matters is treating brand terms as a separate, tightly capped campaign so their very low ACoS does not flatter the rest of the account. Blending brand and non-brand into one number is one of the most common ways an Amazon account looks healthier than it is.
How much should a small D2C brand budget to start on Amazon PPC?
Less than most brands expect, and concentrated on fewer SKUs than they want. A small daily budget spread across an entire catalogue produces too few clicks per keyword to learn anything, which is the most common reason early PPC feels random. Pick the two or three SKUs with the best listings and the strongest reviews, put the whole budget behind those, and let the search term data from that concentrated spend tell you where to expand next.
