Key takeaways
- ACoS (ad spend ÷ ad sales) is too high only when it is above your break-even ACoS, your profit margin before ads. For example, at a 30% margin, 25% ACoS makes money and 35% loses it.
- Diagnose before you cut bids. The excess sits in search terms with no orders, targets above target ACoS, broad and auto leakage, weak listings or misplaced budget.
- Fix in order of impact: negate terms that spent more than your profit per unit without an order, then re-bid over-target targets with bid = order value × conversion rate × target ACoS.
- Give proven converters and testing terms different bid rules, and judge every change on data at least 7 days old (14 days for vendors, Sponsored Brands and Sponsored Display).
Your Amazon ACoS is too high only when it is above your break-even ACoS, the share of the selling price you keep before paying for ads. If it is above that line, don't cut every bid by the same amount. Find where the extra spend goes (search terms with no orders, targets above target ACoS, broad and auto leakage, weak listings, budget in the wrong campaigns) and fix those in order of impact, starting with the fix that costs no sales.
ACoS, advertising cost of sale, is ad spend ÷ ad sales. For example, ₹20,000 of ad spend that brings ₹80,000 of attributed sales is a 25% ACoS, which is a ROAS (ad sales ÷ ad spend) of 4.
Is your ACoS actually too high? Compare it with break-even ACoS
An ACoS figure means nothing until you set it against your margin. Break-even ACoS is the point where the profit on an ad-driven sale exactly covers its ad cost:
Break-even ACoS = profit per unit before ads ÷ selling price
Profit before ads is the price minus product cost, Amazon's fees, shipping and GST. Take fees from Seller Central's current rate card, or use our free Amazon profit calculator, which gives profit per unit after fees and GST and the break-even ACoS. Then decide what you want to keep after ads:
Target ACoS = break-even ACoS − profit margin you want to keep after ads
Worked example. Say you sell a steel water bottle at ₹999 and keep ₹300 per unit before ads. Break-even ACoS is ₹300 ÷ ₹999 = 30%. You want to keep 10% of the price, so your target ACoS is 20%.
| Your ACoS | What it means | What to do |
|---|---|---|
| Below target (under 20%) | Profitable, with room to spare | Test higher bids on your top converters |
| Between target and break-even (20% to 30%) | Profitable, but under your goal | Tune bids and search terms |
| Above break-even (over 30%) | Each ad sale loses money: at 40%, about ₹100 per ₹999 sale | Diagnose before you spend more |
Running above break-even can be deliberate, for a launch that needs sales history or a campaign that wins repeat buyers. Give it a spend limit and an end date, and watch TACoS (ad spend ÷ total sales, organic included) alongside ACoS. Our guide to TACoS vs ACoS covers when each should lead.
Why is your Amazon ACoS too high? Five places to look
A high ACoS always traces back to one of three inputs:
ACoS = CPC ÷ (conversion rate × average order value)
CPC is cost per click (spend ÷ clicks) and conversion rate is orders ÷ clicks. So the problem is a click that costs too much, a click that doesn't convert, or an order that's too small. Pull 30 to 60 days of data and leave out the latest 7 days (14 for vendors, Sponsored Brands and Sponsored Display). A click can still become an order credited to your ad within that window, so recent clicks look worse than they are; older data is settled.
| Where the excess hides | Report to check | What it looks like |
|---|---|---|
| Search terms with clicks and no orders | Search term report | No orders after spending more than your profit per unit, or after two to three times the clicks you'd expect per order |
| Targets above target ACoS | Targeting report (keywords and product targets) | ACoS over target, with enough clicks to judge |
| Broad and auto leakage | Search term report, by campaign | A term you bid on in exact still spends through auto or broad, at a worse ACoS |
| Listings that don't convert | Advertised product report, Business Reports | One product has a high ACoS on most targets and a low unit session percentage (units ordered ÷ sessions) |
| Budget in the wrong campaigns | Campaign report, budget status | Under-target campaigns run out of budget; over-target ones never do |
Both limits in the first row are arithmetic. A term that has spent more than your profit per unit (₹300 for the bottle) with no order has already lost money, even if its next click converts. And you expect an order every 1 ÷ conversion rate clicks: at a 10% conversion rate, one per 10 clicks. For example, if a term truly converted at 10%, the chance of 30 clicks with no order would be about 4% (0.9 multiplied by itself 30 times).
For weak listings, brand-registered sellers can check the Search Query Performance report in Brand Analytics: a purchase share well below your click share for a search means shoppers click your product and buy another. Our post on Search Query Performance and bids shows how to read it. To size each leak, add up the spend above target by group:
Excess spend = spend − (target ACoS × ad sales)
Worked example: where the bottle's extra spend went
In this example, more than half of the monthly ad spend is above target, and almost half of that excess can go without losing a sale. Say the bottle account spent ₹1,00,000 on ads last month for ₹2,50,000 of ad sales: a 40% ACoS, 10 points above break-even.
| Group | Spend | Ad sales | ACoS | Excess over 20% target |
|---|---|---|---|---|
| Search terms with no orders | ₹25,000 | ₹0 | n/a | ₹25,000 |
| Targets above target ACoS | ₹45,000 | ₹90,000 | 50% | ₹27,000 |
| Targets at or under target | ₹30,000 | ₹1,60,000 | 18.75% | None |
| Total | ₹1,00,000 | ₹2,50,000 | 40% | ₹52,000 |
Negating the zero-order terms alone takes the account to ₹75,000 ÷ ₹2,50,000 = 30%, break-even, with no loss of measured sales. Now say re-bidding brings the over-target group to 30% while it keeps ₹60,000 of its ₹90,000 in sales. The account then spends ₹48,000 for ₹2,20,000, a 21.8% ACoS. Ad sales fall 12%, but profit on ad-driven sales goes from a ₹25,000 loss (30% of ₹2,50,000 minus ₹1,00,000) to an ₹18,000 gain (30% of ₹2,20,000 minus ₹48,000). That is the trade: some volume for a lot of margin.
How to reduce ACoS on Amazon, in order of impact
To lower ACoS on Amazon India or any marketplace, start with fixes that cut spend without cutting sales, then those that trade some volume for margin, then the slower work on conversion. Don't cut every bid by the same percentage; it trims your top converters along with your worst.
1. Negate search terms that spent past your limit
Add each zero-order term past your limit as a negative exact in the campaign that bought it. For words that can never fit, such as "free", "second hand" or a model you don't sell, add a negative phrase, which blocks any search containing it. Sponsored Products negative keywords come in these two match types only; there is no negative broad. For our clients, cutting 30–40% of wasted ad spend is a typical outcome, never a guarantee. More in our guide to Amazon negative keywords.
Check: after two settled weeks, spend on zero-order terms as a share of total spend should fall while orders hold.
2. Re-bid targets above your target ACoS
For any keyword or product target with enough clicks to judge, set the bid from its own numbers:
Target bid = average order value × conversion rate × target ACoS
For the bottle, a target converting 6% of clicks needs a bid of about ₹12 to run at 20% (₹999 × 6% × 20% = ₹11.99). Lower bids buy fewer clicks, so move in steps and expect some volume to go. Two settings can lift what you pay above your base bid: dynamic bids "up and down" (switch over-target campaigns to "down only") and placement adjustments, which only ever raise bids. If product pages convert worse than top of search, lower the base bid and raise the top-of-search adjustment. Put break-even ACoS in the formula and you get the most a click can cost, which our guide to Amazon advertising costs in India uses to set budgets.
Check: each re-bid target's ACoS and orders over the next two settled weeks.
3. Move proven converters into exact match
When a search term from an auto or broad campaign converts at or under target, add it as an exact keyword in a manual campaign with its own budget, and as a negative exact in the campaign where you found it. One bid, set by the formula, now decides its price. In auto campaigns, bid each targeting group (close match, loose match, substitutes, complements) on its own ACoS.
Check: in the search term report, the term's spend should move to exact. Compare its ACoS before and after.
4. Fix conversion before raising bids
A better listing lowers ACoS on every target at once. In the bottle example, at a ₹12 CPC, lifting conversion from 6% to 7.5% takes ACoS from 20% to 16%. Check price against the products shown beside yours, rating and review count, the main image and stock on every variant. A two-pack or bundle raises order value, which works the same way if conversion holds.
Check: conversion rate by product in the advertised product report, and unit session percentage in Business Reports, month on month.
5. Move budget to campaigns under target
A campaign that spends its daily budget stops showing ads until the next day. If an under-target campaign runs out while an over-target one never does, move budget from the second to the first. In the worked example, the next rupee belongs with the targets at 18.75%.
Check: the share of total spend going to campaigns at or under target should rise week on week.
Give proven converters and testing terms different bid rules
A term with a track record and a term you're still testing need different rules. Sort by evidence, not search volume: a popular search with plenty of clicks and no orders is still a testing term.
| Rule | Proven converters | Testing terms |
|---|---|---|
| Entry rule (example) | 3 or more orders in the last 60 settled days, at or under target ACoS | Everything else: new keywords, broad and phrase, auto targeting groups |
| Where they run | Exact match campaigns with their own budget | Auto, broad and phrase campaigns with a capped budget |
| Bid | Order value × the term's own conversion rate × target ACoS | The same formula with your product's average conversion rate, never above it |
| Dynamic bidding | "Up and down" only while ACoS is well under target, otherwise "down only" | "Down only" |
| When to act | Re-bid every two weeks on settled data | Negate once spend passes your profit per unit with no order; promote when it meets the entry rule |
The limit: terms that get a few clicks a month may never reach either threshold. Group those by theme (size, colour, use) and judge the group.
ScaleSKUs runs these checks on your account's data from Amazon's Ads API, lists the search terms, targets and budget-capped campaigns that need action, and drafts the changes. Nothing changes on Amazon until you approve it, and you can try it free for 30 days.
Frequently asked questions
What is a good ACoS on Amazon India?
One at or below your target: break-even ACoS minus the margin you want to keep, so it differs by product. Across the accounts we manage, ₹39.8 Cr of ad spend has run at a 16.5% ACoS, but that blends many categories and price points. Set your target from your own margin.
Should I pause a campaign with a high ACoS?
Not before you look inside it. A campaign at, say, 45% ACoS can hold targets at 15% next to search terms with no orders, and pausing it switches off both. Negate and re-bid at the search term and target level first, then pause what still sits above break-even.
Why did my ACoS rise when I changed nothing?
One of its three inputs moved: CPC, conversion rate or order value. Competitors raising bids, a price change, a variant going out of stock or a new rival listing can each do it. Chart CPC and conversion rate by week in your targeting report to see which one changed.