ACoS on Amazon and Flipkart ads: which number is profitable for you

"What is a good ACoS?" has no single answer, because the same number can make one seller money and lose another seller money. This post shows you how to work out the break-even ACoS on Amazon and Flipkart ads for your own products, the level above which ads lose money, judge each campaign against it, and see what ads cost your whole business.

The prices, fees, GST rates and campaign numbers in the example are made up. Use your own rate card, ad reports and invoices. This post is not tax advice: check GST on your products and ads invoices with your CA.

What ACoS and ROAS measure

Amazon Ads defines ACoS (advertising cost of sales) as ad spend divided by attributed sales, the sales your ads brought. ROAS (return on ad spend) is the inverse: ad sales divided by ad spend.

• ACoS = ad spend ÷ ad sales × 100. ₹3,600 of spend that brought ₹19,176 of ad sales is an ACoS of 18.77%.
• ROAS = ad sales ÷ ad spend. The same campaign has a ROAS of 5.33. One is the inverse of the other: 100 ÷ 18.77 = 5.33.
• Cost per click (CPC) = spend ÷ clicks, what you paid on average for each click.

Amazon Ads says there is no general threshold for a good or bad ROAS: a good one is one that is profitable for you. It also says your ACoS needs to be lower than your profit margin to keep a profit. That is the number to find.

Flipkart Ads, on Flipkart Seller Hub, offers Cost per Click and Smart ROI campaigns. Flipkart says a Smart ROI campaign aims to maximise return on investment by ensuring sales for every rupee spent, and may not give as much visibility as a CPC campaign. Its public pages do not give the formula for its ROI figure, so check how your report works it out before you compare it with ROAS. The maths below works on your own numbers from either marketplace.

Find your break-even ACoS

Your break-even ACoS is the ACoS at which the ads take everything one order leaves you. Above it, every ad sale loses money. To find it:

1. Start with the selling price the buyer pays.
2. Take out the GST inside it, if you are registered: price × rate ÷ (100 + rate).
3. Subtract the marketplace fees per order from your rate card, the product cost and packing. What is left is your profit per order before ads.
4. Divide that by the selling price. That share, as a percentage, is your break-even ACoS.
5. Take your target profit, as a share of the price, off it. That is your target ACoS.

The formulas are our measure, built on Amazon Ads' rule that ACoS must stay below your margin. One more cost can count: if you cannot claim back the GST on your ads invoice, each ₹100 of reported spend costs ₹100 plus that GST, so divide your break-even ACoS by (1 + the GST rate). Under section 16 of the CGST Act, a registered business can claim input tax credit only when the conditions are met, so ask your CA.

A worked example: one steel bottle, four campaigns

Kaveri Steel Bottles in Coimbatore is a made-up seller of a 1 litre steel bottle at ₹799 on Amazon.in. With a made-up GST rate of 18%, the price before GST is ₹677.12. Fees from its made-up rate card come to ₹170 an order, the bottle costs ₹260 and packing ₹30. One order leaves ₹217.12 before ads.

• Break-even ACoS: 217.12 ÷ 799 = 27.2%.
• Target ACoS for 10% profit on the price: 27.2% − 10% = 17.2%.
• If the seller could not claim back a made-up 18% GST on the ads invoice, break-even would fall to 27.17% ÷ 1.18 = 23.0%. Here the GST is claimed back.

Last month's four campaigns, from a made-up report:

• Exact keywords: ₹3,600 spent, 24 orders, ₹19,176 of ad sales. ACoS 18.8%, under break-even. After ads it made about ₹1,611 (our measure: ad sales × 27.17% − spend).
• Auto targeting: ₹5,200 spent, 20 orders, ₹15,980. ACoS 32.5%, above break-even: a loss of about ₹858.
• Broad keywords: ₹2,400 spent on 380 clicks, no orders. A loss of ₹2,400.
• Brand: ₹900 spent, 9 orders, ₹7,191. ACoS 12.5%, a profit of about ₹1,054.

Together the four spent ₹12,100 for ₹42,347 of ad sales: an ACoS of 28.6%, just above break-even, and a loss of about ₹593. The average hides the story. Two campaigns make money, one loses a little and one spends with nothing to show. Pausing the broad campaign alone brings the total to an ACoS of 22.9% and a profit of about ₹1,807.

For the auto campaign, the highest average cost per click that stays within break-even is break-even ACoS × ad sales ÷ clicks: 27.2% × ₹15,980 ÷ 1,300 clicks, about ₹3.34, against ₹4.00 now. Within the 17.2% target it is about ₹2.11. That is our measure, and it assumes clicks keep turning into orders at the same rate.

Put your own price, fees and campaigns into the Amazon and Flipkart Ads ACoS Calculator. It works out what one order leaves you before ads, your break-even and target ACoS, and for each campaign you type or paste from a report, the ACoS, ROAS, cost per click, click-through rate, conversion rate, profit or loss after ads and the highest cost per click within your target. Campaigns above break-even and spend with no orders come first in a do-these-first list. It never signs in to Amazon Ads or Flipkart Ads, and no fee or tax rate is built in.

Why a "good ACoS" depends on your margin

Take a 25% ACoS. On a product that breaks even at 35%, the campaign keeps about 10% of its ad sales as profit. On a product that breaks even at 18%, the same campaign loses about 7% of its ad sales. Same ACoS, opposite result.

So compare each campaign with the break-even ACoS of the product it sells, not with a number from a forum. A few more points, all our suggestions:

• Thin margins leave little room for ads. If one order leaves almost nothing before ads, fix the price, the cost or the fees first. Ads can only add cost.
• Set a limit before you judge a new campaign, for example what one or two orders would earn you before ads, rather than stopping it after a handful of clicks or letting it run for weeks.
• Check the listing the ad sends buyers to. The price, photos and reviews decide whether a click becomes an order.

Our post on margin and markup explains why the margin, not the markup, is the number to compare with ACoS.

TACoS: what ads cost the whole business

ACoS judges campaigns. It cannot tell you what ads cost the business, because it counts only ad sales. For that we use TACoS (total advertising cost of sales) as our measure: ad spend ÷ total sales, ads and organic together, for the same period × 100. Amazon Ads mentions TACoS too, in its guide to common holiday advertising myths, as a way to see whether total sales grow in relation to ad spend.

The bottle seller sold 190 bottles last month, ₹1,51,810 in all, so ₹12,100 of ad spend is a TACoS of 8.0%. Our reading: if TACoS rises month after month while total sales stay flat, ads may be buying sales you would have made anyway.

TACoS also matters when you work out profit per product. ACoS treats every sale as an ad sale and shows a higher ad cost per order than you really pay. In the SKU Profit Analyzer, type ad spend per order (a month's spend for the SKU ÷ all its orders), or put TACoS in the "Ads % of sales" column. It works out profit per order and per month for every SKU and channel after fees, returns, ads and shipping, finds the loss-makers and lists fixes, with no fee rate or courier price built in.

Check your fees before you set a target

Your break-even ACoS is only as good as the fees behind it. Amazon.in's fees and pricing page says its referral fee is a percentage of the sale that varies by category, with a closing fee based on the price. Flipkart's fees and commission page says its commission varies by category and its fixed fee depends on the selling price. Copy yours from your own rate card into the Marketplace Fee and Profit Calculator to see the payout, profit and break-even price per order, then use that profit before ads in the ACoS sum. Our post on your real profit per order lists every fee line, returns included.

Mistakes to avoid

• Comparing ACoS with your gross margin. Take out GST, fees and packing first, or break-even looks higher than it is.
• Judging only the account average. A healthy average can hide a campaign that spends with no orders.
• Using ACoS as the ad cost in SKU profit. Use ad spend per order or TACoS.
• Forgetting GST on ads when you cannot claim it back.
• Reading Flipkart's ROI as ROAS without checking how your report works it out.
• Keeping the same targets during a sale. A discount lowers the profit per order, and with it your break-even ACoS, so work it out again at the sale price.

Your next step

Pick your three best-selling products and work out each one's break-even ACoS from your rate card. Then put last month's campaign report into the ACoS Calculator and start with what it flags first. If you are planning a sale, our post on sale discounts that do not eat your margin shows how a discount changes the profit your ads have to work with. All our free tools for online sellers work in your browser. If you want a second look at your numbers, get in touch.