A 25% ACOS can be a great result on one product and a slow bleed on another — and the difference has nothing to do with the ad campaign itself. It comes down to a number most sellers never calculate precisely: true margin after Amazon fees and landed cost. Amazon PPC is an auction where you bid real dollars against that margin, and if you don't know the margin, you're bidding blind.
This guide covers the fundamentals that actually drive profitability: the three campaign types, how the auction prices a click, ACOS versus TACOS with worked math, your break-even ACOS, budget pacing, and where ad spend ends up in your books. Every number in the examples is illustrative — your fees and costs will differ, so check Amazon's current fee schedules and run the math on your own SKUs.
The Three Amazon PPC Campaign Types
Amazon's ad console offers three core campaign types, and most sellers should think of them as a funnel of increasing complexity. Amazon Sponsored Products are the keyword- and product-targeted ads that appear in search results and on product pages, styled to look like organic listings. They are where nearly every seller starts and where most ad budgets live, because they capture shoppers at the moment of highest purchase intent.
Sponsored Brands (banner ads with your logo and multiple products, available to brand-registered sellers) and Sponsored Display (audience and retargeting ads that can follow shoppers on and off Amazon) are the other two. They have their place, but a profit-minded seller earns the right to use them by first making Sponsored Products work at an acceptable ACOS. If you can't buy a profitable click on your own category keywords, broader awareness spend usually just scales the loss.
- •Sponsored Products: keyword/ASIN-targeted, in search results and on listings — start here
- •Sponsored Brands: banner placements with logo and product collection — requires Brand Registry
- •Sponsored Display: audience targeting and retargeting on and off Amazon
- •Within Sponsored Products: automatic campaigns mine search terms; manual campaigns let you control bids on proven winners
How the Amazon PPC Auction Actually Works
When a shopper searches, Amazon runs an instant auction among every ad targeting that query. Your bid is the maximum cost-per-click you're willing to pay, but you generally pay just enough to beat the next competitor's effective bid — not your full maximum. That's why bidding $2.00 often produces actual CPCs of $1.20 or $1.40. The bid sets your ceiling; competition sets the price.
Crucially, the auction isn't decided by bid alone. Amazon weights relevance and expected click-through and conversion performance, because Amazon gets paid per click and only wants to show ads people will click and buy from. A listing with strong images, reviews, and conversion history can win placements at lower bids than a weak listing bidding aggressively. Before scaling Amazon advertising cost, fix the listing — every conversion-rate improvement is a permanent discount on your CPCs.
Bidding strategies (down-only, up-and-down, fixed) and placement multipliers for top-of-search adjust your bid situationally, but the fundamentals hold: you pay per click, not per sale, so every click on a product that doesn't convert is pure cost.
ACOS vs TACOS: The Worked Math
ACOS (advertising cost of sales) is ad spend divided by ad-attributed sales. Spend $300 on ads that generate $1,200 in attributed sales and your ACOS is 300 ÷ 1,200 = 25%. It tells you how efficiently the ads themselves convert spend into revenue — and nothing about the rest of your business.
TACOS (total advertising cost of sales) is ad spend divided by total sales, organic included. Same $300 spend, but the product did $3,000 in total sales that month: TACOS is 300 ÷ 3,000 = 10%. TACOS is the better long-run health metric because Amazon PPC feeds organic rank — ad-driven sales improve your position in search, which drives organic sales the ads never get credit for.
Read them together. ACOS rising while TACOS holds steady usually means ads are getting less efficient but organic is picking up the slack — often fine. TACOS creeping up month after month means the product is becoming structurally dependent on paid traffic, and that dependency comes straight out of margin. A launch tolerates a high ACOS deliberately; a mature product should see TACOS drift downward.
See what Amazon owes you — free
Connect your seller account and get a free reimbursement audit. No credit card, keep 100% of what you recover.
Break-Even ACOS: Judge PPC Against True Margin
Here is the calculation most sellers skip. Your break-even ACOS is your true pre-ad margin: what's left after Amazon's referral fee (typically 8–15% of sale price depending on category), FBA fulfillment fees, and your full landed cost — factory cost plus freight, duties, and inbound shipping. Not your invoice cost. Landed cost.
Illustrative example: a product sells for $30. Referral fee at 15% takes $4.50. Assume an FBA fulfillment fee of $5.50 (check Amazon's current rate card for your size tier) and a landed cost of $9.00. True pre-ad profit: $30.00 − $4.50 − $5.50 − $9.00 = $11.00, a 36.7% margin. That 36.7% is your break-even ACOS — run ads at exactly that ACOS and ad-attributed sales make zero profit. Run at 25% and you keep about $3.50 per ad sale; at 45% you lose roughly $2.50 every time the ad 'works.'
The trap: sellers who compute margin from invoice cost instead of landed cost overstate their break-even ACOS, set bids accordingly, and lose money on campaigns their dashboards score as wins. The chart below shows how each step up in ACOS comes directly out of this example product's net margin.
Budget Pacing: Don't Let the Algorithm Spend Your Month by the 10th
Daily budgets on Amazon are soft caps — Amazon can spend meaningfully more than your daily budget on a high-traffic day and balance it out across the month. Combine that with the fact that a too-small budget exhausts early in the day, and you get the classic pacing failure: your ads run during low-converting morning hours, go dark by afternoon, and you never appear during the evening window when your shoppers actually buy.
Practical pacing discipline: set daily budgets you could sustain every day of the month, review spend-by-campaign weekly rather than reacting to single days, and starve losers to feed winners — pull budget from campaigns above break-even ACOS with adequate click volume, and add it to campaigns converting below target. If a campaign hits its budget cap daily while running below your target ACOS, that's not a problem, it's a signal: raise the budget before raising bids.
Watch attribution lag, too. Amazon attributes sales to clicks over a multi-day window, so yesterday's ACOS always looks worse than it will after the window closes. Make bid decisions on data that's had several days to settle, not on this morning's report.
Where Amazon Advertising Cost Lands in Your Books
Amazon PPC spend is invoiced by Amazon Advertising and, for most sellers, deducted directly from your bi-weekly settlement payouts. That convenience is an accounting trap: the deposit hitting your bank is net of ad spend, referral fees, FBA fees, refunds, and a dozen other adjustments. If you book the deposit as revenue, you understate both sales and advertising expense, and your P&L can't tell you whether PPC is profitable.
Done correctly, each settlement gets broken out: gross sales as revenue, ad spend to an advertising expense line, Amazon fees to their own lines, refunds against revenue. That's exactly the decomposition BeanHawk automates — it posts summarized settlement journals to QuickBooks Online and Xero, so ad spend shows up as its own expense line against true gross revenue instead of vanishing inside a net deposit. With landed cost flowing through inventory valuation as well, the 'true margin' from the break-even section is a number your books produce, not a spreadsheet you rebuild quarterly.
The discipline pays off beyond ads. Once advertising cost, fees, and landed COGS are cleanly separated by month, TACOS trends, contribution margin per SKU, and the decision to scale or kill a product become readable straight off your P&L. As with anything touching your books and taxes, the right treatment can vary by entity and jurisdiction — confirm specifics with your accountant or tax professional.