What Is ACOS on Amazon? Formula, Definition & What It Measures

ACOS (Advertising Cost of Sales) = ad spend ÷ ad revenue × 100. It measures how many cents you spent in ads for every dollar of ad-attributed sales. Amazon's official ACOS guide confirms the formula with a worked example: $50 ad spend ÷ $100 ad revenue = 50% ACOS. Lower ACOS means your ads generated more revenue per dollar spent.

ACOS is a channel-efficiency ratio, not a profitability statement. It only sees ad-attributed sales — not organic sales, not your cost of goods, not Amazon referral fees or FBA costs. As of January 1, 2026, Amazon replaced its 14-day view-through attribution window with a shopping-signal enhanced last-touch model, so the "ad revenue" in your ACOS calculation now reflects that model by default. The mechanics behind that change — and what it means for historical comparisons — are covered in Amazon Ads reporting and attribution. The short version: a campaign’s ACOS from before January 2026 is not directly comparable to its post-January figure.

The most expensive misconception in Amazon advertising is that a lower ACOS is always better. It is not. There is a ceiling on how low ACOS should go, and that ceiling is your break-even ACOS — the ACOS at which the ad generates zero profit. An ACOS well below break-even means you are under-investing in visibility you can profitably afford. The break-even identity, verified against Amazon’s own guide: break-even ACOS = your pre-ad profit margin. A product with a 36% pre-ad margin can run up to 36% ACOS and still break even on the ad. Driving it to 15% when the margin is 36% leaves profitable reach on the table.

Key Takeaways

• ACOS = ad spend ÷ ad sales × 100. ROAS = ad sales ÷ ad spend. They are arithmetic reciprocals: 25% ACOS = 4× ROAS, 20% ACOS = 5× ROAS.
• Break-even ACOS equals your pre-ad profit margin — not 100%. 100% ACOS means ad sales equal ad spend before product costs; your actual break-even is your margin percentage.
• TACOS (total ad spend ÷ total sales) is the strategic health metric. Falling TACOS at stable spend means organic sales are compounding faster than ad spend — the flywheel working.
• CVR on Amazon = orders ÷ clicks (not impressions — that is CTR’s denominator).
• Search-term impression share is paid SOV only — your share of ad impressions for a term, not organic results. Third-party tools model total SOV separately.
• New-to-Brand metrics are not available on Sponsored Products. NTB requires Sponsored Brands, Sponsored Display, or Amazon DSP.
• No universal "good ACOS" exists. Amazon states this explicitly. The right target sits between your break-even ACOS and your desired net margin, set per campaign objective.

Amazon Ads Metrics: Definitions & Formulas at a Glance

Every number on your Amazon Ads dashboard traces to one of nine metrics. The table below defines each one, states the exact formula, and shows what dimension of performance it measures. ACOS and ROAS encode the same information — they are reciprocals. TACOS adds the organic dimension that ACOS cannot see. SIS and NTB are the forward-looking metrics that predict whether your category position is strengthening or eroding.

Three practical notes on formula precision. First, CVR on Amazon is orders divided by clicks — not impressions. Impressions are CTR’s denominator. Conflating them produces a CVR number that is 100× too small. Second, the "ad revenue" in every ACOS and ROAS calculation now uses Amazon’s 2026 default attribution model; pre-2026 figures used the 14-day view-through window. Third, impression share measures paid visibility only — Helium 10 and Pacvue model total SOV (including organic results) separately from Amazon’s SIS report, which covers paid placements only.

ACOS vs ROAS: The Same Number, Two Formats

ACOS and ROAS are not two different metrics — they are the same efficiency ratio expressed in different units. ACOS is a percentage (spend as a share of revenue); ROAS is a multiplier (revenue per dollar of spend). The conversion formula is ROAS = 100 ÷ ACOS%, which means 25% ACOS = 4× ROAS, 20% ACOS = 5× ROAS, and 50% ACOS = 2× ROAS. Amazon’s official guide identifies ROAS as "the opposite of ACOS" and demonstrates the reciprocal with two worked examples.

The practical consequence of the reciprocal: inconsistency across tools is the only real risk. Sponsored Products campaigns default to ACOS in the Amazon console; DSP campaigns typically report in ROAS. A team that reads one report in ACOS and another in ROAS without converting can misread identical performance as a divergence. The metric that genuinely adds new information — because it includes organic sales that neither ACOS nor ROAS can see — is TACOS, covered in the section below. See Amazon Ads optimization and automation for how search-term harvesting and bid rules translate an ACOS target into spend decisions.

Break-Even ACOS: Why It Equals Your Profit Margin

Break-even ACOS is the maximum ACOS you can run before the ad generates zero net profit from the sale. Amazon’s ACOS guide states this directly: "to achieve profit, the Amazon ACOS must be lower than the profit margin." The identity: break-even ACOS = pre-ad profit margin %. Pre-ad profit margin is what remains after subtracting COGS, referral fee, and FBA fees from the selling price — before subtracting ad cost.

The Feedvisor $25 product worked example. Feedvisor’s 2026 ACOS guide traces the calculation step by step: (1) Start with selling price: $25.00. (2) Subtract COGS + shipping: −$8.00. (3) Subtract Amazon referral fee (15%): −$3.75. (4) Subtract FBA fulfillment fee: −$3.86. (5) Subtract other variable costs: −$0.30. (6) Total non-ad costs = $15.91. Pre-ad profit = $25 − $15.91 = $9.09. Break-even ACOS = $9.09 ÷ $25 × 100 = 36.4%. To achieve a 15% net margin on this product, the target ACOS = 36.4% − 15% = 21.4%.

Target ACOS is set between break-even and the desired net margin, deliberately, per objective. A launch campaign acquiring rank and reviews runs at or above break-even. A mature defense campaign on a high-review ASIN runs well below. Neither is wrong — the error is applying the same target ACOS to every campaign regardless of objective. Amazon Ads bidding and budgets covers how dynamic bidding and placement adjustments translate an ACOS target into bid and budget decisions.

What Is a Good ACOS on Amazon? 2026 Category Benchmarks

Amazon states explicitly: there is no specific number for a good ACOS. The correct benchmark is your own break-even ACOS (Table 3 above). Category medians from Autron’s June 2026 benchmark report provide a sanity check against competitors in the same space — not a target. A 38% ACOS in Clothing & Apparel (category median 42%) is outperforming the field; a 38% ACOS in Books (median 19%) signals a structural problem with bidding or listing quality.

Three structural patterns in the data. First, CVR and ACOS move inversely across categories: Books and Food & Grocery have the lowest ACOS and the highest CVR, because high purchase intent at low CPCs produces efficient conversion. Electronics runs 29% ACOS with only 9.5% CVR — high CPCs and comparison-shopping behavior compress the ratio. Second, the Clothing & Apparel median (42%) is an outlier driven by high return rates, broad keyword competition, and thin margins common in fashion PPC. Third, CPC is rising: Ad Badger reports CPC up ~$0.10 year-over-year, with a May 2026 peak of $1.27. CTR is flat for the first time since 2019 — competition has compressed visibility without compressing click rates further. Brands in beauty PPC, skincare PPC, and supplements & nutrition PPC face the highest CPC-to-margin pressure: Beauty median CPC $1.18 at a 24% category ACOS suggests margins in the 28–35% range to maintain profitability.

What Is TACOS? The Strategic Health Metric for Amazon Brands

TACOS (Total Advertising Cost of Sales) = total ad spend ÷ total sales (ad + organic) × 100. The "+ organic" in the denominator is what distinguishes TACOS from ACOS and makes it the strategic metric. Perpetua’s worked example: $2,500 ad spend ÷ $30,000 total sales = 8.33% TACOS. If that same account had $20,000 in ad-attributed sales, its ACOS would be 12.5% — a completely different picture of efficiency for the same ad spend.

TACOS trend direction is the diagnostic. NovaData summarizes the interpretation: falling TACOS means organic sales are growing faster than ad spend — the Amazon flywheel is compounding. Rising TACOS at rising spend means ads are doing all the work and organic is flat. Saras Analytics adds the profitability framing: if TACOS falls below the organic profit margin, the overall business is profitable after advertising. No per-category TACOS dataset is publicly available; the lifecycle ranges below are directional vendor-cited figures from Saras, NovaData, and Perpetua.

Brands in pet supplies PPC, food & beverage PPC, and home goods PPC often reach mature TACOS ranges (5–10%) within 12–18 months when organic rank compounds from a strong initial launch. Categories with high return rates — clothing, electronics — sustain higher TACOS longer because repeat-purchase cycles are slower and brand loyalty is harder to build on Amazon.

New-to-Brand Metrics: Real Customer Acquisition vs Repeat Buyers

New-to-Brand (NTB) metrics classify each ad-attributed order as either a first-time brand buyer or a repeat customer, using a rolling 365-day lookback. Amazon’s NTB help documentation defines the threshold: if the shopper has not purchased from the brand on Amazon in the prior 365 days, the order is New-to-Brand. The lookback is rolling, not lifetime — a customer from 14 months ago qualifies as NTB.

The SP exclusion matters more than most sellers realize. Sponsored Products generates the majority of Amazon ad sales for most accounts — yet it produces zero NTB data. A brand-defense campaign on its own brand keywords can run a 10% ACOS while acquiring almost no new customers; the ACOS looks exceptional but the campaign is only harvesting people already searching the brand name. To measure real acquisition cost, NTB analysis requires Sponsored Brands or Sponsored Display campaigns. The NTB Order Rate on Sponsored Brands — NTB orders ÷ clicks — is the cleanest per-click acquisition cost signal. Adbrew’s example: 60 NTB orders out of 2,000 clicks = 3% NTB order rate. This is what justified that SB campaign’s higher ACOS: every three clicks, the brand acquired a net-new customer.

Search-Term Impression Share: Amazon’s Paid Share of Voice

Search-term impression share (SIS) is your account’s percentage of total ad impressions for a given search term across all advertisers. Perpetua’s report on the SIS metric provides the example: a 40% SIS on "running shoes" means your ads captured 40% of all paid impressions for that term — competitors took the other 60%. Impression rank is the ordinal counterpart: Rank 1 = highest impression share among all advertisers; Rank 3 = two competitors generated more impressions than you.

SIS is a leading indicator. ACOS, sales, and units are trailing metrics — by the time they move, the category has already shifted. When a competitor floods a head term, your impression share erodes first; sales rank and ACOS follow weeks later. Tracking SIS weekly on the 10–20 terms that drive the most revenue gives an early-warning signal to act before the P&L confirms what already happened. Amazon Ads targeting covers how keyword structure and match type selection determine which terms contribute to your SIS.

One precision the glossary pages consistently blur: Amazon’s SIS report measures paid SOV only. It does not capture organic positions, organic clicks, or competitor organic ranks. Third-party tools model total SOV across the full SERP — that is a different and broader signal. Both matter; neither is a substitute for the other. Use Amazon’s SIS report for weekly paid-visibility tracking; use a total-SOV tool for quarterly category share analysis and competitor positioning audits. Brands in categories like electronics PPC and home decor PPC with highly competitive head terms find SIS rank the most actionable weekly metric for protecting category position.

Which Amazon Ads Metrics to Watch by Campaign Objective

No single metric governs all campaign types. ACOS is the right efficiency lens for Sponsored Products harvest campaigns; TACOS is the right strategic lens for brand health; NTB rate is the right acquisition lens for Sponsored Brands prospecting. The table below maps objective to the metric stack that gives the clearest read.

Sellers in watches & jewelry PPC, outdoor & camping PPC, and handbags & accessories PPC typically weight TACOS most heavily because repeat-purchase rates are lower and new-customer acquisition is the primary growth lever. Categories with subscription or replenishment dynamics — food & grocery DTC PPC — weight NTB less heavily because customer lifetime value justifies a higher acquisition-phase ACOS.

TACOS vs ACOS: Why a Falling TACOS Is the Strategic Win

A campaign can show a 40% ACOS — above the category median, a number that triggers concern — while TACOS quietly falls from 18% to 12% over the same quarter. These two numbers are not contradictory. The 40% ACOS reflects the ad channel working hard in competitive terms. The falling TACOS reflects organic sales growing faster than ad spend — the ads seeding rank, reviews, and repeat purchase, which then compound as organic traffic. ACOS cannot see that organic growth; TACOS can.

The inverse can also be true: a "great" ACOS of 12% while TACOS rises from 8% to 15% means organic is eroding and ads are compensating. The business looks efficient at the ad level while the underlying organic engine stalls. This is the most common pattern MB Adv Agency identifies in accounts that report good ACOS numbers but declining category rank: the ACOS metric hides the organic signal that TACOS exposes. Reading TACOS over at least 90-day windows before evaluating any single campaign’s ACOS is how we distinguish a genuinely efficient account from one that is winning an efficiency metric while losing the business.

See Amazon Sponsored ads for how Sponsored Products, Sponsored Brands, and Sponsored Display each contribute differently to the TACOS calculation — SP drives most ad-attributed sales and dominates ACOS; SB and SD contribute disproportionately to organic rank and NTB acquisition, which is why their ACOS looks "worse" in isolation.

Three Amazon Metrics Misconceptions That Cost Sellers Money

Misconception 1: “A lower ACOS is always better.”

This is the most expensive misconception in Amazon advertising. An ACOS far below break-even means you are under-investing in visibility you can profitably afford. Amazon’s own ACOS guide is clear: "to achieve profit, ACOS must be lower than the profit margin" — which means the target is below your margin, not as far below as possible. A 15% ACOS on a 45%-margin product is not a trophy; it is evidence you capped reach to protect a number that did not need protecting. Lower ACOS is only better when it is not purchased at the cost of rank, organic compounding, or new-customer acquisition.

Misconception 2: “Break-even ACOS is 100%.”

This is a category error. 100% ACOS means ad sales equal ad spend — the ad transaction broke even before product costs. Your actual break-even includes COGS, referral fees, and FBA fees. On the Feedvisor $25 example, break-even ACOS is 36.4% — not 100%. Running at 100% ACOS while selling a $25 product that costs $15.91 to make and fulfill is a loss of $15.91 per sale. The 100% figure circulates because it is the break-even for the ad spend itself, stripped of all product economics — a number with no operational meaning for a seller.

Misconception 3: “ACOS tells you whether the account is profitable.”

ACOS is a channel-efficiency ratio. It compares ad spend to ad-attributed sales and knows nothing about COGS, referral fees, FBA costs, returns, or organic sales. An account at break-even ACOS is at $0 ad profit per sale, not account profit. Saras Analytics and NovaData both frame TACOS — not ACOS — as the profitability health metric because TACOS captures total revenue, including the organic sales that ad spend generates indirectly. Account-level profitability requires a margin model: revenue minus COGS, fees, ad spend, and returns — a calculation that starts after ACOS, not with it.

Frequently Asked Questions: Amazon Ads Metrics

Methodology

Formula benchmarks draw from Amazon Ads, "What is ACOS?" and Amazon Ads, "New-to-brand metrics". Category benchmarks from Autron, "Amazon PPC Benchmarks by Category (2026)," June 3, 2026, cross-checked against Ad Badger, "Amazon Advertising Benchmarks 2026" and Trellis, "Amazon Ads Benchmarks 2026". TACOS lifecycle ranges from Saras Analytics and NovaData (2026). SIS data from Perpetua (2026). NTB definitions from Amazon Ads and SellerStack (2026). Break-even ACOS worked example from Feedvisor (2026). All benchmarks are vendor estimates from managed-account datasets — not mbadv client data. Last updated: June 2026. Reviewed by MB Adv Agency, June 2026.