Product Research · Aug 9, 2026 · 7 min

Amazon Sponsored Products ROAS vs. POAS: Why Your Ads Look Profitable but Lose Money

Sean Travis

Founder · Kaldon

TLDR

Amazon Sponsored Products campaigns can show a strong ROAS while losing money on every sale because blended ROAS does not account for COGS, FBA fees, referral fees, or SKU-level margin variation. POAS (Profit on Ad Spend) subtracts all variable costs before dividing profit by ad spend, revealing actual unit economics. A product with 3x ROAS may have negative POAS if its landed cost and fees exceed ad-attributed revenue. Sellers launching cloned bestsellers with thin margins face near-impossible profitability even at high ROAS, while products with genuine demand gaps and pricing power can remain profitable at lower ROAS thresholds.

TLDR. Amazon Sponsored Products campaigns can show a strong ROAS while losing money on every sale because blended ROAS does not account for COGS, FBA fees, referral fees, or SKU-level margin variation. POAS (Profit on Ad Spend) subtracts all variable costs before dividing profit by ad spend, revealing actual unit economics. A product with 3x ROAS may have negative POAS if its landed cost and fees exceed ad-attributed revenue. Sellers launching cloned bestsellers with thin margins face near-impossible profitability even at high ROAS, while products with genuine demand gaps and pricing power can remain profitable at lower ROAS thresholds.

Why ROAS looks profitable but POAS reveals the loss

Amazon Sponsored Products campaigns can show a strong ROAS while losing money on every sale because blended ROAS does not account for COGS, FBA fees, referral fees, or SKU-level margin variation. As of August 2026, sellers are reporting increased scrutiny around true profit per channel after Amazon moved ad billing to balance deduction (effective August 1, 2026 for eligible accounts) and expanded Sponsored Products placements into off-Amazon creator inventory (effective August 10, 2026). These changes compress cash flow and introduce new attribution questions, making the gap between reported ROAS and actual profit more visible.

ROAS (Return on Ad Spend) divides attributed revenue by ad spend. A campaign spending $1,000 and generating $3,000 in sales reports 3x ROAS. POAS (Profit on Ad Spend) subtracts all variable costs (COGS, FBA fees, referral fees, fulfillment, returns) from revenue before dividing by ad spend. If that same $3,000 in sales carries $2,200 in total costs, POAS is $800 profit divided by $1,000 spend, or 0.8x. The campaign shows 3x ROAS but loses $200 per $1,000 deployed.

Blended ROAS across all campaigns or all SKUs hides individual losers. A seven-SKU catalog may show 4x blended ROAS while three SKUs run negative POAS and subsidize the winners. Without SKU-level margin calculation, you scale spend into products that destroy capital.

How to calculate margin-adjusted POAS for Amazon Sponsored Products

POAS requires five inputs per SKU: ad spend, attributed revenue, COGS (landed cost including inbound freight and prep), Amazon fees (referral + FBA + monthly storage), and returns/damage allowance. The formula is:

POAS = (Attributed Revenue - COGS - Amazon Fees - Returns Allowance) / Ad Spend

Step-by-step:

Pull Sponsored Products campaign data for the SKU over the evaluation period (30 or 60 days). Record ad spend and attributed sales.
Calculate landed COGS per unit: factory price + freight + prep + customs + quality control. Multiply by units sold from ad-attributed orders.
Calculate Amazon fees per unit: referral fee (typically 15% of sale price) + FBA fulfillment fee (varies by size tier) + monthly storage (prorated). Multiply by units sold.
Estimate returns and damage: apply a 2-5% allowance based on category norms or historical return rate. Subtract this from gross margin.
Subtract total costs from attributed revenue, then divide by ad spend.

Example: A supplement sells for $29.99. Sponsored Products spend is $500, attributed sales are $1,799.40 (60 units). Landed COGS is $8.50 per unit ($510 total). Referral fee is 15% ($269.91). FBA fulfillment is $4.12 per unit ($247.20). Monthly storage is $12 prorated over 60 days ($12). Returns allowance is 3% of revenue ($53.98). Total costs are $510 + $269.91 + $247.20 + $12 + $53.98 = $1,093.09. Profit is $1,799.40 - $1,093.09 = $706.31. POAS is $706.31 / $500 = 1.41x. ROAS is $1,799.40 / $500 = 3.6x. The campaign looks strong on ROAS but only returns $1.41 for every ad dollar when costs are included.

If COGS were $12 per unit instead of $8.50 (a common result when cloning competitive products with no pricing power), total COGS rises to $720, total costs become $1,303.09, and profit drops to $496.31. POAS falls to 0.99x, meaning the campaign loses money despite 3.6x ROAS.

This is why launching products with genuine unmet demand and pricing power matters more than optimizing bids. Products with thin margins cannot survive profitably even at high ROAS.

Why cloned bestsellers make profitable ad spend nearly impossible

Cloning existing bestsellers compresses margin in three ways. First, buyers compare directly to the original, limiting price premium. You cannot charge $34.99 when the leader charges $29.99 without a meaningful differentiation story. Second, supply chains optimize for the leader’s volume, not yours, so your landed COGS is higher. A factory quoting $6.80 per unit at 10,000 MOQ for the bestseller will quote you $9.20 at 1,000 MOQ. Third, you compete in the same ad auction as the leader and every other clone, driving CPCs higher without improving conversion.

The result is a margin squeeze visible in POAS. A product with $8.50 landed COST, $4.12 FBA fee, and 15% referral fee needs to sell above $26 just to break even before ad spend. At $29.99 retail, gross margin is $29.99 - $8.50 - $4.50 (referral) - $4.12 = $12.87, or 42.9%. If CPC averages $1.20 and conversion rate is 12%, cost per order is $10. POAS at that efficiency is ($12.87) / $10 = 1.29x. Workable but fragile. If CPC rises to $1.50 (common after Amazon’s July 2026 relevance update, which increased bid pressure), cost per order becomes $12.50 and POAS drops to 1.03x. A 25% CPC increase erases profitability.

Products discovered through unmet demand research face less auction competition and carry pricing power because they solve problems the market is paying for but nobody ships yet. Higher retail price and lower CPC create POAS headroom. A $39.99 product with the same cost structure has $17.37 gross margin per unit. At $1.50 CPC and 12% conversion, POAS is $17.37 / $12.50 = 1.39x, 35% better than the clone despite identical ad efficiency.

For a detailed breakdown of how margin and ad cost interact at launch, see Product Research ROI: Real Ad Costs and Margin Calculation.

SKU-level profitability frameworks: tagging winners and killing losers

Blended ROAS obscures which SKUs fund growth and which destroy it. A SKU-level framework tags every product as a winner (POAS > 1.5x), sustainer (POAS 1.0–1.5x), or loser (POAS < 1.0x) and allocates budget accordingly.

Winners get increased ad spend, expanded keyword coverage, and placement priority. These SKUs generate enough profit per order to fund customer acquisition and scale. Sustainers receive maintenance spend to hold rank and capture organic-adjacent demand but do not receive expansion budget. Losers are paused or restructured (price increase, cost reduction, creative overhaul) within 30 days. Continuing to spend on negative-POAS SKUs is a choice to subsidize lost sales with winning-SKU profit.

Implement this in practice:

Export Sponsored Products performance by SKU for the trailing 60 days.
Join cost data (COGS, fees, returns allowance) to each SKU.
Calculate POAS per SKU using the formula above.
Segment SKUs into winner/sustainer/loser tiers.
Reallocate budget: shift 80% of incremental spend to winners, 15% to sustainers, 5% to testing new SKUs. Pause losers unless a specific fix (price change, cost renegotiation, creative refresh) is scheduled.
Re-evaluate monthly. SKUs move between tiers as cost structure, pricing, and conversion improve or degrade.

This approach prevents the common mistake of scaling total budget when blended ROAS looks strong while three SKUs burn capital. For brands managing multiple products, the six-figure Amazon brand economics model shows how SKU-level profit management compounds into sustainable growth.

How Amazon’s August 2026 placement changes affect POAS visibility

Amazon extended Sponsored Products into off-Amazon creator placements starting August 10, 2026. Campaigns now serve ads in Amazon Influencer Program content outside Amazon’s domain. Early seller feedback on Reddit and Amazon forums describes this traffic as lower-intent and poorly converting, with some reporting the expansion happened without clear opt-in.

Off-Amazon placements introduce two POAS risks. First, attribution windows may not align with buyer behavior. A user clicking a creator link and purchasing 10 days later may fall outside the 7-day attribution window (standard for seller accounts) or be attributed incorrectly if the user also clicked an on-Amazon ad. Second, conversion rates on off-Amazon placements are typically 30–50% lower than on-Amazon search because intent is weaker. If CPC remains constant but conversion falls, cost per order rises and POAS drops.

Sellers should segment on-Amazon and off-Amazon performance separately. Amazon’s campaign reporting now includes placement-level data under the Advertised Product Report. Filter by placement type (Amazon search, product pages, off-Amazon) and calculate POAS for each. If off-Amazon POAS is negative, exclude those placements using campaign-level negative placement targeting or adjust bids downward by 40–60% to reflect lower conversion.

The billing change (ad costs deducted from proceeds starting August 1, 2026) also affects POAS indirectly by compressing working capital. Sellers previously floated ad spend on credit cards and captured 1–2% rewards plus 30-day payment terms. Now ad costs deduct immediately from available balance, reducing flexibility to test and scale winners quickly. Budget allocation becomes more rigid. POAS frameworks matter more when capital is constrained because reallocating from losers to winners is the only lever left.

Common ROAS traps that hide unprofitable campaigns

Five ROAS traps create the illusion of profitability:

1. Branded keyword ROAS inflation. Campaigns bidding on your own brand name capture demand you already own. A 6x ROAS on branded terms may represent zero incremental sales. Branded and non-branded campaigns should report POAS separately. Only non-branded POAS reflects true acquisition cost.

2. Attribution window mismatch. Amazon defaults to 7-day click attribution for sellers and 14-day for vendors. Comparing your 7-day ROAS to a competitor’s 14-day benchmark overstates relative performance. Longer windows capture more attributed sales and inflate ROAS. POAS calculation should match the attribution window used in reporting.

3. Ignoring organic halo. Sponsored Products campaigns improve organic rank, which drives non-attributed sales. A campaign may show 2.5x ROAS but generate an additional 1.5x in organic sales from the same buyers. Blended POAS (total profit from SKU / total ad spend on SKU) captures this better than campaign-only POAS, but requires tracking total SKU revenue alongside ad-attributed revenue.

4. Campaign-level averages over time. A campaign running 90 days may show 3x ROAS overall but 1.8x ROAS in the last 30 days due to CPC increases or conversion decay. Trailing 30-day POAS is more actionable than lifetime average.

5. Cross-SKU blending within campaigns. A Sponsored Products campaign containing five SKUs reports one ROAS number. If two SKUs carry 60% margin and three carry 25% margin, blended ROAS hides the fact that three SKUs lose money. Split SKUs into separate campaigns or ad groups to isolate POAS per product.

For operators managing multiple products or brands, Kaldon’s Discover phase surfaces unmet-demand opportunities with margin and pricing assumptions built in, preventing launch into categories where POAS profitability is structurally impossible.

What to do when ROAS is strong but POAS is negative

If a campaign shows 3x+ ROAS but POAS below 1.0x, the product cannot support its ad cost at current pricing and cost structure. Four fixes exist:

1. Increase price. Raise retail price by 10–15% and monitor conversion. If conversion drops less than 10%, gross margin increases enough to recover POAS. This works best for products with differentiation or where competitors are also priced low.

2. Reduce COGS. Renegotiate factory pricing, consolidate shipments to lower freight, or remove unnecessary packaging. A $1 COGS reduction on a $9 landed cost product improves gross margin by 11% and can shift POAS from 0.9x to 1.2x.

3. Lower CPC via better creative and targeting. Improved main image, title, and bullet points increase click-through rate, which lowers CPC in the auction. Tighter keyword targeting (exact match on high-intent terms, negative match on low-converters) reduces wasted impressions. A 20% CPC reduction directly improves POAS by the same percentage.

4. Pause and reallocate. If price cannot increase, cost cannot decrease, and creative is optimized, the SKU is not viable for paid acquisition. Pause Sponsored Products, shift budget to winning SKUs, and rely on organic rank or liquidate inventory.

Most negative-POAS situations trace back to launching the wrong product (cloned bestseller with no pricing power) rather than poor campaign execution. The fastest fix is to launch better products. Kaldon’s 5-phase workflow discovers unmet demand, validates margin, and generates content in one platform, replacing the 6+ tool stack most sellers use. Start a free trial to see how unmet-demand research prevents negative-POAS launches.

Benchmark POAS by category and what strong looks like in 2026

POAS benchmarks vary by category based on margin structure and competitive intensity. Recent 2026 data shows blended eCommerce ROAS around 2.87x and Amazon-specific ROAS averaging 3.4x, but POAS benchmarks are lower because they account for costs.

Strong POAS by category:

Supplements / consumables: 1.5–2.5x POAS. High repeat rate and subscription potential justify lower per-order profit. LTV matters more than first-order POAS.
Home / kitchen: 1.3–1.8x POAS. Competitive but decent margins if differentiated. Clones struggle to break 1.2x.
Electronics / accessories: 1.1–1.5x POAS. Thin margins and high return rates compress profitability. Premium positioning required.
Apparel / private label fashion: 1.2–1.6x POAS. Margin depends on branding strength. Generic items rarely exceed 1.1x.
Baby / pet: 1.4–2.0x POAS. Emotional purchase decisions support higher pricing and lower CPC if messaging is strong.

Any SKU below 1.0x POAS is unprofitable and should be paused or fixed. SKUs above 2.0x POAS in competitive categories signal strong product-market fit and pricing power, and deserve maximum budget allocation.

These benchmarks assume standard FBA fee structure and 7-day attribution. Off-Amazon placements, longer attribution windows, and seller-fulfilled logistics will shift the numbers.

Why POAS matters more than ROAS for long-term brand economics

ROAS measures revenue efficiency. POAS measures profit efficiency. Revenue does not pay salaries, inventory invoices, or platform fees. Profit does.

A brand scaling to six figures on strong ROAS but weak POAS will hit a capital wall. Ad spend increases, revenue increases, but cash flow stays flat or negative because each order generates insufficient profit to reinvest. The business looks healthy in top-line metrics (sales up 40% year-over-year) but cannot fund inventory for the next order or hire the next operator.

POAS-first thinking forces better product selection, better pricing strategy, and better cost discipline. It prevents scaling into unprofitable SKUs and redirects budget toward products that actually fund growth. For a full breakdown of how unit economics compound into brand-level cash flow, see six-figure Amazon brand economics.

Sellers building on unmet demand rather than cloning bestsellers start with structural POAS advantages (higher price, lower CPC, better conversion) and scale faster with less capital. That is the core insight behind Kaldon’s Discover phase: find the demand gap, validate the margin, and launch with pricing power built in. See how it works.

Frequently asked questions

What is the difference between ROAS and POAS for Amazon Sponsored Products?

ROAS (Return on Ad Spend) divides attributed revenue by ad spend and does not account for costs. POAS (Profit on Ad Spend) subtracts COGS, FBA fees, referral fees, and returns from revenue before dividing by ad spend, showing actual profit per ad dollar. A campaign can have 3x ROAS but 0.8x POAS if costs exceed 73% of revenue.

How do I calculate POAS for a Sponsored Products campaign?

Pull ad spend and attributed sales from campaign reports. Calculate total costs: COGS per unit times units sold, plus Amazon fees (referral + FBA + storage), plus a 2-5% returns allowance. Subtract total costs from attributed revenue, then divide by ad spend. The formula is (Revenue - COGS - Fees - Returns) / Ad Spend.

Why does cloning bestsellers make it harder to achieve profitable POAS?

Cloned products face direct price comparison to the leader, higher COGS due to lower MOQ, and higher CPCs from auction competition, compressing margin. A cloned product with thin margin may only achieve 1.0–1.2x POAS even at high ROAS, while a differentiated product with pricing power can sustain 1.5–2.0x POAS at the same ad efficiency.

What is a strong POAS benchmark for Amazon Sponsored Products in 2026?

Strong POAS varies by category. Supplements and consumables typically achieve 1.5–2.5x POAS. Home and kitchen products run 1.3–1.8x. Electronics and accessories range from 1.1–1.5x due to thin margins. Any SKU below 1.0x POAS is unprofitable and should be paused or restructured.

How do Amazon’s August 2026 off-Amazon placements affect POAS?

Off-Amazon creator placements (live August 10, 2026) show lower conversion rates than on-Amazon search, increasing cost per order and lowering POAS. Sellers should segment placement performance separately and exclude or reduce bids on off-Amazon placements if POAS is negative. Early reports describe this traffic as lower-intent and poorly converting.

Sources & citations

amazon-advertisingroas-calculationprofit-metricssponsored-productsmargin-analysis

Last updated Aug 9, 2026

Try Kaldon

See the pipeline for yourself.

Start with 2 free analyses. No credit card required.

Start Free