ACOS (Advertising Cost of Sales)

ACOS (Advertising Cost of Sales) measures how much of the revenue generated by an Amazon advertising campaign was spent on ads. It's expressed as a percentage:

ACOS = (Ad Spend / Ad Sales) × 100

A campaign that spent €20 to generate €100 in sales has an ACOS of 20%. Lower ACOS means more efficient spend — but "lower" is not automatically "better." Pushing ACOS too low usually means starving the campaign of impressions and leaving incremental sales on the table. The right ACOS is the one that hits the profit goal for the SKU, not the lowest number you can drive on a dashboard.

ACOS, TACOS, and ROAS — getting the definitions straight

The terminology around advertising efficiency is sloppy across the industry. Pin the definitions down before optimizing:

• ACOS — Ad Spend / Ad-Attributed Sales. Campaign-level efficiency. Reported per campaign, ad group, target, search term, and placement.
• TACOS (Total ACOS) — Ad Spend / Total Sales (organic + ad). Brand-level efficiency. The metric that actually correlates with P&L health. A declining TACOS at a constant or growing total revenue is the signature of a brand whose organic flywheel is working.
• ROAS (Return on Ad Spend) — Ad Sales / Ad Spend. The mathematical inverse of ACOS, expressed as a multiple. ACOS 25% = ROAS 4.0×. Common outside Amazon, less common inside.
• Target ACOS — see Target ACOS. The ACOS ceiling derived from product margin; what you plan against, not what you measure.

A campaign report showing 35% ACOS at the line level can still be acceptable if TACOS at the brand level is 12% — because most of the revenue is organic and the ads are functioning as rank fuel.

Why ACOS depends entirely on the attribution window

ACOS is not a property of a campaign. It is a property of a campaign as measured through a specific attribution window. Change the window and the same campaign produces a different number:

• 1-day attribution — only sales completed within 24 hours of click. Lowest reported sales, highest reported ACOS.
• 7-day attribution — Seller default. Captures most impulse and short-consideration purchases.
• 14-day attribution — Vendor default. Captures mid-consideration purchases.
• 30-day attribution — Captures long-consideration purchases, payday-driven carts, and seasonal stockup behavior.

Wider windows always show lower ACOS for the same campaign, because more downstream conversions are credited back to the click. A product where customers research for two weeks before buying (think robot vacuums, mattresses, premium kitchen appliances) will look unprofitable at 7-day attribution and highly profitable at 30-day attribution — even though nothing about the campaign itself changed.

ACOS by ad type — and why apples-to-apples comparisons fail

The reported ACOS number means subtly different things across ad types:

• Sponsored Products — ACOS is calculated on the advertised ASIN and on other ASINs in the brand's catalog that the same click drove to purchase ("halo" or "brand halo" sales). What's reported in Seller Central is typically advertised-ASIN-only; the halo lives in the Brand Analytics or AMC reports.
• Sponsored Brands — ACOS is calculated on all branded ASINs purchased after the click, not just the click target. SB ACOS therefore looks better than SP ACOS on the same keywords, even if SB is less efficient on a pure per-click basis.
• Sponsored Display — ACOS includes view-through conversions (the user saw the ad but didn't click), which inflates reported sales attribution vs. SP. View-through ACOS and click-through ACOS are very different animals.

Comparing ACOS across ad types without normalizing for attribution model is one of the most common mistakes in campaign reviews.

The relationship between ACOS, CVR, and Max CPC

ACOS is mathematically a function of two operational levers — Conversion Rate (CVR) and Cost Per Click (CPC) — and one fixed input: the product's average selling price (ASP).

ACOS = CPC / (ASP × CVR)

Rearranged, the maximum CPC you can afford at a given target ACOS is:

Max CPC = ASP × CVR × Target ACOS

This is the formula every disciplined bid optimizer runs internally. It explains the most common ACOS failure modes:

• ACOS keeps drifting up. CVR dropped (listing change, price change, review hit, stockout). Lower bids until CVR recovers.
• ACOS is too low and growth is flat. CPC ceiling is too tight relative to CVR — bid higher to take more impressions while staying under target.
• ACOS is fine on the campaign but margin is gone. Halo sales the campaign isn't credited for plus return rate isn't priced into target ACOS.

What "good" ACOS looks like

There is no universal benchmark. A defensible target ACOS depends on:

• Gross margin of the SKU after COGS, FBA fees, returns reserve, and overhead allocation.
• Lifecycle stage. Launch campaigns intentionally run above target ACOS to buy rank. Mature campaigns should run at or below target.
• Strategic role. Defensive branded campaigns (your own brand keywords) often run at 5–10% ACOS and exist to suppress competitor conquest. Top-of-funnel category campaigns may run at 60–80% ACOS and exist to feed search-term harvesting.

A healthy account doesn't run a single ACOS target. It runs a portfolio of ACOS targets, each tied to the role of the campaign in the funnel.

Common ACOS mistakes

• Treating ACOS as the only KPI. It ignores incremental sales, organic lift, and brand-building value.
• Optimizing ACOS at the campaign level instead of search-term level. A campaign at 25% ACOS often contains terms at 8% and terms at 90% — the average hides the work.
• Using the wrong attribution window for the product. Diapers ≠ robot vacuums.
• Forgetting halo and TACOS. Cutting an "expensive" campaign that was actually driving most of the organic sales is one of the most expensive mistakes in PPC.
• Comparing ACOS between ad types without normalizing. SP, SB, and SD attribute differently — by design.