What this page is, and isn't — read this before the numbers
ROAS and ACOS benchmarks are the single fastest-moving, most context-dependent figures on this site, which is exactly why this page is tagged Continuous rather than Quarterly or Evergreen: a number that's reasonable today can be stale within a single ad-platform algorithm change or a seasonal competitive shift. This page exists to feed Step 2 of the Benchmark Library framework and should always be read alongside ROAS vs. ACOS vs. TACOS, which explains why a single ROAS or ACOS figure in isolation — without knowing your break-even point — tells you almost nothing about whether a campaign is actually working. Your own break-even ACOS, calculated from your real margin (see Break-Even ROAS, Explained), is always a more meaningful target than any external category figure.
Overall ACOS/ROAS ranges
Trellis's March 2026 aggregated benchmark report puts average Amazon Sponsored Products ACOS at roughly 30%, with a commonly cited "healthy" target range of 20-40% depending on margin structure and growth stage, and a general ROAS range of 2.5x-5.0x depending on campaign objective (defense/retention campaigns typically run more efficient than aggressive-growth or new-launch campaigns). By ad type, the same source reports:
| Ad type | Typical ACOS range |
|---|---|
| Sponsored Products | 20-35% |
| Sponsored Brands | 25-40% |
| Sponsored Display | 30-50% |
ROAS by category
Triple Whale's August 2026 update (median performance across 2,800+ brands, August 2025-July 2026) reports the following category-level medians. Note this dataset reports ROAS, CPA, CVR and CTR rather than ACOS directly — a rough ACOS-equivalent can be approximated as 1/ROAS (e.g. a 3.5x ROAS is roughly a 29% ACOS-equivalent), though this approximation ignores differences in how each metric handles shipping, tax, and multi-touch attribution, so treat it as a rough translation, not an exact conversion.
| Category | ROAS (median) | Approx. ACOS-equivalent | Conversion rate |
|---|---|---|---|
| Electronics | 3.93x | ~25% | 5.04% |
| Home & Garden | 3.54x | ~28% | 7.14% |
| Sports & Outdoors | 3.56x | ~28% | 6.89% |
| Apparel & Accessories | 3.46x | ~29% | 6.84% |
| Toys, Art & Collectibles | 3.44x | ~29% | 7.47% |
| Pets & Animals | 2.74x | ~36% | 16.33% |
| Food & Beverage | 2.73x | ~37% | 14.96% |
| Beauty | 2.72x | ~37% | 14.10% |
| Health & Wellness | 2.46x | ~41% | 11.56% |
Read the pattern here carefully rather than just the ROAS column: the categories with the lowest ROAS (health & wellness, beauty, food & beverage, pets) tend to have the highest conversion rates — these are typically lower-price-point, higher-frequency-purchase, more habitual-buy categories where advertisers can tolerate a lower ROAS per sale because purchase frequency and repeat-purchase potential (see the Benchmark Library's note on this) make the customer, not just the single order, the more relevant profitability unit. Electronics and home & garden show the opposite pattern: higher ROAS, lower conversion rate, consistent with higher-consideration, less frequent purchases where each ad dollar needs to work harder per click to be justified.
A caveat about what these numbers actually measure
Both cited sources aggregate advertiser-platform data (Trellis, Triple Whale) rather than a single official Amazon-published benchmark — Amazon itself does not publish category-level ACOS or ROAS benchmarks. That means both tables above reflect the mix of brands using those specific third-party tools, not a random or comprehensive sample of all sellers in each category. Treat the relative pattern across categories (which ones run more efficient, which run less) as more reliable than the exact absolute figures, which will vary by the underlying tool's customer base.
How to use this
- Calculate your own break-even ACOS from your real contribution margin first — see Break-Even ROAS, Explained — before comparing yourself to any category figure above.
- If your category has a naturally high repeat-purchase rate (consumables, supplements, pet food), consider whether a lower first-order ROAS is still justified by lifetime value — see CAC and LTV for Ecommerce — rather than optimizing purely to a category-average ROAS that doesn't account for repeat purchases.
- Track TACOS (not just ACOS) over time per ROAS vs. ACOS vs. TACOS — a category benchmark for ACOS alone can't tell you whether your advertising is also lifting organic sales, which is often the more durable signal.
- Re-check the live source before a real budget decision — this page is tagged Continuous specifically because ad-cost benchmarks are the fastest-moving figures on this site.
FAQ
Why doesn't this page give a single "good ROAS" number? Because "good" depends entirely on your break-even point, which depends on your margin — a 2.5x ROAS can be very profitable on a high-margin product and unprofitable on a thin-margin one. See Break-Even ROAS, Explained for the calculation that actually answers this for your specific product.
Why do low-ROAS categories (beauty, food & beverage, health) show up as "efficient" advertisers in this data? Because ROAS on a single order understates value in categories with strong repeat-purchase behavior — an advertiser in these categories can rationally accept a lower first-order ROAS if the value of a retained customer over time (see CAC and LTV for Ecommerce) more than makes up for it.
Does Amazon publish its own category ACOS/ROAS benchmarks? Not as of this writing — every category figure on this page comes from third-party platforms aggregating their own advertiser data, not an official Amazon source. Treat the relative pattern across categories as more trustworthy than the exact absolute numbers.