My Amazon ACoS Is Too High — How to Fix It

If your Amazon ACoS is too high and climbing, you have one of two problems — and the fix for one makes the other worse. Either your campaigns are structurally leaking money on clicks that were never going to convert, or your campaigns are fine and your listing is failing to convert the traffic they buy. Sellers who guess wrong burn months cutting bids on a conversion problem or rewriting bullets to fix a targeting problem. The diagnosis takes about an hour with two reports you already have: the search term report and Brand Analytics. This page walks the decision tree — what to pull, what thresholds separate the branches, and the fix path for each — so you stop treating the symptom and fix the actual leak.

First: Know Your Break-Even ACoS

“Too high” is meaningless without a reference point, and the reference point is not a category benchmark or a number your last agency picked. It’s your break-even ACoS: the profit margin left after product cost, freight, FBA fees, and the 8-15% referral fee, before ad spend.

The math: a $30 product with $9 landed cost, $6 FBA fulfillment, and a $4.50 referral fee leaves $10.50 — a 35% margin. Your break-even ACoS is 35%. Every point below it is profit on ad-attributed sales; every point above is buying revenue at a loss. Sellerboard or a proper profitability analysis will give you this per ASIN in minutes, and it varies enough across a catalog that a single account-wide target is usually wrong.

Two situations where ACoS above break-even is acceptable on purpose: launches, where you’re buying rank and reviews, and defense of top organic positions that generate profitable organic volume. That’s why ACoS should never be judged without its sibling metric — if organic sales are growing while ACoS runs hot, your blended picture may be healthy. The full breakdown of when each metric matters is in our guide to ACoS vs TACoS. If your ACoS has been above break-even for 60+ days with no launch or rank-defense justification, you have a real problem. Diagnose it.

The Core Diagnostic: Campaign Problem or Conversion Problem?

ACoS is spend divided by ad revenue, which decomposes into exactly two levers: what you pay for clicks, and how many clicks become orders. So every high-ACoS account falls into one of two branches:

Branch A — campaign structure problem. Your ads buy the wrong clicks, or pay too much for the right ones. Symptoms live in the search term report: irrelevant queries eating spend, no negative keywords, broad match running unsupervised, bids set once and never revisited.

Branch B — listing conversion problem. Your ads buy reasonable clicks at reasonable prices, but the listing converts them below the category norm, so every order carries the cost of too many wasted clicks. Symptoms live in Brand Analytics and your business reports: conversion rate well under the category median, weak reviews, a price out of position.

The branches demand opposite responses. Branch A is fixed inside the ad console. Branch B cannot be fixed inside the ad console — cutting bids on a Branch B account just shrinks a losing operation, and scaling spend on it scales the loss.

Misdiagnosis is expensive in both directions. Treat a Branch B account like Branch A and you’ll spend three months negating keywords and trimming bids while the real problem — a listing converting at half the category rate — keeps taxing every click you buy. Treat a Branch A account like Branch B and you’ll rewrite bullets and reshoot images while an unsupervised broad match campaign keeps torching 30% of the budget on queries that have nothing to do with your product.

Before the Tree: Segment, Don’t Average

One preliminary step that saves a lot of wrong turns: never diagnose from account-level ACoS. It’s an average, and averages hide the distribution that actually matters.

A typical account showing 42% blended ACoS breaks down like this when you segment by campaign and then by search term: a core of exact match keywords running at 18%, a middle tier around break-even, and a long tail of auto and broad spend running at 90%+ that drags the whole number up. That account doesn’t need a strategy overhaul — it needs the bottom tier cut. A different account showing the same 42% might be uniformly mediocre across every term, which points at conversion, not structure.

Segment three ways before judging: by campaign type (auto vs. broad vs. exact), by ASIN (margin and CVR vary — one bleeding product can hide six healthy ones), and by placement (top-of-search vs. rest-of-search often convert at wildly different rates). Fifteen minutes of segmentation tells you whether you’re hunting a concentrated leak or a systemic one.

The Decision Tree, With Thresholds

Run these checks in order. Check the search term report first — it’s the fastest signal and the most common culprit.

Step 1: Pull 60 days of search term reports (Sponsored Products, all campaigns) and sort by spend descending.

• Is 20%+ of total spend on search terms with zero orders? → Branch A. In an unmanaged account this figure is often 30-40%.
• Are irrelevant queries in your top 20 spenders? You sell stainless steel water bottles and you’re paying for “kids plastic sippy cup” via an auto campaign or careless broad match → Branch A.
• Is one search term appearing in multiple campaigns at different bids, competing against yourself with no single-campaign ownership? → Branch A.

Step 2: Check your conversion rate against category reality. Pull unit session percentage from Business Reports for your top ad-driven ASINs, and compare against Brand Analytics — the Search Query Performance dashboard shows your conversion share versus the market on your actual query terms.

• Is your CVR below ~66% of the category median? (Category median is 12% and you’re at 7%?) → Branch B, even if your search terms look clean.
• Is your CVR fine (at or above median) but ACoS still high? → Back to Branch A: the problem is click cost, not conversion. Look at bids and placement multipliers.

Step 3: If both look bad, fix Branch A first. Structural waste is faster to remove, and you can’t read conversion data clearly while a third of your traffic is irrelevant — junk clicks depress measured CVR and contaminate the Branch B diagnosis.

One more gate before you touch anything: check whether ACoS rose because spend rose or because sales fell. If ad revenue and organic revenue dropped together, you may be looking at a lost Buy Box, a suppressed listing, or a rank collapse — that’s not a PPC tuning problem, and it’s covered in our sales dropped emergency guide. A denominator collapse will masquerade as an ACoS spike and waste weeks of campaign tinkering.

Fix Path A: Campaign Structure

Work these in order of speed-to-impact.

1. Negate the proven losers — today. Any search term with 20+ clicks (or 2x your product price in spend) and zero orders becomes a negative exact. Do this account-wide. This is the single fastest ACoS reduction available and it costs zero sales, because you’re cutting spend that never converted.

2. Set up the harvest-and-negate cycle. Weekly: converting search terms from auto and broad campaigns graduate into exact match campaigns where you control the bid; the queries that keep missing get negated at the source. Auto and broad campaigns are discovery engines, not permanent homes for spend. If terms have been converting in an auto campaign for a year without graduating, you’ve been paying discovery prices for known winners.

3. Fix match type architecture. Proven terms in exact match with deliberate bids. Broad match on a short leash — modest bids, tight negative lists, reviewed weekly. One campaign owns each hero keyword, so you’re not bidding against yourself.

4. Rework bids from data, not defaults. Sort spend descending: any term above break-even ACoS over 60 days gets a bid cut sized to the gap (a rough rule: new bid = current bid x target ACoS / actual ACoS). Check placement reports — if top-of-search converts at triple the rest-of-search rate, shift budget there via placement multipliers instead of raising base bids everywhere. Then look at bid optimization as an ongoing weekly discipline rather than a one-time repair.

5. Check the bidding strategy setting itself. “Dynamic bids — up and down” gives Amazon license to double your bid on placements it predicts will convert, and its predictions serve Amazon’s revenue before yours. On campaigns with thin data or high ACoS, switch to “dynamic bids — down only” until performance stabilizes. It’s a two-minute change that routinely shaves several points of ACoS on struggling campaigns.

What not to do: the panic move of cutting every bid 30% account-wide. Blanket cuts hit your profitable placements and your wasteful ones equally, so sales fall alongside spend, ACoS barely improves, and you lose organic rank on the keywords where ad velocity was doing real work. Every cut in this fix path is surgical and targeted at spend that’s demonstrably not converting.

Run this sequence and expect the wasted-spend share to fall visibly within 30 days. This is exactly the structural work behind the TACoS 28% to 11% turnaround in our case studies — most of the early gains came from negation and architecture, not clever bidding.

Fix Path B: Listing Conversion

If the search term report is clean and CVR is the outlier, more ad-console effort is wasted effort. Find which conversion lever is broken:

Price position. Open an incognito search for your main keyword. If you’re priced 15%+ above the page median without the reviews or brand strength to justify it, that’s your CVR gap. Check Keepa for competitor pricing history — you may be the only one who didn’t move.

Review deficit. Under ~50 reviews or below 4.2 stars while page-one competitors hold thousands at 4.5+ means traffic will comparison-shop away from you no matter what the bullets say. Vine enrollment and review velocity work come before ad scaling.

Content and imagery. Main image quality, A+ Content, and title relevance to the queries actually sending traffic. Test with Manage Your Experiments rather than guessing. The full diagnostic — including mobile rendering and the order shoppers actually read a listing — is in our guide to listings that don’t convert.

The sequencing rule: fix conversion first, then re-scale spend. Every point of CVR improvement lowers ACoS across all traffic simultaneously — a listing that goes from 7% to 11% conversion just cut the ACoS on every keyword by a third with no bid changes at all.

Keeping It Fixed

High ACoS is rarely a one-time event; it’s what accounts drift toward without weekly search term hygiene and periodic conversion benchmarking. The maintenance cadence — weekly negation, monthly bid reviews against break-even by ASIN, quarterly CVR checks against Brand Analytics — is covered across our Amazon PPC management guides.

If you’d rather not own that cadence in-house, this diagnostic is literally the first week of our PPC management engagement: we pull your search term data, benchmark your conversion rates, tell you which branch you’re on, and show you the wasted-spend number before you commit to anything.