How to Lower Your Amazon ACoS Without Killing Sales

Most sellers try to lower ACoS by cutting bids across the board. Sales drop, rank slips, and a month later the ACoS is back where it started — on less revenue. I've rebuilt enough accounts to know the problem is almost never "bids too high." It's spend leaking into clicks that never convert.

Here's the order of operations I run on real accounts, including one doing $300–400K/yr in ad spend.

Step 1: Cut the wasted spend (the fastest win)

Pull your search term report for the last 60 days. Sort by spend, and look for search terms with 10+ clicks and zero orders. Ten clicks is my minimum before judging a term — below that you're reacting to noise.

Every one of those terms gets a negative exact. This is the highest-leverage fifteen minutes in PPC: you remove pure loss without touching anything that converts. On messy accounts, this step alone routinely takes several points off ACoS.

Here's what this looks like in practice. Say you run a broad match campaign on "kitchen storage rack." The search term report shows the ad fired on "cheap kitchen rack kids" — 14 clicks, $18.20 spent, zero orders. Negative exact it. Do that across 20–30 terms in one session and you've typically freed $150–400 in weekly spend with zero impact on sales. On a $300–400K/yr account with 400–500 campaigns, the first 60-day search term audit alone can cut several points off ACoS before you touch a single bid.

Step 2: Fix bids by placement, not by feeling

Amazon reports performance by placement — top of search, product pages, rest of search. The same keyword can be profitable at top of search and a disaster on product pages.

Check the placement report before cutting any keyword. Often the fix isn't "lower the bid" — it's removing a placement multiplier that's paying a premium for the placement that doesn't convert. I run dynamic bids, down only on most campaigns, so Amazon can reduce bids on low-probability auctions but never inflate them.

Here's the actual walkthrough. In Campaign Manager, go to Reports, create a Placement report, and set it for the same 60-day window. Look at spend, orders, and ACoS broken out by placement. A common pattern looks like this:

The fix is not cutting the keyword bid — it's reducing the product page placement multiplier to zero or near-zero. The keyword is working at top of search. A blanket bid cut kills that too.

Step 3: Set your target from break-even, not from a benchmark

You can't know if 30% ACoS is good without knowing your margin. Break-even ACoS = (price − COGS − Amazon fees) ÷ price. Run your numbers through the break-even ACoS calculator — that's your ceiling. I set target ACoS around 70% of break-even so each sale keeps real profit.

Now every keyword decision is mechanical: above break-even with enough clicks = cut or fix. Between target and break-even = optimize. Below target = scale.

Quick example. A product sells for $35. COGS $12, Amazon fees roughly $8. That leaves $15 per unit. Break-even ACoS = $15 ÷ $35 = 42.8%. Target ACoS = 42.8% × 70% = 30%. Any keyword above 43% is a guaranteed loss. Anything between 30–43% needs review. Below 30% is where you scale.

One account I managed from January to August 2024 ran $234,583 in ad sales on $55,746 spend — 23.76% ACoS, 4.21 ROAS. That's what eight months of staying disciplined on break-even targets looks like. Not a lucky month.

Step 4: Move spend into what converts

Lowering ACoS isn't only subtraction. Take the budget you freed and put it behind proven converters — the search terms with strong order history. Harvest them from auto and broad campaigns into exact match campaigns where you control the bid precisely.

That's the part most "lower your ACoS" advice misses: the goal is a lower ACoS on more revenue, not a prettier percentage on a shrinking account. ACoS down, revenue up — in that order, but both.

The harvest process: pull auto and broad search terms that have three or more orders and a click-through ACoS under your target. Add those as exact match keywords in a separate campaign. Then negative-exact those same terms in the source auto or broad campaign — this forces spend toward the controlled exact match where you set the bid precisely.

Opening exact bids? I start around 2x Amazon's suggested bid. That buys early visibility without getting buried, and I adjust down based on actual performance after two to three weeks of data.

Order of operations at a glance

Common objections

"My ACoS is fine, I just need more sales." Run the search term report anyway. Clean accounts scale faster. If you're funding terms with 20+ clicks and zero orders, that money isn't buying growth — it's buying noise.

"I've already got negatives in place." How old are they? Search behavior drifts. I re-run the 60-day audit at minimum every 6–8 weeks. Terms that looked borderline three months ago may have accumulated enough data to confirm they're losers.

"Dynamic bids down-only will hurt my visibility." Down-only doesn't reduce your maximum bid. It lets Amazon apply a discount on low-probability auctions. Your top-of-search exposure stays intact at the bid you've set. The keyword still competes at the same baseline — Amazon just doesn't inflate it where conversion probability is low.

What improvement actually looks like week by week

Weeks 1–2: Run the search term audit, add negatives, pull the placement report, fix multipliers. You'll see less wasted spend in the next reporting cycle. ACoS may tick down a few points.

Weeks 3–4: Harvest high-performing search terms into exact campaigns. Initial data comes in. You're now funding winners at a bid you control.

Weeks 5–8: Exact campaigns start taking over spend share from broad and auto on your best terms. ACoS compresses further. You can begin cutting bids on keywords with eight or more orders sitting above break-even.

Month 3 and beyond: The structural work is done. Maintenance is monthly search term audits, quarterly bid reviews on exact campaigns, and scaling anything that's sitting below target ACoS. This is where the gap widens between accounts managed on data and accounts managed on instinct.

What not to do

• Don't blanket-cut bids. You'll starve winners and losers equally.
• Don't judge keywords on fewer than 10 clicks. That's noise, not data.
• Don't pause everything above break-even instantly. Check placement splits first — many "bad" keywords are one placement adjustment from profitable.

If you want this run on your account, that's literally what the free audit is: I go through your campaigns, find the wasted spend, and send you a short video with the three highest-leverage fixes.