Amazon Negative Keywords: What They Are and How to Use Them

What Negative Keywords Actually Do

A negative keyword is a search term you tell Amazon not to match your ad against. You aren't removing a keyword you're bidding on — you're blocking a search query the shopper typed, so your ad never enters that auction. It's the opposite of a target: targets say show my ad here, negatives say never show my ad here.

Amazon gives you two negative match types — negative phrase and negative exact — set at the ad group or campaign level. There is no negative broad match on Amazon, which trips people up if they've run Google Ads, where negative broad exists. If you want wider coverage, negative phrase is the closest equivalent.

Negative keywords apply to keyword-targeted and auto campaigns. If you're running product targeting instead, the equivalent tool is negative product targeting — you exclude an ASIN or category, not a search string. Same purpose, different mechanism.

How to Find Negative Keywords Worth Adding

The only reliable source is your search term report, found in Campaign Manager under Sponsored Products or Sponsored Brands reporting. It shows the actual queries that triggered your ad, not the keyword you bid on — those are often different, especially under broad match and auto targeting.

Pull the report and sort for the pattern that matters: spend with no sales. Say a search term — 'kids yoga mat' — spent $86 last month across 14 clicks on a product built and marketed for adults, with zero orders. That's a clean negative-exact candidate. Multiply that pattern across an account and it adds up fast: if 8% of a $12,000 monthly budget sits in terms with zero conversions and enough clicks to call it a pattern, that's roughly $960 a month leaking into searches that were never going to buy.

Don't act on one click. A single click with no sale on a low-bid keyword is noise, not evidence. Let a term collect enough clicks to mean something for your typical conversion rate before you negate it. Competitor names and obviously mismatched intent — a budget brand's ad showing on a luxury search, or the reverse — are the exception; negate those on relevance alone, without waiting on spend data.

Negative Phrase vs Negative Exact Match

The two match types behave differently, and picking the wrong one either lets waste through or blocks traffic you wanted to keep.

Mistakes That Waste More Than They Save

• Negating on one bad day. One click, no sale, on a cheap keyword isn't a trend. Pulling it too early means giving up on a term that might convert fine at normal volume.
• Setting the negative at the wrong level. A negative added inside one ad group only blocks that ad group. If an auto campaign and a manual campaign can both match the same search term, blocking it in one leaves it live in the other.
• Using phrase match where exact was needed. A phrase negative on a root term can quietly block variants you actually wanted, because it catches anything containing that phrase in order.
• Negating a term you're also targeting elsewhere. If the exact phrase is a positive keyword in another ad group, the negative usually wins, and that ad group goes dark for a term it was supposed to win.

We've made the first one ourselves — pulling a term after a single expensive, no-sale click, only to watch it convert fine the following week once volume caught up. The fix isn't a rule of thumb. It's checking enough data before acting.

When Adding Negatives Doesn't Fix the Problem

Sometimes wasted spend doesn't drop after you add the negative. Before assuming it failed, check three things: level — did you block it everywhere it can appear, not just where you found it; match type — a phrase negative won't catch a variant with an inserted word, so you may need exact plus phrase, or a small family of negatives; and attribution lag — Amazon's reporting window can mean spend from before the negative went live is still settling into this week's numbers.

If ACOS still hasn't moved once the negative is confirmed live and the search term report shows the spend genuinely gone, the leak wasn't targeting — it's somewhere else. Bids too high on terms that are actually relevant, or a listing that isn't converting once the click lands. Negative keywords fix relevance. They don't fix a weak page.

Every proposed change carries three things before it runs: the evidence behind it, a measurement plan, and a rollback trigger. That standard applies just as well to a negative keyword as to a bid change — write down why the term is bad, decide how you'll check the effect, and know in advance what you'll do if removing it costs more than it saved.

Where This Fits With Automated Management

Negative keywords are one control among many an account needs pulled correctly and on schedule. Dr. PPC — built by Full Circle, which has managed more than $500M in revenue across 100+ brands — reads the search term report alongside everything else in the account and proposes the negative, with the evidence and a rollback trigger attached, for a human to approve or let run automatically depending on the autonomy level chosen. Orbit, the software underneath it, is included at no extra cost, so the search-term tracking this page just walked through is already built in rather than something to buy separately. None of that is required to apply what's above — a seller doing this manually in Seller Central can use every method on this page without it. It's just where the work goes if the manual version becomes the bottleneck.