Broad, Phrase, Exact: How Amazon's Three Keyword Match Types Actually Control Your Spend

If you've ever looked at a search-term report and wondered why your ad showed up for a query that had nothing to do with your product, the answer is almost always match type. Amazon gives you three ways to tell it how loosely or tightly a shopper's search has to match your keyword: broad, phrase, and exact. Get the distinction right and you control exactly which searches you pay for. Get it wrong and you fund a stream of irrelevant clicks. This is foundational, and a surprising number of sellers run on autopilot without really understanding the lever in their hand.

The three match types, plainly

Each match type trades reach for control. The wider you go, the more searches you trigger and the less you control relevance.

• Broad — the widest. Your ad can show for searches that include your keyword in any order, plus synonyms, related terms, and variations. Maximum reach, minimum control over relevance.
• Phrase — the middle. Your ad shows when the search contains your keyword phrase in order, with words allowed before or after but the core phrase intact. A balance of reach and relevance.
• Exact — the tightest. Your ad shows only for that exact term and very close variants like plurals. Minimum reach, maximum control — and usually your most efficient spend.

How they work together, not in isolation

The mistake is treating these as a pick-one choice. The match types are a system. Broad and phrase are your discovery engine — they surface the real searches shoppers use, including the long-tail phrases you'd never have guessed, which you find in your search-term report. Exact is your conversion engine — once a discovery campaign reveals a term that converts, you graduate it to its own exact-match campaign where you can bid precisely and protect the budget from drift. Discovery finds; exact harvests. Running only exact means you never discover; running only broad means you never control.

Negatives are the other half of control

Broad and phrase match earn their keep only if you actively prune what they catch. They will, by design, match to searches that aren't right for you — and those clicks cost money. The fix is negative keywords: as your search-term report reveals queries that spend without converting, add them as negatives so the same waste doesn't repeat. There's also a structural use — adding your exact-match winners as negatives in your broad campaigns so the two don't compete for the same search and you can read each clearly. Negatives aren't an afterthought; they're how you keep the wide match types honest.

A simple structure that keeps spend clean

You don't need anything elaborate. A workable setup: an automatic campaign and a broad/phrase campaign doing discovery on a product, feeding winners into a dedicated exact-match campaign with tighter bids. Keep the discovery and harvest campaigns separate so the data stays clean and you can budget each on its own performance. This separation is what lets you bid aggressively on proven terms while keeping exploratory spend on a short leash — and it makes your weekly optimization a quick, obvious routine instead of a guessing game.

Match type can't fix a thin margin

Tightening your match types reduces waste, but it can't make an unprofitable product profitable. Even a perfectly efficient exact-match campaign loses money if your bids exceed what the product's margin can carry. Know your real per-unit profit — after referral fees (around 15%), fulfillment, and landed cost — and set your bids beneath it. Match types control *which* searches you pay for; your margin controls *how much* you can afford to pay. You need both right.

Frequently asked questions

Which match type should I start a new product with?

Start with an automatic campaign plus a broad or phrase campaign to discover the terms shoppers actually use, then graduate the converters into exact match. Leading with exact alone means you only ever bid on terms you already guessed — you miss the long-tail phrases that discovery uncovers.

Why is my ad showing for unrelated searches?

That's broad match doing what it does — matching synonyms and related terms. The fix is to review your search-term report and add the irrelevant queries as negative keywords. If it's persistent, consider moving the keyword to phrase or exact for tighter control.

Do I still need broad match once I have my exact-match winners?

Yes, for continued discovery — shopper language changes and new long-tail terms keep appearing. Just add your exact winners as negatives in the broad campaign so the two don't overlap, and keep pruning what broad surfaces.