Table of contents
What is Amazon keyword targeting?
Keyword research finds the words; Amazon keyword targeting deploys them.
Every keyword you target carries four decisions. Match type — how strictly Amazon must match a shopper’s search to your keyword: exact, phrase, or broad. Bid — what each click on that target is worth to you, based on its intent strength and your margins. Placement — which campaign and ad group it lives in, so budgets and performance read cleanly. Negatives — which search terms must never trigger it, so the wrong traffic never reaches your listing.
Take “insulated water bottle”. As an exact-match target in your core campaign with a confident bid, it's a money keyword. As a broad-match target in a discovery campaign at a low bid, it's a scout that finds new phrases. Same keyword, two completely different jobs — the targeting decides which one it does.
If you haven't built your keyword list yet, start with our guide to Amazon keyword research . Targeting works best on a researched list, not a guessed one. The free keyword explorer expands one seed into a tagged keyword universe in your browser.
Key takeaway: Targeting = match type + bid + placement + negatives. The same keyword does different jobs depending on how you target it.
Match types: exact, phrase, and broad in 2026
Amazon gives you three keyword match types — the Amazon PPC keyword match types every seller should know. Think of them as three levels of trust:
Exact match is maximum trust. Your ad is eligible only when the search closely matches your keyword. Amazon still allows close variants like plurals and word order. Use it for keywords you've researched and believe in — proven converters and your strongest launch terms. Exact targets give the cleanest performance read and usually the most efficient ACoS of the three.
Phrase match is medium trust. The search must contain your phrase in order, with other words allowed before or after. “Best insulated water bottle for gym” can match “insulated water bottle”. Use it to stretch proven keywords into longer, still-relevant searches without handing Amazon a blank check.
Broad match is discovery mode. Word order does not matter and Amazon may match related terms and synonyms. Use it at low bids in a dedicated discovery campaign whose job is finding new search terms — never as the home of your money keywords.
The classic beginner mistake: stacking the same keyword in all three match types inside one campaign “to be safe”. That creates duplicate targets competing for the same searches and muddies every report. Keep your Amazon keyword match types in separate campaigns or ad groups — structure is next.
Key takeaway: Exact = proven winners. Phrase = controlled expansion. Broad = low-bid discovery. Never stack the same keyword across match types in one campaign.
Campaign structure that makes targeting work
Match types only work when the structure around them is clean. Here's a targeting structure that scales:
1. Core exact campaign. Your researched winners in exact match, each with its own bid. This gets the majority of your budget — the profit engine.
2. Phrase expansion ad groups hold proven keywords in phrase match, grouped by theme: material, size, use-case. Slightly lower bids than exact; their job is controlled reach.
3. Broad discovery campaign. Unproven, discovery-stage terms in broad match at low bids. A capped daily budget keeps discovery from cannibalizing the core. Its only job is feeding you new search terms.
4. Automatic campaign. Let Amazon find what you missed — also on a low budget, also a pure discovery source.
Keep exact, phrase, and broad targets in separate campaigns (or at minimum separate ad groups). That way, each match type’s ACoS reads honestly. When everything sits in one campaign, you cannot tell whether exact or broad is carrying the weight — and you cannot fix what you cannot see. Our free campaign builder turns a grouped keyword list into this structure in minutes.
Key takeaway: Separate campaigns for exact, phrase, broad, and auto. Discovery gets capped budgets; the core gets the money.
The free targeting workflow: three tools, one clean list
Here is the targeting workflow, demonstrated with a realistic example: you sell an insulated stainless steel water bottle.
Step 1 — expand. Run your seed through the free Amazon keyword explorer : “insulated water bottle” becomes a tagged universe — “stainless steel water bottle”, “water bottle 32oz”, “gym water bottle insulated”, “kids water bottle”, and hundreds more.
Step 2 — group. Sort the universe by intent with the keyword grouper : core product terms, material seekers, size seekers, use-case buyers. Each cluster maps to a campaign from the structure above.
Step 3 — de-duplicate. Before a single target goes live, run the full list through the duplicate keyword cleaner . “Water bottle insulated” and “insulated water bottle” are one target for bidding purposes. Duplicates across campaigns mean you bid against yourself and split performance data across two rows.
Step 4 — mine the stems. After your first week of data, paste your search-term report into the N-gram analyzer . It breaks every search term into word stems and shows which ones convert: if “stainless” and “32oz” keep appearing in converting terms, those stems become new exact targets. Merge their variations with the keyword phrase combiner so each stem is targeted exactly once.
Try it: free N-gram analyzer Paste a search-term report and find the word stems that actually convert — free, in your browser.
Open the N-gram analyzerKey takeaway: Expand → group → de-duplicate → mine stems. Clean the list before it becomes campaigns, and duplicates never reach your ad account.
Bidding and budgets per target
The principle: your bid should reflect the target’s quality.
Exact core targets get your real bids — sized against your margins, not your hopes. If a click cannot plausibly pay for itself at your conversion rate and target ACoS, the bid is a guess. Size it with the free keyword bid calculator before it goes live.
Phrase targets bid slightly below their exact counterparts: same intent family, slightly less certainty.
Broad and auto discovery bid low, always. Their job is cheap information, not sales. Cap discovery budgets so a surprise viral search term cannot drain the account overnight.
Budget split follows the same logic: the core exact campaign gets the lion’s share, phrase gets a meaningful slice, and discovery campaigns get fixed caps. When a discovery term proves itself, it graduates: promoted to exact in the core campaign, where it earns a real bid and real budget. Over time, hand this thinking to automation: the free bid optimizer updates per-target bids from your actual performance data.
Key takeaway: Bid to the target’s certainty: full bids for exact, measured bids for phrase, low bids and capped budgets for discovery.
Negative targeting: the other half
Every targeting strategy has a shadow strategy: negatives. While match types decide what you show for, negatives decide what you refuse to pay for.
Negative exact blocks specific searches that burn money — the “kids water bottle” clicks when you only sell adult sizes, the bargain-hunter modifiers when your price point cannot compete there.
Negative phrase blocks whole stems. One negative phrase on “glass” can stop a hundred irrelevant “glass water bottle” searches before they cost you a click.
Harvest negatives from the same search-term reports you mine for winners. Any term with spend and no sales after a fair test earns a negative. Run your reports through the free negative keyword analyzer to spot the burners step by step instead of eyeballing spreadsheets.
The discipline that matters: add negatives at the right level. Campaign-level negatives protect a whole campaign’s theme; ad-group negatives fine-tune within it. And never negate a stem that also appears in your converting terms. Check the n-grams first with the N-gram analyzer before you block anything broad.
Key takeaway: Negatives are targeting too. Block proven losers at campaign level, fine-tune at ad-group level, and never negate a converting stem.
Maintaining your targets: the weekly loop
Targeting isn't a launch task; it's a loop. Markets shift, competitors bid, language drifts. Run this weekly rhythm:
Monday — pull search terms. Export last week’s search-term data for every campaign.
Wednesday — mine and promote. Run the report through the N-gram analyzer and the search term harvester . Converting terms graduate to exact targets in the core campaign, and new stems become phrase targets.
Friday — prune and protect. Negate the losers, and run your target list through the duplicate keyword cleaner again. Duplicates creep back in as teams and tools add targets over time.
Monthly — restructure. Check whether ad groups have drifted off-theme and whether discovery budgets still match their harvest rate.
Key takeaway: Weekly: pull, promote, prune. Monthly: restructure. Targeting compounds — small weekly wins become a moat.
From targeting to revenue with BidPilot
Every tool in this workflow is free and runs in your browser. Start with the keyword explorer for expansion, the N-gram analyzer for stem mining, the duplicate keyword cleaner for list hygiene — no signup, nothing uploaded. Targeting is the part of PPC worth understanding deeply before you automate any of it.
Once your targeting structure is live, BidPilot Pro takes the weekly loop: adjusting bids toward your target ACoS, harvesting new converting search terms into your campaigns, and negating the waste — every change logged and visible. Flat $49 per month, 30-day free trial.
Key takeaway: Learn targeting with the free tools. Automate the weekly loop once the structure is proven.
Related reading
- Campaign Targeting: The 2026 Amazon Seller Guide — Covers campaign targeting.
- Amazon Keyword Ads: The 2026 Seller's Guide — Covers amazon keyword ads.
- Amazon Keyword Advertising: The 2026 Seller's Guide — Covers amazon keyword advertising.
- Amazon Keyword Optimization: The 2026 Seller's Guide — Covers amazon keyword optimization.
- Product Targeting: The 2026 Amazon Seller's Guide — Covers product targeting.
About the author
BidPilot Editorial Team — Amazon PPC specialists who run campaigns, build free seller tools, and publish hands-on guides from real account experience.
Amazon Keyword Targeting: FAQ
Amazon keyword targeting is how you deploy keywords in PPC: choosing the match type (exact, phrase, broad), the bid, the campaign and ad group placement, and the negatives for every keyword. Keyword research finds the words; targeting decides what each word is allowed to do and what it costs. Good targeting puts proven keywords in exact match with confident bids, exploratory terms in low-bid discovery, and blocks irrelevant searches with negatives.
Use exact match for keywords you have researched and believe in — it gives the cleanest performance read and usually the most efficient ACoS. Use phrase match to extend proven keywords into longer related searches. Use broad match only at low bids in a dedicated discovery campaign to find new search terms. Never put the same keyword in multiple match types inside one campaign.
Run your full target list through BidPilot’s free duplicate keyword cleaner before launch and monthly afterwards. It finds word-order duplicates like “water bottle insulated” versus “insulated water bottle” so each concept is targeted once. Duplicates split your performance data across rows and make you bid against yourself in the auction.
An N-gram is a word stem pulled from your search terms — for example “stainless” from “stainless steel water bottle”. BidPilot’s free N-gram analyzer breaks your search-term report into stems and shows which ones appear in converting searches. Stems that convert become new exact-match targets; stems that only spend become negative phrase targets.
Keep ad groups tightly themed: 5–15 closely related keywords per ad group is a practical range. The real rule is thematic purity — every keyword in an ad group should deserve the same bid and the same ad copy. When a keyword needs a different bid, it needs its own ad group.
Review search terms weekly: promote converting terms to exact match, negate losers, and de-duplicate the list. Restructure campaigns monthly. Targeting compounds — the weekly loop is where profitable accounts separate from the rest.
Start with exact match on your best-researched keywords so every click teaches you something clean, plus a small low-bid broad or automatic discovery campaign running in parallel to find terms you missed. Starting broad-only with big bids pays for Amazon’s guesses instead of your research.
Target the right keywords. Stop paying for the wrong ones.
Clean your list with the free tools, then automate the weekly loop with a 30-day trial.
30 days free · No credit card · Cancel anytime