- Keyword targeting follows shopper search intent, while product targeting reaches shoppers around specific products or categories.
- Sponsored Products supports both keyword and product targeting; Amazon also uses product/contextual tactics across other eligible ad products.
- Exact, phrase and broad match should have different jobs in discovery and control; they are not simply three versions of the same campaign.
- Targeting decisions should be based on relevance, conversion, CPC, margin and enough data—not only on whether a target has one sale.
1. What keyword targeting does
Keyword targeting allows an advertiser to bid on words/phrases connected to what customers search. It is most useful when the seller wants to reach specific search intent and can identify queries that describe the product or problem it solves.
The advertiser is not buying the keyword permanently. The campaign competes for eligible ad placements based on targeting, bid, budget, relevance and Amazon’s ad auction/serving systems.
2. Exact, phrase and broad match have different roles
Exact match gives the most control around a defined query intent. Phrase match allows broader search-term coverage while preserving the phrase relationship. Broad match is designed for wider discovery and can surface queries the advertiser did not explicitly list.
A useful structure separates discovery from proven intent so search-term performance can be read clearly and strong queries can be managed deliberately.
3. What product targeting does
Amazon Ads describes product targeting as a way to reach shoppers browsing similar or complementary products or categories. In Sponsored Products, advertisers can target specific products/ASINs or categories and may have refinement options depending on the interface.
This is useful for competitor conquesting, complementary-product placement, own-ASIN defence/cross-sell and category discovery.
4. Keyword versus competitor-ASIN intent
A keyword such as “personalized doctor pen” reaches a shopper expressing that search intent. Targeting a competing personalized-pen ASIN reaches a shopper already viewing a specific alternative. The audience context is different even when the products are similar.
Use keyword campaigns to cover search demand and product targeting to influence consideration around product/detail-page and category contexts.
5. Own-ASIN targeting can be useful too
Product targeting is not limited to competitors. A seller can target related products within its own catalogue to support cross-sell or defensive placement where the economics make sense.
Do not automatically target every own ASIN. Choose relationships that make sense to the shopper, such as pack sizes, complementary products or a premium/entry option.
6. Category targeting needs control
Category targeting can expand reach quickly but can also create broad traffic. Use available refinements such as brand, price, rating or Prime eligibility where supported and commercially relevant.
Review product/category targets individually. A category can contain many products that look related at a high level but do not share the same buyer intent.
7. Use search-term data to move from discovery to control
Automatic, broad or phrase campaigns can discover converting customer queries. When a query has enough evidence and is commercially important, it can be added to a controlled exact-targeting campaign or ad group with an appropriate bid.
That does not require instantly negating it everywhere else. The seller should decide whether the goal is isolation, budget control, rank/visibility or simply additional coverage.
8. Negative targeting should be evidence-led
Negatives can stop irrelevant or inefficient traffic, but premature negatives can also remove useful discovery. Evaluate relevance, clicks, spend, conversion and product margin before excluding a term or target.
If a term is clearly irrelevant, exclusion can be immediate. If it is highly relevant but has weak conversion, investigate the offer/listing and bid before assuming the keyword itself is the problem.
9. Build a portfolio, not one giant ad group
Separate major targeting jobs so budgets, bids and results are understandable: discovery keywords, proven exact terms, competitor ASINs, own-ASINs and category testing. This makes optimisation decisions more precise than mixing dozens of unrelated targets into one campaign.
Structure should still fit catalogue size. A small seller does not need hundreds of campaigns simply because each targeting type exists.
Targeting decision matrix
| Targeting type | Use it when | Example |
|---|---|---|
| Keyword exact | You want controlled coverage of proven high-intent queries | personalized doctor pen |
| Keyword phrase/broad | You want search-query discovery with different levels of control | personalized pen / gift pen |
| Competitor product | You want visibility around a specific alternative | Relevant competing ASIN |
| Own product | You want cross-sell or defensive placement | Pack size or complementary own ASIN |
| Category | You want broader product-context discovery | Relevant category with available refinements |
Frequently asked questions
What is the difference between keyword and product targeting?
Keyword targeting is based on shopper search terms; product targeting focuses on specific products/ASINs or categories and the shopping context around them.
Can Sponsored Products target competitor ASINs?
Yes. Sponsored Products supports product targeting, which can include specific products/ASINs and categories subject to current campaign options.
Should every converting search term be moved to Exact?
Not automatically. Promote commercially important proven queries when greater bid/budget control is useful. Consider data volume, overlap and campaign objective.
When should I use category targeting?
Use it for broader product-context discovery when the category is relevant and the available refinements can keep traffic commercially sensible.
Should I negative a relevant term after several clicks with no order?
Use sufficient evidence. Check spend, CPC, relevance, conversion history and product margin. A relevant non-converting term may need a bid/listing fix rather than immediate exclusion.
Was this guide helpful?
Your feedback helps CES decide which guides need clarification or updating.