Amazon automatic targeting matches Sponsored Products with relevant search and product contexts using Amazon's systems. Manual targeting lets the advertiser select keywords or products. Use automatic campaigns to learn about demand and manual campaigns to test chosen targets, with distinct budgets and review criteria.
Amazon identifies four automatic targeting groups: close match, loose match, substitutes, and complements. The operational question is how much uncertainty each campaign should carry, and what evidence would justify another step.
Automatic vs manual targeting: define the campaign goal
A mature best seller, a new variation, and a product with limited stock have different constraints. Decide whether the current job is discovery, protecting relevant demand, evaluating a product context, or supporting a launch. Assign a budget and review question to that job.
For an illustrative compact appliance, discovery might ask whether shoppers search by room size or a technical feature. A controlled manual test could focus on one verified intent. Neither campaign needs to cover every possible use case or duplicate the entire catalog.
How to review Amazon automatic targeting groups
Automatic groups represent different matching contexts. A combined result can conceal a useful group alongside one that does not fit the product. Inspect the underlying traffic before changing the whole campaign, and check which ASIN received it.
Translate observations into hypotheses. A promising ASIN context may justify a product test; a useful query may justify a keyword test. A high click count alone does not establish relevance. The search-term review workflow helps organize the evidence into actions.
Move automatic targeting evidence into a manual test
Manual targeting is useful when you can name the opportunity and define a review. Keep the starting target set small enough to inspect. If automatic delivery suggests a competitor product, look at that offer before assigning a manual product bid.
Consider price, inventory, attribution timing, and whether the conversion came from the intended product. A single order is not a permanent winner. The ASIN-targeting guide covers the comparison a shopper is likely to make and the evidence needed to keep testing.
- Give each test a product and a purpose.
- Record the intended spending boundary.
- Separate relevance evidence from conversion evidence.
Give automatic and manual campaigns distinct roles
Moving a query into manual targeting does not automatically require blocking it in discovery. The decision depends on budget ownership and the need for further learning. Use negatives for an explicit relevance or traffic-routing reason, with the scope recorded.
Review both campaigns together after changes. A cleaner manual campaign with falling total relevant orders may represent a transfer rather than growth. The branded-versus-non-branded guide adds another useful view of the demand being captured.
Let automatic campaigns earn their learning budget, and manual campaigns earn their control. Promote evidence into a defined test rather than building campaigns by habit.
Source material
Amazon Ads: Sponsored Products targeting
Amazon Ads: keyword match types and automatic targeting groups
