Most underperforming Amazon accounts we review do not have a bidding problem. They have a targeting problem that bidding is being asked to fix. When a campaign shows a poor return, the instinct is to lower bids, which reduces spend and the losses with it, but also reduces the volume of the queries that were working. Amazon ad optimization targeting is the layer above that: deciding which queries, products and audiences the campaign is allowed to reach at all, so that bid changes are adjusting a set of things you actually want.
The distinction matters because the two levers have different failure modes. A bid that is too low costs you volume quietly. Targeting that is too broad costs you money loudly and teaches the algorithm the wrong things about your listing, because impressions and clicks against irrelevant queries shape what Amazon thinks your product is for.
Structure Before Tactics
A campaign structure exists so you can control budget and read results, and those two purposes often conflict. Too few campaigns and you cannot see which product group is failing. Too many and each one has too little data to make decisions on. The structure we usually land on groups products by margin and by search behaviour, so that products competing for the same queries do not sit in separate campaigns bidding against each other.
Keep discovery and harvest separate. One campaign exists to find queries, running broad with modest bids and a tight budget cap. Another exists to exploit the queries you already know convert, running exact match with bids that reflect their real value. Mixing the two means the discovery spend eats the budget your proven terms needed.
Match Types Are a Sequence, Not a Preference
Broad, phrase and exact are stages in a pipeline. Broad and automatic campaigns generate query data. You read the search term report, promote the terms that convert into exact match campaigns where you can price them properly, and add them as negatives in the discovery campaign so you stop paying twice for the same query. Without that last step, the discovery campaign keeps competing with your own harvest campaign.
Negatives deserve more time than they usually get. In the accounts we manage, the negative list built over the first three months is often worth more than any bid adjustment made in the same period, because it permanently removes a category of wasted spend rather than reducing it by a percentage.
- Run discovery and harvest as separate campaigns with separate budgets
- Promote converting search terms to exact match, then negate them in discovery
- Build negatives from the search term report weekly in the first quarter
- Group products by margin so target ratios mean something within a campaign
- Check product targeting reports, not just keyword reports, for placement waste
Product and Category Targeting Is Underused
Keyword targeting gets the attention, but product targeting is where a lot of the available advantage sits, particularly for products that are an obvious upgrade or complement to something else. Targeting the detail pages of competitors with weaker reviews, higher prices or slower delivery puts your listing in front of a buyer at the exact moment they are evaluating. Targeting complementary products works for accessories and consumables.
The discipline is the same as with keywords. Review the report, keep what converts, negate what does not, and do not leave broad category targeting running indefinitely because it once produced a good week.
How Amazon Ad Optimization Targeting Interacts With the Listing
Targeting decisions are limited by what the listing can convert. A query can be perfectly relevant and still fail if the listing does not answer the buyer’s question, and no amount of bid work fixes that. When we see a term with strong click through and weak conversion, the first place we look is the listing rather than the campaign. Often the image or the title is promising something slightly different from what the query wanted.
This works in the other direction too. Improving a listing raises the value of every query it serves, which changes what you can afford to bid, which changes which terms are viable. Treating advertising and listing work as separate workstreams means neither one gets the benefit.
A Review Rhythm That Holds Up
Weekly: search term report, negatives, obvious spend leaks. Monthly: structure review, promotion of terms to exact, product targeting audit. Quarterly: a harder look at whether the campaign groupings still match how the catalogue sells, because product ranges change faster than campaign structures do.
Resist the urge to change everything at once. If you adjust structure, bids and targeting in the same week you will have no way to attribute the result, and the next amazon ad optimization targeting decision will be made on a guess. One change at a time is slower on paper and faster in practice.
Keep reading: Amazon Ad Optimization · Amazon