Amazon's advertising algorithm changes dramatically? Seriously affect conversion rate

In the past two days, many places have reported that the broad matching logic of Amazon’s in-site advertising has changed, which directly affected the conversion rate of advertising campaigns!

This news mentioned that the change in the advertising logic can be summarized as follows: the broad match mode will split the keywords set by the seller and expand them separately. For example, if the seller sets a keyword combination of A+B, before the logic is changed, the broad match will only separate the two words for delivery, that is, A, B and A+B, and will not expand the separated words again. After the revision, the expansion after the word splitting is added, that is, on the basis of the previous one, there will be combination words such as A+X and B+X. As for what kind of word X is, the algorithm will grab it according to A and B respectively.

In fact, Amazon has a relatively detailed explanation in the background, and some sellers have also retained official annotations before and after the advertising logic changes.

The above are the rules before the change, and the below are the rules after the change. I found a Chinese version and put it below for you to take a look.

The first sentence here clearly states that in the broad match mode, the customer's search terms can contain keywords in any order, and may also include singular, plural, variants, synonyms and related words. It seems that Amazon's split-word matching seems to be in line with this logic, but this logical change that everyone is discussing actually deviates from the goal of broad matching.

For example, "cotton shirts" is a relatively standard A+B keyword, but the weight of this word is obviously different. As a modifier, cotton only narrows the scope of the main word shirts. If you expand "cotton" without the main word, and use a bunch of words that have nothing to do with the product, such as "cotton dress" and "cotton sheets", this change can only bring more useless traffic. The result is that the exposure is very large, but the conversion is very small. Over time, Amazon will reduce the push traffic due to poor conversion of advertising activities, and fall into a vicious cycle.

Matching search terms that are completely unrelated to the main term is not the purpose of Amazon’s in-site advertising. The occurrence of this situation only means that there is a certain deviation between the current situation and the purpose. Whether to modify the algorithm or stick with it depends on why Amazon wants to make changes at the algorithm level .