Dives into how to optimize your Amazon ads placement to help reduce overlapping between product and keyword targeting.
Before diving at the core of the subject, let’s make sure that we all understand the various ad terms and peculiarities of targeting and placements. First, keep in mind that targeting is not a placement but instead what’s triggered it.
Let’s introduce two types of targeting (Figure 1). The first type is product targeting (on the left-hand side), where sellers specify the Amazon Standard Identification Number (ASIN) they want to be targeted in their campaigns. The second type is keyword targeting (on the right-hand side), where sellers enter keywords to be targeted. Moreover, products and keywords targeting can display on one of two placements: search placements and product placements.
Let’s further discuss what search placements and product placements are. To understand the difference between the two placements, we can visit two landing pages on Amazon, commonly referred to as real estate. Similarly to the real estate market, advertising is about purchasing ads in the most praised areas at the best possible prices. The first type of real estate we want to look at is the search term result page (figure 2), where we usually find search placements. In the example below, we differentiated between the keyword targeting “vitamins” circled in green and the search placements circled in red. Therefore, on the one hand, we have a keyword (green) that has been entered in the search bar to be targeted, and on the other hand, we have search placements (red) that have been targeted through this particular keyword targeting.
On the flip side, the product page (i.e. the second type of real estate on Amazon) shelters the second type of placement: product placements circled in red (figure 3). Product placements differ from search placements as they are reached through sponsored products or sponsored displays.
After defining targetings and placements, we could expect that keyword targeting displays search page placement, and product targeting displays product page placement. However, this is a common misbelief as Amazon mixes placements independently from the targeting types, leading to overlapping. One way to prove it is to pull placements reports for product targeting and look at the types of placements. We can see that most of the impressions, clicks, and sales are coming from the product pages placements, but that a few proportions of those metrics are also attributed to search traffic from search placements.
Let’s illustrate this situation with two examples. Our first example demonstrates how a product displayed on a product page can come from both product and keyword targetings. For example, assume the user connects to amazon.com and types “vitamins'' in the search bar. He lands on the search term result page where products are displayed from the keyword targeting campaign with “vitamins” for the exact keywords. Next, he clicks on one of the products with ASINXXX, which loads the product’s page. On this product page, there are placements below the product, which are product page placements from not only campaigns targeting the product ASINXXX but also campaigns targeting the initial searched keyword “vitamins”.
In contrast, our second example reveals how a product displayed on a search page can come from the two types of campaigns. Imagine that the user is already on a product page of a product with a specific ASINXXX. On this page, there is also a search bar where the user types “vitamins”. He then lands on a search page related to the search result of “vitamins”, where sponsored products can be displayed from either a campaign that is targeting directly the keyword “vitamins” or a campaign that is targeting the initial seed ASINXXX.
Overall, keyword and product targeting overlap in the two examples, as they lead to both search page and product page placements (figure 4).
The fact is this overlap will always happen, but some solutions can help minimise it. One of them would be to use a combination of high and no-bid modifiers. Let’s further explain; we start by setting low base bids for both keyword and product targeting. Next, we link our keyword targeting to a high bid modifier for search page placement and a no-bid modifier for a product page placement. In a similar manner, we link our product targeting to a no-bid modifier for search page placement and a high bid modifier for product page placement (figure 4).
Let’s look at an example, pretend that we create a keyword campaign, so we want our ads to display only on the search pages. The idea is to start by setting a low base bid around 0.2$ and use an incredibly high top of search (first page) bid adjustment of approximately 500% as well as no-bid adjustment for product pages. In other words, we are allowing Amazon to bid up to 1.2$ per bid on search placement whereas staying at 0.2$ for product placement. Therefore, we would have most, if not all, of the traffic on search placements.
In the long run, when sellers maintain and update this strategy based on performances, they can adjust one of two things. If the keyword is poorly performing, they should steer at the keyword level on the base bid, or if the overall campaign is poorly performing, then the campaign steering should be on the bid adjustment. However, this solution is not without risk.
While alleviating the overlap between the targetings, the bid adjustment steering solution can result in significant overspend.
We urge sellers to stay ultra vigilant with adjustment steering and to firmly control each type of campaign where it’s getting displayed. The reason is that if a seller uses an incredibly high top of search or product adjustment of 900% (i.e. the highest), his default bid can be multiplied by ten. Additionally, if the seller applies a dynamic up and down modulation, it allows the default bid to go up or down by 100%. To put it in another way, when combining dynamic bidding and adjustment, a base bid of 2$ can increase up to 40$. Although unlikely, with a substantial second-price bid or with a competitor bidding very high would result in paying somewhere around 38$ for a click (remember that Amazon advertising’s eco-system is a second-priced auction where the highest bid wins and ends up paying the second-highest bid). As a rule of thumb, we recommend avoiding mixing adjustments with dynamic up but keeping only dynamic down and conserving a low base bid as it will get quickly multiplied.
Overall, we demonstrated the existence of an overlap between keyword and product targeting. We explained how to master advertising placement to cope with the overlapping and warned sellers of the potential risk involved. Finally, it is worth noting that this is an advanced strategy which it can be difficult to maintain, that is why we only recommend it to experienced sellers spending at least 2000$ a month in advertising.
We will constantly share insightful articles about Amazon ads with you.
Endlich hat es Amazon seinen Partnern ermöglicht, stündliche Statistiken für Anzeigen auf den globalen Marktplätzen abzurufen — also haben wir getan, was wir tun mussten: Wir haben die Zahlen analysiert und nach einer Möglichkeit gesucht, unser Tool zu verbessern.
Finally, Amazon allowed its partners to retrieve hourly stats for Ads on the global marketplaces, so we did what we had to: we crunched the numbers looking for a way to improve our tool.