“What on earth are ads for my pet food products doing on beauty websites?” Questions like this or similar ones may arise when using our app. This is because you will indeed find ads for your shop on unrelated pages. Whether it's furniture on gaming sites or sausage products on sports sites, at first glance it probably seems pretty random and sometimes out of place – so it's certainly not unreasonable to question whether this is appropriate. Wouldn't it make more sense to block certain sites and only display your ads on sites that are thematically relevant to your products? Well, if that were the case, there would be no article here, because we would have already taught our algorithm to do just that. So there are good reasons why it doesn't...
The idea of using website targeting to specifically target certain topic-relevant websites with your advertising or to exclude irrelevant sites using blacklists is not a bad one in principle. This allows campaigns to be focused on target groups and thus made significantly more efficient. Retargeting campaigns work differently: here, visitor behavior in the shop is analyzed and the user is then targeted again on external websites. The focus here is on users who have already shown interest in one of your products, so you can be sure that they belong to your target group. And they happen to be browsing the pages they happen to be browsing – and these may of course include pages where you would not necessarily expect to find your main target group. People are individuals, surprising and not so easy to pigeonhole. This is a trait that pushes website targeting to its limits – but not retargeting!
The algorithm optimizes primarily on a user-based level, and placement is just one of many relevant criteria. First, behavior in the shop is analyzed: the number of retargeting touchpoints at various levels (category, product, shop search, shopping cart, and order completion page) as well as the length of stay. When it comes to ad delivery, factors such as recency (the time between the shop visit and the ad contact), location, browser, operating system, device, ad, placement, and much more are also taken into account, including website placement (referrer) and the context of the content.
While the campaign is running, the algorithm works its magic and shows what it can do: Based on the factors mentioned above, it optimally adapts to each user, pushes the delivery with the highest purchase probability, and throttles those with the lowest. In short, it ensures that the user is always contacted with the right web shop ad at the right time and in the right place.
Conclusion
Even if it may feel at first as if your own advertising is sometimes being delivered in a way that is “not appropriate,” you can, should, and must trust the algorithm! It is already learning from a four-digit number of performance campaigns, automatically optimizing ad delivery, and is thus able to continuously improve performance—without excluding any websites, locations, or devices.
.png)