-
Cracking Facebook’s Ad Algorithm: How Standard Events Shape Your E-commerce Marketing Success
If all products use the same standard events for ad placement, will the system learn to target the same audience?
For example: If several e-commerce platforms all use "Add to Cart" for adding items to the shopping cart and "Purchase" for completing transactions, will this cause Facebook to learn to target the same group of people?
This question touches on Facebook's user categorization logic, but we likely won't find official documentation to support our theories. We can only speculate.
I estimate the logic might work as follows:
- Facebook likely tags users based on basic information such as age, gender, friend connections, and interests (books, movies, etc.). These tags are used to determine ad exposure during campaign delivery.
- Facebook probably adds tags based on user behavior, including GPS information, photo-taking habits, page likes, post interactions (likes, comments), potentially considering post attributes, page types, or even clustering based on similarities with other users who've engaged with the same content.
- Facebook likely considers users' past ad interactions, such as app downloads, in-game purchases, or e-commerce activity. (They might not know specific products purchased but could potentially identify key information through clustering large volumes of conversions for the same product to create tags.)
While we understand Facebook's clustering based on user behavior and attributes, we should also consider that Facebook (and other ad platforms) have strong attribute tagging capabilities for their standard event callbacks.
E-commerce universally uses purchase events, games use level completion and tutorial completion events, in-app purchases, etc. Some platforms even have specific events for different stages of the online shopping journey.
Standard events help ad platforms quickly identify similar users in the early stages. Users who trigger add to cart and purchase events can be collected through standard events and given a unique user tag, which is likely more precise than targeting based on age, gender, or interests alone.
So, when many of our e-commerce products choose to use "Add to Cart" and "Purchase" events for Facebook callbacks, Facebook likely considers both the product type and standard event conversions for tagging. As more products convert, the platform may tend to expose ads more to audiences who have previously converted on these events.
The underlying logic is still audience tagging based on product behavior, but more directly, it might prioritize users who have triggered the same standard event, as this attribute is stronger than some other attribute tags.
We can speculate that using inappropriate standard events for a product category might have some negative impacts. For example, using gaming tags for an e-commerce product might lead to inaccurate audiences during the cold start phase, but this issue likely resolves itself after the product accumulates data for a while.
Regarding the impact of standard events, some leading e-commerce platforms intentionally avoid using standard events for their campaigns, possibly for two reasons:
- They don't want other companies to benefit from their event tags.
- They want to avoid inaccurate standard events affecting their conversions (they might have more specific or customized events for their unique shopping process).
From a campaign management perspective: For smaller online stores, this might not matter much. For larger e-commerce platforms, it's worth considering reducing the impact of imprecise standard events by using custom events instead.
-
1 Reply
-
Advising against standard events is certainly a bold statement. Is that all from your own speculations? Any source or proof to support it?
-
There is nothing wrong with standard events. Meta, like Google and Amazon, knows what people do and don’t buy and they use that to show them ads based on a dozen other factors. Not using standard events is a waste of time.
Log in to reply.