Facebook AN Traffic Cheat Warning

In October or November last year, I found that the MAI (installed) running FB easily went to the AN position, and these positions have retained no follow-up behavior, or retained no.

​​In recent days, I have found that some series have abnormal retention, running the retention of AEO (yes, I put the retention of a R1 point directly to the AEO) goal, and found that the retention directly ran out of 70%.Damn!How can I deserve 70% retention for this product!Hurry to check, ha ha, do quite true, have retained, have registered (also thief good), have front behavior, but to the key realisation-related behavior and depth of the event is basically gone.

The data just started to get weird in the last week, before it was in the normal range.

Sure enough… a check is AN.Do not have to go into what is the reason, properly is AN cheating flow.

It’s not really surprising to see AN cheating, 0 retention or 0 follow up on previous MAI runs, after all, no goals were set.

This run AEO appeared cheating, reflecting on the event, running retention, obviously too “hasty”, after all, in case you encounter a brush amount of publisher (which did not meet it), retention brush good, which in the view of the FB this is a proper standard good traffic, hurry up and add!

The point is that I’m guessing it’s not just these AEO series that are going to be on their knees, he looks like he’s going to take my entire all ads inside the ditch.

Since the goal is R1, the R1 that runs out of AN is particularly good, so when I learn about similar people, I will go to some publishers of AN to expose it, including my other MAI series and other AEO series. I guess all my old campaigns are going to have some impact.

Solutions:

Turn off all the R1 AEO series, MAI series strict screening, see the AN proportion increased in time to kill.

Key: now immediately replace the other AEO target, all cut to the new target placement, the direction of the entire advertising study to break back, not lack of volume immediately hand the old series of possible problems directly killed.

Considering that the retention of this kind of too easy to be brushed out by the machine, it is recommended that we do not play this kind of operation in the future. (Google run R1 now have not found the bug, but 99% can’t run away, sooner or later)

Leave a Reply

Your email address will not be published. Required fields are marked *