So how do you target Facebook ads?
Posted: Thu Jan 30, 2025 4:41 am
In the meantime, a very simple exercise will suffice. We combine these two sets into one, and then check what the sizes of both sets are independently of each other. Then we will quickly see that their sum is definitely smaller than it should be, considering how large both groups are independently of each other. Most of these groups are exactly the same people.
Another example:
In the consumer world, there are highly segmented groups related to running:
5 km run,
10 km run.
It turns out, however, that these groups overlap by about 90%. This means that lebanon rcs data there is no point in looking for such very precise targeting, because it is simply not possible. On the other hand, even if it were possible to target something very strongly (e.g. to the level of several thousand recipients), it still would not be good.
Why?
First, a potential reach of several thousand people is not a group that we will definitely reach. These are users who, for example, log into Facebook, rarely, once a month. Our group is therefore several times smaller in reality. It will be difficult for us to regularly and frequently enough display ourselves to this type of recipient. So that this ad actually works on them, or works in a time that is satisfactory for us, e.g. several days or weeks.
In addition, the smaller the target group we create, the more expensive it will be. This is because the algorithm will have a harder time finding recipients who will be tailored to our high requirements.
Our goal is to set a certain framework for the algorithm. Based on what we ourselves target, and then based on user activity under ads, the algorithm, through the learning phase and subsequent optimization, continuously improves our targeting. So we don't have to think about it in such detail anymore. So it's a marriage of our intelligence with artificial intelligence.
Another example:
In the consumer world, there are highly segmented groups related to running:
5 km run,
10 km run.
It turns out, however, that these groups overlap by about 90%. This means that lebanon rcs data there is no point in looking for such very precise targeting, because it is simply not possible. On the other hand, even if it were possible to target something very strongly (e.g. to the level of several thousand recipients), it still would not be good.
Why?
First, a potential reach of several thousand people is not a group that we will definitely reach. These are users who, for example, log into Facebook, rarely, once a month. Our group is therefore several times smaller in reality. It will be difficult for us to regularly and frequently enough display ourselves to this type of recipient. So that this ad actually works on them, or works in a time that is satisfactory for us, e.g. several days or weeks.
In addition, the smaller the target group we create, the more expensive it will be. This is because the algorithm will have a harder time finding recipients who will be tailored to our high requirements.
Our goal is to set a certain framework for the algorithm. Based on what we ourselves target, and then based on user activity under ads, the algorithm, through the learning phase and subsequent optimization, continuously improves our targeting. So we don't have to think about it in such detail anymore. So it's a marriage of our intelligence with artificial intelligence.