Less is more - for best results, do not change Performance Groups more than necessary.
Why (in short)?
Frequent changes mean that our algorithm has to keep adapting. And it takes time for the algorithm to fine-tune the learnings for each and every change. The fewer changes, the better algorithm works towards reaching your set goal.
The Algorithm needs time to fine-tune every Performance Group adjustment
Our smart AI uses historical data and works towards the specific goal you set. It analyzes:
Campaigns within a Performance Group
their behavior
the market’s behavior
goal type
value
and several other factors.
Applying a lot of changes in a short time can affect the performance of your advertising campaigns.
Why? Every change triggers the algorithm to adjust to new constraints (e.g. new goal type or value, or changing the campaigns in the Performance Group).
If changes are done in a (too) short time, the algorithm needs to adapt too often. It needs time to adapt to new data, a new goal or a new Performance Group.
Every change is followed by a mini learning phase
Remember the learning phase? Let me remind you: when you first started with Adspert we advised you to put all your campaigns into one Performance Group per platform (or marketplace) and let it run for at least 2-3 weeks. The learning phase gave the algorithm enough time to learn all about your campaigns.
Now imagine that after every change to the Performance Group the algorithm needs time to adjust. This is a very soft equivalent of restarting the learning phase all over again. Not like the initial one when you first started with Adspert.
Our algorithm already knows your account, but each change brings a much milder variation of the learning phase, meaning the algorithm will need time to fine-tune the learnings and adjust to them.
When you adjust a goal, restructure a Performance Group, or make any other change, we recommend allowing the algorithm at least 7 days to adapt before making another adjustment.
To help you with this, Adspert displays the date of your last change as a reminder.
Furthermore, when you enter the goal adjustment section, Adspert will show a warning if your last goal change was made less than 7 days ago, advising you to wait before updating it again.
The more stability there is, the better the algorithm can work towards the goal you have set.