What is the Learning Phase?
When starting the optimization with Adspert, please note the algorithm first needs to go through a learning phase to learn about your campaigns.
The same applies when you are already working with Adspert but you are adding new campaigns into the optimization. Those campaigns also will go through a Learning Phase.
Why?
Think of our AI as your new employee. When you welcome a new team member, you probably give them some time to get their head around how things work, right? As well as your new employee, our AI needs to learn about your campaigns, analyze the data, and figure out what's working well in your account and what's not.
When you let our AI learn all about your account, you get better results in the long run. Our algorithm gains more bidding flexibility. It learns faster and better about all the factors that improve the performance of your campaigns. And when the algorithm works better, it boosts the ad optimization which brings you more sales and more profits. All that happens automatically, saving your time, letting you focus on your core business and help it grow. And that’s what it’s all about, isn’t it?
How to kick-off a learning phase?
Create one Performance Group per platform (or marketplace).
Add all your campaigns to it
Set the Cost per Day goal. - Please also see this article for more explanation
Let it run for 2-4 weeks (the learning phase takes around 2 weeks for Amazon, Google Ads and Microsoft Advertising (former Bing Ads) and 4 weeks for eBay Ads)
The best part? Our AI learns on its own. As long as you follow those steps, you don’t need to do anything else to let it learn.
After the learning phase, you can then divide your campaigns into several Performance Groups.
For next steps please have a look at this article: Learning Phase is Over - Now What?
How long exactly is the Learning phase?
While there isn't a fixed endpoint as AI continually learns, dedicating approximately 2-4 weeks can often suffice to familiarize the AI with your campaigns and market. This period allows for the accumulation of a substantial database, providing the algorithm with a robust foundation to operate from.