Over the last few weeks I've been splitting one of my client's pmax campaign into a top sellers and mid/low campaigns using Austin's strategy and advice. The client only had one pmax campaign in the account that was performing very well, and we decided to follow Austin's pmax campaign structure method to try to scale it.
Of the total available daily budget of 9€, the top sellers campaign has 7€ and the mid/low has 2€. The top sellers has had a tROAS of 3000% even before creating the mid/low campaign. Initially, the mid/low campaign ran without any tROAS, and after Austin's insights, I changed it's bidding to tROAS of 3200% on May 7th to prevent Google splitting traffic.
I started splitting products on April 17th, and some time after that the campaign began to struggle. As I have more 3800+ SKUs, at first I've been moving around 30/40 products every day that haven't had any conversions in the last 365 days, until about a week ago when I noticed that the campaign was struggling and I stopped moving products to see if it would recover.
The attached graph shows overview data of both campaigns, from the beginning of April until now. The number of clicks and impressions as dropped considerably, and the spend isn't anywhere near half the available budget (in the last few days both combined haven't even spend 2€). The actual ROAS of 2202% in that period of time did not reach the rROAS of 3000%.
Could it be that the tROAS is so high that Google is being extremely careful with bidding? Or moving 30/40 products a day, even if they don't have any conversion data, is too much for the campaign to handle? Or another reason?