-
PMax Segmentation by ROAS
I’ve been running PMax for a client for a while now. It’s performing pretty well, but we’re looking at ways to optimize. One idea was to segment the campaign into two campaigns based on ROAS, with the high-ROAS campaign getting the bulk of the budget. The current PMax budget is set at \~$800 per day. This would split it to \~$600/$200.
Is there a way to automatically move products from one campaign to another based on ROAS (e.g., if a product in the low-ROAS campaign starts performing well, move it to the high-performing campaign and vice versa)? The product feed contains approx. 12,000 products.
Is this a waste of time given that PMax is supposed to automatically optimize? The spend for top-performers is already at \~70%.
-
1 Reply
-
Lemme ask ya, what happened to the conversion volume in your branded Search campaign when you started using Pmax?
-
This could really well for the ad accounts as long as there is enough data in both PMax campaigns once you do the split.
-
You can try it but I don’t see it working out too well. I did something similar to test products that would go on promotion which would be moved to campaign with a lower ROAS target. The issue I came across is that the new SKUs would need to relearn when moved to the new PMAX campaign and did not respond as expected. I ended up ditching PMAX and transitioning back to standard shopping where I could set different ROAS targets by ad group.
-
It is good to allocate a separate campaign for your best assets and products, so you have more control and better analysis of data
-
I don’t think there is very good support for P-Max in Google scripts yet. But in theory you might be able to inspect ROAS per product for say the past 30-days and if greater than “x” you pause in one campaign and enable in the other… and vice versa if lower than “x.”
It would be really easy to label products using Merchant Center rules, but unfortunately, Google only reports on clicks in that system.
-
I would be cautious when splitting the campaigns as you could end up hurting performance. When you split out the products into a new campaign, Google goes into learning mode again and you may not get the same results. I would leave the best performing products in your current campaign and segment the lower value products into the new campaign. That way the leanings stay with your better performing products.
There is software out there that can automatically move over products based on performance but it’s pricey.
To make things easier you can also try consolidating your products to only show the lowest price variation. For example if you sell shirts that range from $10 to $50, you should focus on only showing the $10 shirt so that you’re more competitive when displayed on Google
Shopping along other ads. I only say this because you mentioned over 12k products so I assume you have variations of the same product just at different price points. This is a strategy that has worked very well for me especially when the budget is tight and or the space is very competitive.
Log in to reply.