I am debating whether to migrate my traditional Shopping campaigns over to Performance Max (PMax). I've heard mixed reviews about the "black box" nature of PMax and the lack of keyword-level data. For a mid-sized e-commerce store, does the AI-driven automation actually provide a better Return on Ad Spend (ROAS) than manual control, or will I just lose transparency and waste budget?
3 answers
Transitioning to Performance Max can be a double-edged sword for e-commerce retailers. While the AI is excellent at finding customers across YouTube, Gmail, and Display, it often prioritizes "low-hanging fruit" like brand searches unless you set up brand exclusions. I recommend running a split test. Keep your best-performing Standard Shopping campaigns but introduce PMax with a specific "Asset Group" for your top-selling products. Ensure you provide high-quality video and image assets, as the machine learning needs these to beat the performance of simple product feeds used in traditional shopping.
Are you planning to use Audience Signals in your PMax setup to guide the algorithm, or are you going to let it find customers from scratch?
I switched to PMax last year and saw a 20% increase in ROAS, but the lack of search term data is frustrating for long-term SEO planning.
I agree with Jennifer. The results are great, but I had to use a third-party script to actually see where my money was going in terms of search queries!
Christopher, I’m definitely planning to upload my first-party customer lists as signals. I’ve heard that providing a "starting point" for the AI helps it find similar high-value users much faster than just letting it run wild on its own. Does the algorithm prioritize these signals heavily, or are they just suggestions? I want to make sure I’m not just wasting time segmenting my email lists if the AI is just going to ignore them.