Product feed optimization?

Ask Me Anything Q&A Google Shopping AMA

Topic summary

A seller asks how to optimize a Google Shopping product feed under a tight budget with many variants. Inventory is ~150 SKUs (stock-keeping units) but expands to 3,000+ due to sizes and bundle options, and a <$50/day budget means most products get little or no spend.

Key question: should they reduce variants (e.g., list one size) to avoid diluting budget and concentrate spend on fewer items?

They seek which feed elements most “move the needle,” specifically:

  • Product titles: “design-led” (brand/style descriptors) vs “intent-led” (search-intent/keyword-focused)
  • Google product categories: how much they matter and how to use them
  • Images: mockups vs lifestyle photos for performance
  • Any other must-do optimizations or prioritization tips

No responses or decisions yet; guidance requested remains open.

Summarized with AI on January 29. AI used: gpt-5.
1

Feed with many variants vs less variants, say i have 150skus but with sizes and bundle options that increases to 24 per sku so suddenly i have over 3000 skus. Campaign starting budget less than $50 per/day most products don’t get spend. so less variants say in 1 size makes more sense right else im spreading myself too?

Most important, what will move the needle most such as titles(design led vs intent led terms), google Categories etc.. Images say mockups vs Lifestyle anything else we should know?

Thanks, mike

2

Hello Mike, nice to see you on the community, I don’t need to introduce myself, we already know each other :slightly_smiling_face:

Yes, you don’t want to spread your budget too finely, as then to increase performance will be a very long journey.

So you need to ask yourself, do people in your case buy just the top, just the shorts or both. Or is it all random.

In tha case, show all options.

Then for sizes, as I see you have 9 sizes, times 3 options, thats already 27 products. That is a lot. However the problem is that consumers do search for specific sizes sometimes. And you don’t want to miss out on these searches.

But I would do a test.

Test for 3 weeks, all 3 options, but just submit 1 size, lets say medium. But don’t mention it is size medium.

Then test another 3 weeks, with all sizes, so all 27 products. And see how the performance is. Make sure you add the size attribute and add it in the title as well as the description.

Make sure to test a product that has traffic, don’t test something that has barely any traffic. Give it it’s own budget.

So lets say $5 or $10 per day. For the total 6 weeks for this specific test.

Currently what moves the needle are titles, product type, product highlights, product details, images, price, sale price, variant attributes, as well as annotations.

https://www.youtube.com/watch?v=Ztru3Ouciu4