On this page
The purchase event is the campaign
The stores that stall optimized for view content or add to cart because purchases were rare, and then scaled a campaign that learned to find browsers.
If purchases are too rare for learning, fix the offer or the site before you open more ad sets. The learning threshold is on Facebook ads cost.
Show the product being used
A packshot on a gradient is a starting static. The ads that hold are a person using the thing, a before and after when the claim is honest, and a demo of the one feature that makes the product. Catalog cards retarget. They rarely introduce.
Creator video for that job is on UGC videos. Before-and-after structure is on before and after ads.
When the catalog earns its place
A catalog campaign needs a clean feed: in-stock products, real prices, images that match the site. Point it at people who already visited or bought. Using it as the only prospecting is how a store shows random SKUs to strangers.
The shared Meta setup, pixel through placements, is on how to advertise on Facebook.
Questions
Should an ecommerce store start with Advantage+ shopping?
Advantage+ sales is a reasonable first campaign when the purchase event is solid and you have several distinct ads. It is a bad first campaign when you have one image and no purchase data.
Do I need a creator for every product?
For the hero product, yes if you can. The rest of the catalog can ride dynamic ads to people who already know you. Filming 80 SKUs before the hero converts is how the budget disappears into production.
Why do my add-to-carts look great and my purchases do not?
The campaign learned to find people who add to cart. Change the optimization to purchase once that event fires reliably, and expect the CPA to be the real one, which is higher and more honest.