Meta Catalog Ads Strategy for Shopify: 27x ROAS, Unchanged Since 2016

This is my Meta ads account over five years:

Year	Return on ad spend (ROAS)
2016	6.6
2017	8.3
2018	11.6
2019	20.5
2020	27.08

It’s the same strategy the whole way. I started it in 2016 and never changed it. Not new creatives, not a new agency: catalog ads.

In this guide: why catalog ads keep winning, exactly how I set them up, the bid trick I use, the mistakes that kill them, and two things that push the return even higher.

A note on the screenshots behind these numbers: Meta deletes old data, so nobody can open 2016 in Ads Manager anymore. Luckily I saved them. Later I had to move to a new ad account, so 2023 onwards is a different account: same brand, same strategy.

I tried everything else

I didn’t sit on one idea. Over these years I tried video ads, carousels, engagement posts, creators, new angles and new audiences. Some had great months.

But look at the ad names in any of those yearly lists: copy, copy 2, copy 3, copy 4. Creative ads come and go. The one that kept coming back year after year was “All Products – product catalog sales”: number one by ROAS in 2016 and 2017, top three in 2019 and 2020. And it has no designed creative at all. It’s just my products.

What catalog ads are, and why they win

Catalog ads, also called dynamic product ads (DPA) and now Advantage+ catalog ads, take your product catalog and show each person the products they’re most likely to buy. Someone looked at a product on your site? They see that exact product. Someone who looks like your buyers? They see what people like them buy.

Why do they win? Four reasons.

1. Relevance. Every person sees a product they care about, not a brand message. Relevant ads get more clicks from the same impressions; in my account that meant cheaper clicks and cheaper impressions than my creative ads.
2. The ad is the product. Someone sees a product, clicks, and lands on that exact product. The ad and the page match every single time.
3. Meta does the testing for you. It tries thousands of product × person combinations every day. No creative team can test that many ads.
4. Social proof that keeps growing. This is the one most people never think about.

Years of social proof, inside the ad

My catalog ads have been running for years: the same ads. Over that time they collected thousands of comments, likes and shares. Real people writing things like “I bought it as a gift, she loves it”, “Second order, great quality”, “When will this be back in stock?”

Think about what that means for someone seeing the ad for the first time. They don’t see an ad from a brand they’ve never heard of. They see a product with hundreds of people underneath saying it’s good. It works like reviews, right inside the ad.

That’s very hard to get with normal ads. You make a new creative, it starts with zero comments, and after a few weeks you turn it off and make another one. All that proof restarts from zero every time. A catalog ad that runs for years keeps all of it. So don’t restart it. Keep the same ads running and let the proof grow.

It survived COVID, war and iOS privacy

Then came the hard years: COVID, then a war close to home, and the privacy updates when iPhone users could turn off tracking and Meta lost a lot of its data. Growth got harder for everyone, including us.

The new account, from September 2023 to October 2026: 25 campaigns, 1,073 purchases, a return on ad spend of 16.63. Same catalog strategy. And I touch these ads about once a month, on a seven-figure brand.

How my catalog campaigns are set up

I run the catalog at three levels:

Level	Who sees it	Result (new account)
Broad	New people, no interests, the whole catalog; Meta finds who wants what	283 purchases, ROAS 15.60
Social remarketing	People who liked, commented on or watched my Facebook and Instagram content but never visited the site	249 purchases, ROAS 36.30
Website remarketing	Everyone who visited the site in the last 180 days; they see the exact products they viewed, and similar ones	—

Remarketing will always show a higher return, because these people already know you. That’s normal. The broad campaign is the one that grows the brand, because it finds new buyers every day.

Split the catalog into product sets

Inside these campaigns I split the catalog into product sets: bestsellers, new arrivals, premium high-price products, and cheaper products. Why? Because a $500 product and a $30 product shouldn’t compete for the same budget with the same bid. The cheap one will always win the auction and your best products never get shown. With product sets, each group gets its own budget and its own bid.

An honest note on these numbers: they’re Meta’s own reporting, with 7-day click and 1-day view attribution. No Meta number is perfect. But I’ve used the same settings for years, so the comparison between ads is fair, and the money in the bank tells the same story.

Who this is for, and who it isn’t

There are a lot of experts with a lot of opinions, so let me be honest: this might not work for your brand. It works for mine because I have a big inventory, hundreds of products. Catalog ads need choice: the more products you have, the more matches Meta can make.

If you sell three products, this won’t scale you. Then you need what everyone tells you: lots of creatives, lots of angles, hundreds of tests. I’ve been on both sides.

But if you do have a big catalog, I honestly think many brands could spend less on agencies and big marketing teams, put that money into catalog ads, and make more.

Your catalog is your creative

If catalog ads are the strategy, your catalog is the creative. Most stores connect Shopify to Meta once and never look at it again. Here’s what I check:

• Titles. In the ad, the title is the headline. “SKU 0042” sells nothing; “Black leather crossbody bag” does. Nobody searches on Meta, so this isn’t about keywords there: a clear, accurate title simply tells people what they’re looking at, so more of them click. The same titles help your SEO on Google.
• The first image. That’s the image in the ad. Use a clear, bright photo where the product fills the frame; on a phone, small details disappear.
• Stock and price. When the catalog syncs with your store, sold-out products drop out of the ads automatically and prices update. But only if the sync works. If it breaks, people click a product they can’t buy or see a different price on the page, and you lose their trust in a second. Make sure it syncs automatically, every day or even every hour.
• More products, more photos. Every new product you add is a new ad, for free.

How to get more ROAS from catalog ads

1. The bid ladder

Most people run one catalog campaign and let Meta decide everything. I duplicate it with different bid caps. A bid cap is the most you let Meta bid in the auction. For example: one at $50, one at $60, one at $65 and one at $80. Same catalog, same audience.

After a couple of weeks, see which bid brings the best return. A low bid cap wins fewer auctions, so you get fewer, cheaper sales. A higher one wins more auctions and more volume, but each sale costs more. Somewhere in between is the sweet spot for your store, and you only find it by testing. Keep the winner, pause the rest, and repeat every few months, because the sweet spot moves.

2. Design your catalog

A plain catalog ad is a product photo on white. Today there are tools that let you design it: add your price, a discount, free shipping, your reviews, your logo and a frame in your brand colours. Every product gets it automatically, even new ones. In my experience that adds about 20% to ROAS.

3. Fix the product page they land on

Nobody cancelled CRO. Catalog ads send people straight to the product page, so the product page decides whether the click becomes a sale.

The math: you spend $10,000 on catalog ads at a return of 16, so $160,000 in sales. You don’t touch the ads; you improve the product page and conversion rises 20%. Same clicks, 20% more orders. ROAS goes from 16 to 19.2: $32,000 more from the same $10,000.

The changes that do it are simple. In one client test, a video instead of a photo on the product page brought 16% more revenue per visitor. In another, one five-star review under Add to Cart brought $67,365.

The mistakes that kill catalog ads

I’ve seen these in a lot of accounts, and I made some of them myself.

1. Too many campaigns. A small budget split ten ways means none gets enough sales for Meta to learn. Fewer campaigns, more budget each.
2. Editing every day. Every big change sends the campaign back into learning. That’s why I touch mine once a month.
3. Turning it off too fast. A catalog campaign needs time to find who wants what. Give it at least two weeks.
4. Judging only by remarketing. Remarketing looks amazing, so people move all the money there. But remarketing only works if new people keep coming to your site. Cut the broad campaign and within a few months your remarketing audience is empty.

What I check once a month

• ROAS per campaign and cost per purchase per bid cap. If one bid clearly wins, I move budget there.
• Frequency. If the same people see the ads too often, I need fresh people, so the broad campaign gets more budget.
• Top-selling products from the ads. I make sure they have the best photos and titles, and that nothing sold out is still being shown.
• The comments. They tell you what people love, what they ask and what stops them buying. That’s free research for your product page.

How to start catalog ads this week

1. Open Commerce Manager and check your catalog. Is every product there? Are prices right? Is anything sold out still showing?
2. Fix the basics: clear titles, a bright first image, stock syncing every day.
3. Launch two campaigns, not ten: one broad catalog campaign for new people with your whole catalog, and one remarketing campaign for site visitors who didn’t buy.
4. Wait two weeks with no big changes. Let Meta learn.
5. Start the bid ladder: duplicate the broad campaign with a few bid caps and keep the winner.
6. Design your catalog, then start testing your product page.

And leave the ads running. Every month they run, they collect more comments, more data and more proof.

Recap

Catalog ads, if you have a big inventory. A broad campaign to find new buyers, remarketing to your social audience and site visitors, and product sets so your best products get shown. Treat your catalog like your creative. Duplicate with different bids and keep the winner. Design your catalog. Don’t touch the ads every day, and let the social proof grow. And fix the product page they land on, because that’s where the sale happens.