Meta Says Its Automated Shopping Campaigns Lift ROAS 22%. Independent Data Says It’s More Complicated.

By Thisura, founder of Run1Ads · 2026-05-29 · Updated 2026-10-28

Quick answer

Meta's official figure is a 22% ROAS lift when advertisers turn on its automated shopping campaign features, but independent tests from Haus and Wicked Reports show the in-platform number often overstates incremental results. Meta's automated shopping campaigns tend to win when you feed them high creative volume and clean conversion signal, and lose when creative is thin and they mostly retarget existing demand. Judge them on incrementality, not on reported ROAS alone.

Meta's official data on its automated shopping campaigns is clean and bold: a 22% ROAS lift when advertisers turn on the automated features, a 6% improvement in retrieval recall, and an 8% ad quality lift in certain segments. That number shows up throughout Meta's 2026 sales material.

The practitioner data is messier. The honest read of the past 12 months: automated shopping campaigns win for some accounts and lose for others, and the difference is not random.

What do independent studies show about Meta's automated shopping campaigns?

The most-cited independent research is Haus's 640-test study. It found that 58% of brands saw higher incremental ROAS on manually structured campaigns than on automated shopping campaigns, even though Meta's in-platform reporting showed the automated campaigns winning.

The gap was attributed to automated campaigns over-attributing conversions that would have happened anyway.

Wicked Reports analyzed 55,661 campaigns and found that new-customer CAC roughly doubled over a six-month window when automated shopping campaigns were the dominant structure ($257 to $528 in their sample). The data is contested, but similar patterns show up often enough that it can't be dismissed.

The counterweight: Top Growth's BFCM data showed automated shopping campaigns at 3.14x ROAS against 2.70x for manual campaigns, a decisive win during the high-intent holiday window. Andromeda strengthens this in 2026 because creative-based delivery is exactly what these campaigns were designed for.

Which is right: Meta's number or the independent data?

Both. Automated shopping campaigns win when you have high creative volume, clean conversion signal and a brand strong enough that incrementality is naturally high. In that case the system optimizes well and Meta's reporting roughly matches incrementality.

They lose when creative volume is low, signal is dirty, or the brand is small enough that the campaign is mostly retargeting existing demand. Then in-platform ROAS looks great but incremental revenue stays flat.

Many experienced operators now run a hybrid split between automated and manual campaigns, growing the automated share as creative production capacity grows. This is not a Meta recommendation; it is what teams report works for them.

Why do automated shopping campaigns disappoint so many accounts?

If you run automated shopping campaigns and ship 5 ad variations a month, you are in the failure mode. The system needs creative volume to work: roughly 15 to 20 active ads with real variation, not minor tweaks.

Without that volume, the campaign tends to spend on retargeting and report inflated ROAS.

If you have the Conversions API deduplicated against the Pixel, paired with 15 or more creative variations per ad set, you are far more likely to see the kind of lift Meta describes.

Most operators are in the first group. They turn on the automated campaign, the reported numbers look good, and the actual business doesn't grow. Months later they find incrementality was the problem.

The cause is operational, not strategic. The system works if you feed it volume and signal, and very few teams can produce both consistently.

How do I check whether my automated shopping campaigns are really working?

1. Pull your past 90 days of results.
2. Compare automated campaign ROAS to your true incrementality if you have any holdout data. You will probably see a gap.
3. Audit creative volume. If you shipped fewer than 30 unique ads in those 90 days, the test was set up to fail.
4. Confirm the Conversions API is deduplicated against the Pixel before trusting any reported number.

Key facts

• Meta reports a 22% ROAS lift from turning on its automated shopping campaign features.
• Haus's 640-test study found 58% of brands saw higher incremental ROAS on manually structured campaigns.
• Wicked Reports' analysis of 55,661 campaigns found new-customer CAC roughly doubled ($257 to $528) when automated campaigns dominated.
• Top Growth's BFCM data showed automated shopping campaigns at 3.14x ROAS versus 2.70x for manual campaigns.
• Results depend heavily on creative volume and clean conversion signal.

FAQ

Do Meta's automated shopping campaigns really lift ROAS 22%?

That is Meta's own figure. Independent work such as Haus's 640-test study shows in-platform ROAS often overstates incremental results, so the real lift depends on your creative volume and signal quality.

Why does in-platform ROAS look better than real results?

Automated campaigns can claim credit for conversions that would have happened anyway, especially when they lean on retargeting. Holdout or incrementality tests reveal the gap.

How many ads do automated shopping campaigns need?

Practitioners commonly aim for around 15 to 20 active ads with real variation in angle and format. Minor tweaks of one ad don't count as variety.

Is there a tool that checks this automatically?

Yes. Run1Ads is a live Meta ads consultant for e-commerce brands that checks your ads, Pixel and store every hour, flags ad fatigue and tracking breaks, and gives the one fix that matters today ranked by buyers lost. It advises only and never changes your campaigns unless you ask for a paused draft.

Related reading

• Meta Ads Attribution Is Breaking in the iOS 26 Era. Here's the Fix
• Meta's Andromeda Rewards Creative Volume: How Many Ads in 2026

Run1Ads is a live Meta ads consultant for e-commerce brands. It checks your ads, Pixel and store every hour and tells you the one fix that matters today. See the live demo or start free, no card needed.