Catalog ROAS vs campaign ROAS: why they never match (and which one to trust)
Catalog ROAS and campaign ROAS don't match in Meta because they measure different slices of the same spend. Catalog (DPA) retargeting recaptures people who already visited your site and typically returns 2.5x to 6x ROAS. Cold DABA prospecting finds buyers who have never heard of you and lands around 1.2x to 2.5x.1 Campaign-level ROAS blends both, plus different audiences, attribution settings and reporting breakdowns, so the numbers diverge by design, not because your reporting is broken. Judging both against one blended target is the actual mistake, not the gap between them.
If catalog product breakdowns look empty or match rate is stuck low while spend still runs, that is a different failure mode: Meta catalog content_id mismatches. Fix ID alignment first, then judge the structural ROAS gap on this page.
Catalog ROAS and campaign ROAS don't match in Meta for three reasons. 1) Audience: DPA retargeting reaches warm visitors (typically 2.5x to 6x) while DABA prospecting reaches cold buyers (typically 1.2x to 2.5x), so any blend sits between them. 2) Measurement: breakdowns by product, ad set or campaign can use different attribution settings, windows and reporting views, so totals rarely reconcile line by line. 3) Visibility: Meta only counts what its own pixel and catalog can see, so it misses halo effects that show up in blended MER. Split reporting by campaign type, set separate ROAS benchmarks, rule out a content_id mismatch, then use MER to decide whether to grow, hold or cut spend.
Diagnose the mismatch: what you see and what it usually means
| What you see in Meta | Most likely cause | What to check | What to do |
|---|---|---|---|
| Retargeting catalog ROAS far above prospecting ROAS | Structural audience gap (warm vs cold) | Split DPA retargeting and DABA prospecting into separate rows | Set separate benchmarks, roughly 2.5x to 6x and 1.2x to 2.5x |
| Campaign ROAS lower than the catalog or product view | Campaign blends cold and warm spend, or the views use different attribution settings | Compare the same attribution window and the same date range in both views | Report one agreed window and compare like with like |
| Product breakdowns empty or match rate stuck low while spend runs | Catalog IDs not matching pixel IDs | Check that pixel content_id values equal catalog item IDs | Fix ID alignment first, see the content_id mismatch guide |
| Strong ROAS but new customers are flat | Retargeting is re-counting people who would have bought anyway | Look at new-customer ROAS, not only blended ROAS | Read new-customer ROAS vs blended ROAS |
| Meta ROAS looks weak but revenue and profit are healthy | Halo effects on organic, email and other channels that Meta cannot see | Track total revenue divided by total marketing spend | Use MER, see MER vs break-even ROAS |
Why catalog ads report such different numbers
Dynamic Product Ads (DPA) and Dynamic Ads for Broad Audiences (DABA) get conflated because they run through the same catalog, but they solve different problems. DPA retargets people who already visited your site, browsed products, or abandoned a cart, and brings them back with the exact items they looked at. DABA prospects people who have never been to your site, using Meta's behavioural signal instead of pixel history to guess who might buy.1
Those are structurally different audiences, so they produce structurally different ROAS. A cart abandoner already knows your brand and was close to buying, which is why DPA retargeting can post 2.5x to 6x. A cold prospect has never heard of you; landing anywhere in the 1.2x to 2.5x range there means the campaign is working.
If you are pausing or fixing a prospecting campaign because it cannot touch your retargeting ROAS, you are solving the wrong problem. Split reporting by campaign type first, DPA retargeting vs DABA prospecting, before you decide anything is underperforming.
The comparison, side by side
| Campaign type | Typical ROAS | Who it reaches | What high ROAS actually proves |
|---|---|---|---|
| DPA retargeting | 2.5x to 6x | People who already viewed your site or cart | You recapture warm intent well, not that you can find new buyers |
| DABA prospecting | 1.2x to 2.5x | People who have never visited your site | You generate real new demand; a lower ratio is expected |
| Blended MER | 3x to 5x+ target | All channels, all revenue | Whether total marketing spend pays for itself, halo effects included |
Why campaign ROAS and catalog ROAS differ even for the same ads
The audience gap is the biggest driver, but it is not the only one. Three measurement effects stack on top of it.
- Blending. Campaign ROAS is total purchase value divided by total spend for everything inside the campaign. If one campaign mixes retargeting and prospecting, the result sits between the two ranges and describes neither audience well.
- Attribution settings. Different views in Ads Manager can use different attribution windows. A shorter click window or a view-through setting will move ROAS up or down without any change in actual sales, so always compare the same window and date range.
- Breakdown mechanics. A product or catalog breakdown splits one purchase across the items shown and bought, so line items rarely add up to the campaign total. If breakdowns are missing or oddly low, check for a content_id mismatch before you trust any product-level ROAS.
None of these mean Meta is wrong. They mean each ROAS number answers a narrower question than the one most teams ask of it. If you are scaling budget on the strength of retargeting ROAS alone, check new-customer ROAS too, because returning buyers can flatter the headline number.
Why a single blended number (MER) matters more than either
Once retargeting and prospecting are separated, the next question is which number to report up. The answer is neither in isolation. It is Marketing Efficiency Ratio (MER), sometimes called blended ROAS: total revenue divided by total marketing spend across every channel, not just Meta.2 For a deeper walkthrough, see MER vs break-even ROAS.
MER matters because platform-reported ROAS only counts what its own pixel or catalog can see. In one illustrative example, a DTC brand running paid search, paid social, and influencers sees 2.0 ROAS in Meta Ads Manager, but MER across all channels comes out at 3.5, because paid social was driving halo effects in organic and email sales that platform ROAS could not see.2 Catalog ROAS has the same blind spot: it is real, but it is not the whole picture.
A good blended MER benchmark is often cited around 5.0 or higher, meaning $5 in total revenue for every $1 of marketing. The right target varies by margin, growth stage, and how much revenue is organic versus paid.2
How to use these numbers together
- Split Meta reporting by campaign type first. Do not read "Meta ROAS" as one number. Separate DPA retargeting from DABA prospecting before drawing conclusions.
- Set separate benchmarks for each. Roughly 2.5x to 6x for retargeting and 1.2x to 2.5x for cold prospecting. A prospecting campaign at 1.8x is not broken; it is in range.
- Track MER monthly, not just campaign ROAS. MER tells you whether the whole system, including halo effects, is profitable.
- Use ROAS to optimise, MER to decide. ROAS tells you which campaign, audience, or creative to adjust. MER tells you whether to grow, hold, or cut the budget overall.
1. DPA vs DABA ROAS ranges: Blip, "Dynamic Product Ads and Catalog Strategy."
2. MER definition, benchmark, and blended-ROAS example: Northbeam, "Marketing Efficiency Ratio (MER): Definition, Benchmarks, and How it Differs from ROAS."
FAQ
Why don't catalog ROAS and campaign ROAS match in Meta?
Because they measure different slices of the same spend. Catalog retargeting (DPA) reaches warm visitors and typically returns 2.5x to 6x ROAS, cold DABA prospecting typically returns 1.2x to 2.5x, and campaign ROAS blends whatever sits inside the campaign. Attribution settings and breakdown views add further differences, so the numbers diverge by design.
Why is my catalog retargeting ROAS so much higher than my prospecting ROAS?
Because they reach different people. DPA retargeting shows ads to people who already visited your site, viewed products, or abandoned a cart, which typically returns 2.5x to 6x ROAS. DABA prospecting reaches people who have never heard of you and typically lands at 1.2x to 2.5x. The gap is structural, not a sign either campaign is broken.
Should I judge DPA retargeting and DABA prospecting against the same ROAS target?
No. One blended target guarantees you either kill a healthy prospecting campaign too early or under-invest in a retargeting campaign that is actually performing. Set separate benchmarks: roughly 2.5x to 6x for retargeting and 1.2x to 2.5x for cold prospecting.
Why do my product or catalog breakdowns not add up to my campaign ROAS?
Breakdowns split purchases across the items shown and bought, and different views can use different attribution settings, so line items rarely sum to the campaign total. If breakdowns are empty or match rate is stuck low while spend runs, check for a content_id mismatch between your pixel and your catalog first.
Does the attribution window change catalog ROAS versus campaign ROAS?
Yes. A different click or view window can raise or lower reported ROAS without any change in real sales. Compare views using the same attribution window and the same date range, and agree one window for reporting before you judge any campaign.
Which ROAS should I trust: catalog ROAS, campaign ROAS or MER?
Use each for a different job. Catalog and campaign ROAS tell you which audience, campaign or creative to adjust. MER, total revenue divided by total marketing spend, tells you whether to grow, hold or cut budget overall. Check new-customer ROAS as well if retargeting is carrying the headline number.
What is MER and how is it different from ROAS?
MER (Marketing Efficiency Ratio, also called blended ROAS) is total revenue divided by total marketing spend across every channel. ROAS measures return at the campaign or channel level. ROAS tells you which campaign to optimise; MER tells you whether marketing spend is making the business more profitable overall, including halo effects ROAS cannot see.
Not sure which ROAS number to trust?
A growth operator will split your catalog and campaign reporting, calculate your real blended MER, and show where budget should move. Written roadmap in 24 hours.
Request a Growth Diagnosis →