The short answer
Meta publishes spend only for political and issue ads, so any competitor ad spend number is an estimate. The usable approaches are EU reach, engagement tracked over time, longevity and variant counts. Use estimates to rank ads, not to audit a P&L: on our board the median winning ad is estimated at $223K, and the top 10% start at about $1.07M.
Why Meta doesn't show you the number
The Meta Ad Library was built for transparency on political advertising, then extended to all ads. Spend made it into the first part and not the second. In Meta's API documentation for archived ads, the spend and impressions fields are both marked as available only for political and issue ads.
Ads delivered in the EU get more fields because of the Digital Services Act: total EU reach, reach broken down by country, age and gender, and the targeting settings. Still no spend.
So every competitor ad spend figure you see, in any tool, is a model. The useful question is not "what is the real number" but "which approach gets the ranking right, and where does it break".
Four ways to estimate competitor ad spend
| Approach | What you need | Works best for | Main weakness |
|---|---|---|---|
| EU reach | Ad delivered in the EU | Brands selling in Europe | Says nothing about US-only ads |
| Engagement over time | Reactions and comments, checked repeatedly | Comparing ads in the same niche | Engagement per dollar varies a lot |
| Longevity | Start date | Separating survivors from tests | Old ads often coast on small budgets |
| Variant count | "Multiple versions" and repeated creatives | Spotting what is being scaled now | Some brands always launch many versions |
1. EU reach times your own CPM
If the competitor's ad delivered in the EU, the library shows how many accounts it reached there. Multiply reach by an assumed frequency to get impressions, then by a CPM you trust (your own account's CPM in a similar niche is better than any published benchmark). You get a rough EU budget for that ad.
It is the closest thing to a hard number you will get. It also covers only the EU part of the campaign, and most US DTC brands spend the bulk of their budget at home.
2. Engagement tracked over time
Reactions, comments and shares accumulate as an ad gets delivery. Delivery costs money. If you record engagement on the same ad every week, the growth rate becomes a proxy for how much budget is flowing through it.
This is the approach Ad Radar uses: the estimate is modeled on engagement from the public Meta Ad Library, tracked over time. We don't publish the model, and you shouldn't trust anyone's exact coefficient anyway, because the ratio between engagement and spend shifts by format, niche, offer and audience. A funny pet video earns more reactions per dollar than a joint supplement ad. That is why the estimate is labeled as an estimate and is best used to rank ads, not to quote a budget.
3. Longevity
Days live is cheap to read and hard to fake. A brand doesn't keep paying for an ad for 200 days unless it works. But longevity is a weak proxy for size. On our board, the rank correlation between days live and estimated spend is only about 0.2. Old ads are proven, and often small.
Certification in Play Therapy
- Est. spend
- $88K
- Days live
- 553
- Format
- Image
A year and a half live with a modest estimate. Longevity proves the ad pays for itself. It doesn't prove the ad is big.
4. Variant count
Brands add variants to ads that work. On our board, the median estimated spend climbs with the number of variants:
| Variants | Ads | Median estimated spend |
|---|---|---|
| 1 | 1,743 | $177K |
| 2 | 390 | $301K |
| 3 to 4 | 189 | $399K |
| 5 to 9 | 93 | $613K |
| 10 or more | 28 | $1.12M |
The rank correlation with estimated spend is about 0.33, stronger than days live. Variant count is the best fast signal for "being scaled right now", and we cover it in variant count as a scaling signal. It is noisy at the brand level: some advertisers launch every ad with five versions by habit.
What the estimates look like across 2,443 winners
Every ad on the Ad Radar board has passed $50k in estimated spend. Here is how the estimates spread (snapshot of October 9, 2026):
| Estimated spend | Ads | Share |
|---|---|---|
| $50K to $100K | 511 | 21% |
| $100K to $250K | 820 | 34% |
| $250K to $500K | 536 | 22% |
| $500K to $1M | 311 | 13% |
| $1M to $5M | 237 | 10% |
| $5M or more | 28 | 1% |
$223K
median estimated spend
$1.07M
90th percentile
$236K
median for video ads
$173K
median for image ads
Niche moves the median less than people expect. Across the largest niches on the board, the typical winner sits between roughly $200K and $280K:
| Niche | Ads | Median estimated spend |
|---|---|---|
| Pet | 313 | $275K |
| Apparel and fashion | 81 | $282K |
| Joints | 208 | $237K |
| Skin | 224 | $232K |
| Food and nutrition | 164 | $229K |
| Digestion | 239 | $201K |
| Cleaning | 151 | $197K |
That narrow band is useful when you size a test. If the winners in your niche sit around $200K to $300K, an ad your competitor has fed for three months is probably in that range too, give or take a lot. The Facebook ad spend benchmarks article goes further on what these numbers mean for your own budget.
Young ads can carry big estimates. When a brand finds a winner, it often pushes it through several Pages at once, including creator Pages.
This is Maisie, and our story began three years ago.
- Est. spend
- $2.2M
- Days live
- 53
- Format
- Video, 40s
A creator Page sending traffic to a fresh dog food brand, under two months live and already estimated in the millions. Search only the brand's own Page and you miss this spend entirely.
Where every estimate breaks
Know these before you put a number in a deck:
- Shared posts. Several ads can run on the same post and share its engagement. On our board, 13 groups of ads from the same brand carry identical estimates for this reason. Treat them as one creative.
- Format bias. Video and entertaining creative collect more reactions per dollar than plain offer images. Compare within a format.
- Niche bias. Pet, beauty and food ads attract comments. Finance and B2B ads don't. Compare within a niche.
- Engagement bait. "Comment YES below" inflates engagement without spend. Read the comments before you trust the count.
- Pages you didn't search. A brand's total spend includes partner and creator Pages. Our Ad Library search tricks show how to find them.
- Stopped ads. Outside the EU, inactive ads vanish, and so does their spend history.
How to use an estimate properly
- Rank, don't quote. Use estimates to decide which five ads to study first, not to tell a client what a competitor's monthly budget is.
- Stack signals. An ad with a high estimate, 90+ days live and several variants is a near-certain winner. A high estimate on a three-day-old ad is a guess. The four-signal scorecard puts the signals together.
- Compare like with like. Same niche, same format, same time window.
- Write "estimated" every time. In reports and pitch decks, a spend figure without that word invites the wrong conversation. If you pitch with competitor data, our guide to using competitor ads in an agency pitch shows how to frame it.
If you want estimates already computed and sorted, the Ad Radar board shows every ad with its estimated spend, days live and variants side by side.
Figures marked as estimated spend come from Ad Radar's model of engagement on public Meta Ad Library ads. They are estimates, labeled as such, and are best used to rank ads against each other.