Search this question and every result gives you the same number: three to five ads per ad set. Then you open the Meta Ad Library, look at a brand you admire, and count 400 live ads. One of those numbers is lying to you, and it is worth knowing which.
Neither is, actually. They are answers to two different questions, and the gap between them is where most of the confusion about ad volume lives. So we counted.
On 3 August 2026 we pulled the live active-ad count for 27 brands straight from Meta's public Ad Library. Not a sample of each account, the full population count for every one. The median brand was running 126 live ads. The smallest ran 10. The largest ran 966.
27 brands · active US ads per Page · Meta Ad Library via SOCIALFUEL · measured 3 Aug 2026
The short answer
Three to five ads per ad set is real guidance, and it is about ad sets. Meta's own delivery advice is to keep the number of ads inside a single ad set low, because budget splits across them and each one needs enough delivery to exit the learning phase. That is a statement about how the auction allocates a budget you have already set.
The number you see in the Ad Library is the whole account. It is every active ad across every campaign and every ad set on that Page, added together. A brand running five ads in each of thirty ad sets shows up as 150 live ads and has broken no guidance at all.
So the honest answer to "how many ads should I run at once" is: three to five per ad set, and as many ad sets as your budget can actually feed. The account-level total is an output of that, not a target you pick.
What 27 brands were actually running
Here is the distribution across the 27 accounts we counted.
| Live active ads | Brands | Share |
|---|---|---|
| 10 to 49 | 7 | 26% |
| 50 to 99 | 4 | 15% |
| 100 to 249 | 9 | 33% |
| 250 to 499 | 3 | 11% |
| 500 or more | 4 | 15% |
| Median account | 126 ads | — |
The quartiles tell it faster. A quarter of these brands ran 44 live ads or fewer. A quarter ran 349 or more. The middle of the pack sat at 126.
A few individual counts, since the shape matters more than the average:
- Sephora 966, The Ridge 592, SKIMS 586, Hims 542
- Factor 478, Gymshark 431, Huel 349, Liquid Death 249
- HelloFresh 189, Dior 158, Warby Parker 126, Nike 82
- Squarespace 69, Casper 44, Red Bull 32, Bombas 28
- Oakley 15, Adobe 10, MANSCAPED 10
The spread is the finding, not the median
The gap between the smallest and largest account here is roughly 96 times. That is the part worth sitting with.
Adobe and Nike are not smaller companies than The Ridge. They were running 10 and 82 live ads against The Ridge's 592. Red Bull, a brand most marketers would name in the first ten seconds of a branding conversation, was running 32.
Live ad count does not measure how big a company is, how much it spends, or how well its ads work. It measures how much simultaneous creative testing an account is set up to do, and that is a structural choice. Some brands push volume through the ad account. Some push it through channels that never touch the Ad Library at all.
Anyone quoting a single "right" number for ad volume is quoting the habits of whichever accounts they happened to look at.
Two accounts can be big in completely different ways
This is where a raw count gets misleading, so it is worth breaking open.
Among the 30 highest-impression ads on each Page on the day we measured:
- Sephora's were 27 of 30 dynamic formats (14 DCO, 13 DPA), carrying only 9 distinct headlines between them.
- The Ridge's were 26 of 30 standard video and image ads, carrying 16 distinct headlines.
Both accounts are near the top of our list. They got there by different routes. Catalogue and dynamic-creative setups mint ad records automatically, one per product or per generated combination, so a large retail catalogue can produce hundreds of live ads without a person briefing hundreds of creatives. A count of 966 is not 966 ideas.
That sample is the top 30 by impressions on each Page, not the whole account, so treat it as a texture check rather than a full format breakdown. It is enough to make the point: before you copy a competitor's ad volume, check what kind of volume it is.
Meta's own ceiling is higher than almost anyone reaches
Meta caps active ads per Page, on a tier that moves with the account's peak monthly spend in the previous year. The entry tier allows 250 active ads, rising in steps to 20,000 for the very largest advertisers (Meta's documentation).
Twenty of our 27 brands sat under that entry-tier ceiling. The cap is not what limits ordinary advertisers, and it is not a target either. Budget is the real constraint: every extra ad divides the same money into thinner slices, and an ad that never accumulates enough delivery to leave the learning phase teaches you nothing while still costing you.
So what should you actually run?
The useful version of this question is not a number, it is a ratio. Work backwards from budget:
1. Per ad set, stay in the 3 to 5 range. This is the one place the standard advice is exactly right. 2. Only add an ad set when it has a real job — a different audience, a different offer, a different geography. Not to raise a total. 3. Make sure every ad set can fund its own learning. If splitting the budget again leaves each ad set unable to gather enough conversions, you have found your ceiling regardless of what a competitor runs. 4. Judge the account by what survives, not what launches. Ad volume is an input. Ad longevity is the closest thing to a free performance signal Meta gives away, because nobody keeps paying to run a loser. We measured that separately in how long Facebook ads actually run.
A brand running 400 ads is not beating you because of the 400. It is testing at that width because its budget supports it, and most of those ads will be gone within a month.
How we measured this, and what we cannot tell you
Every figure here came from Meta's public Ad Library on 3 August 2026, read through socialfuel.io.
For each brand we requested the active-ad set for its Page filtered to ads delivering in the US, and read the population count the Ad Library reports, rather than counting the rows on the first page of results. That distinction matters more than it sounds: results come back sorted by impressions, so counting one page of results would have measured our own page size instead of the brand's account. We validated the reported count by paging two accounts to exhaustion, and it matched exactly both times (Bombas 28 of 28, Poppi 38 of 38).
What this does not show you:
- Nothing about spend. Meta withholds spend and impression figures for ordinary commercial advertisers, so anyone publishing a dollar estimate against a brand's ad library is publishing a model, not a measurement. We do not estimate it.
- Nothing about performance. A live ad is an ad someone has chosen to keep running. That is a signal, not proof.
- One Page per brand. Large advertisers often run several regional Pages, and the Ad Library counts per Page, so a global total may be higher.
- A moving number. HelloFresh reported 194 live US ads when we measured them on 31 July and 189 on 3 August. Any count like this is a photograph, not a fact, which is why the date is pinned to every figure here.
- A convenience sample. These 27 brands are ones already tracked in our dataset plus well-known consumer names, not a random sample of advertisers.
Frequently asked questions
How many Facebook ads should you run at once? Three to five ads per ad set, which is Meta's own delivery guidance. Your account-level total is whatever that adds up to across the ad sets your budget can genuinely fund. Across 27 brands measured on 3 August 2026, the median account had 126 live ads.
How many ads do big brands run on Facebook? It varies enormously. Among 27 brands measured on 3 August 2026, counts ran from 10 (Adobe, MANSCAPED) to 966 (Sephora), with a median of 126. Roughly 59% were running 100 or more live ads.
Is there a limit on how many Facebook ads you can run? Yes. Meta caps active ads per Page, starting at 250 for the entry tier and rising with the account's peak monthly spend. Most advertisers never approach it.
Does running more ads mean spending more? No. More ads split the same budget into smaller pieces. That is the actual risk of high ad counts on a modest budget: ads that never gather enough delivery to leave the learning phase, so you pay for data you cannot read.
Why does a competitor have hundreds of live ads? Often because of how the ads are generated rather than how many were briefed. Catalogue and dynamic-creative campaigns create ad records automatically, so a retailer with a large catalogue can show hundreds of live ads built from a handful of headlines.
How can I see how many ads a competitor is running? Meta's Ad Library is public and free, and shows active ads per Page. It is page-scoped and sorted by impressions, so the count you want is the reported total, not the number of results you can see. socialfuel.io reads the full set and decodes the hook, angle and offer behind the ads that are actually surviving, which is the part a raw count cannot give you.
Method: live active-ad counts for 27 Pages, filtered to ads delivering in the US, read from Meta's public Ad Library via socialfuel.io on 3 August 2026. Counts are the Ad Library's reported population per Page, validated against exhaustive pulls on two accounts. Format samples cover the 30 highest-impression ads per Page and are not a full-account breakdown. No spend, impression or performance data is claimed or implied. None of the brands named are affiliated with SOCIALFUEL.