How many nutra offers actually run on Facebook

Where the numbers come from

A single measurement against our showcase database, taken on 20 September 2026. The vertical is decided by a model reading the creative itself, from video frames and text on the image, rather than from what the advertiser called their page. The domain is taken from the ad's destination address.

Every figure below states what it is a share of. A share of all ads and a share of the labelled portion are different claims, and swapping one for the other is a lie.

One ad in five goes nowhere

Of 3,524,945 nutra ads, 3,285,444 carry a destination. A destination is not a landing page, though. 685,936 of them, 21% of those with an address, point not at an offer's site but into a pipe: a messenger, an app store, a page inside Facebook itself, or a marketplace.

The largest nutra "domain" by advertiser count is not an offer at all. It is api.whatsapp.com, with 88,668 advertisers and 266,371 ads. Then come instagram.com and Facebook's own short links. Counting hosts like these as offers is meaningless, because anything at all can sit behind them.

That leaves 2,599,508 ads pointing at ordinary websites, and those are what the rest of this counts.

Forty percent of domains showed exactly one ad

153,492 domains sounds impressive right up until you look at their size. 60,672 of them, 40%, carried exactly one ad across the whole observation period. Those are test launches, abandoned drafts and single-use domains burned within a day.

• 134,264 domains (87%) have fewer than 15 ads;
• 9,645 domains clear a bar of 15 ads and at least two advertisers;
• 1,299 made it into our public directory after every check.

The gap between 9,645 and 1,299 is not strictness for its own sake. In between sit domains with too few creatives to show any variety of approach, and domains whose vertical was guessed from one ad in a hundred.

Bridge farms look like giant offers

72 domains each collected more than sixty distinct advertisers. These are not popular offers, they are infrastructure: one host through which dozens of independent teams run traffic.

• sarb.daily-buzzz.com, 233 advertisers, 10,648 ads;
• sarb.path-2-career.com, 228 advertisers, 8,216 ads;
• heyzeyla.us, 265 advertisers, 4,809 ads;
• trk.purehealthnest.site, 264 advertisers, 699 ads.

Sort nutra by volume in an ad spy tool and domains like these land at the top, looking like the hottest offers on the market. There is nothing to study in them: the creatives belong to different products and different teams, and all they share is a redirect.

A real large offer looks different

The same order of volume, but with a sane advertiser count:

• gloesse.com, 11,316 ads from 2 advertisers;
• get.tryrosabella.com, 10,109 ads from 19 advertisers;
• lymphoria.co, 6,582 ads from 27 advertisers.

That is what an offer looks like: a narrow circle of people running it and a large body of creatives for one product. The ratio of ads to advertisers is the fastest way to tell the two apart. A farm sits around forty; a live offer is in the hundreds.

What the market is made of

The subcategory is read from the ad's own text and is filled in for 10.3% of nutra, so the shares below are of the labelled portion rather than the whole corpus. The ordering holds regardless.

• Anti-aging and skin, 66,497 ads;
• Joints and back, 53,241;
• Weight loss, 42,314;
• Diabetes, 41,322;
• Male health, 22,937;
• Blood pressure, 19,788;
• Hair loss, 16,612.

The common assumption that nutra is mostly weight loss does not survive the count: anti-aging and joints together carry more ads than weight loss and diabetes combined.

What to do with this

One practical conclusion: the offer count in an ad spy tool's marketing says nothing about how much of it you can actually study. Counting domains is easy. Separating offers from pipes and bridges is hard, and that work is what decides whether you find something useful or spend an afternoon scrolling redirects.

We published the result of that separation openly: 1,299 offers, each showing its creatives with their copy, formats, geos and placements, with no account required. It is the same count described above, carried through to pages.

Method

The query runs over the showcase database: every ad labelled with the nutra vertical, the domain taken from the destination address, lowercased, with www stripped. Hosts belonging to messengers, app stores, Facebook itself, trackers and marketplaces are separated out by a pattern list. Advertisers are counted by page name. Measured on 20 September 2026.

What we deliberately do not count: creative lifespan and the distribution across countries. Facebook's end date matches the day our crawler saw the ad, and the geography in ad spy databases is the geo of the search basket rather than a campaign's targeting. Both can be turned into a number, but it would measure us and not the market.