Start with a Sprint

Ecommerce Facebook ad statistics: What 300 live ecommerce ads on Meta look like right now.

These ecommerce Facebook ad statistics come from 300 live ads by 164 product brands in Meta's public Ad Library, read on 10 November 2026. Most led with no discount, 50% were running more than one version, and still images outnumbered video.

By Vistalis · · Download the data (CSV)

Ten findings

  1. 69% (208 of 300) of the ads had been running for 90 days or more, and the median ad had been live for 145 days.
  2. 74% (222 of 300) mentioned no discount, free shipping or bundle in their text, weeks before Black Friday.
  3. Discounts were the most common offer, in 22% (66 of 300) of ads, from 10% (5 of 50) in skincare to 38% (19 of 50) in supplements and vitamins.
  4. Free shipping almost never appeared in the ad text: 1% (4 of 300).
  5. Still images led the format mix at 57% (170 of 300), against 42% (125 of 300) video and 2% (5 of 300) carousel.
  6. Video use ranged from 60% (30 of 50) in pet products to 16% (8 of 50) in apparel and activewear.
  7. 50% (149 of 300) carried Meta's "This ad has multiple versions" label.
  8. "Shop now" was the button on 62% (186 of 300) of ads; "Learn more" on 8% (24 of 300).
  9. 98% (294 of 300) ran on both Facebook and Instagram.
  10. The median ad text ran to 35 words; 47% (141 of 300) used an emoji, 10% (29 of 300) cited reviews or customers and 11% (34 of 300) used urgency.

How long had the ads been running?

Long. On the day we read them, the median ad had been live for 145 days, 95% (286 of 300) had run for at least a month, and only 5% (14 of 300) were under 30 days old.

Read that with care. The Ad Library lists results by total impressions, so the ads it shows first are the ones that have reached the most people, which favors ads that have been running a long time. What it does show is that brands keep ads live for months rather than swapping them out every week.

Median days running, by category
CategoryMedian
Home and kitchen326 days (29 ads)
Haircare248 days (32 ads)
Skincare164 days (50 ads)
Pet products144.5 days (50 ads)
Supplements and vitamins139 days (50 ads)
Apparel and activewear119 days (50 ads)
Snacks and food63 days (32 ads)

Image, video or carousel?

Still images are still the workhorse: 57% (170 of 300) of the ads were a single image, 42% (125 of 300) were video and 2% (5 of 300) were carousels.

The mix depends on what is being sold. Video outnumbered images in supplements and vitamins, pet products, and snacks and food, while only 16% (8 of 50) of apparel and activewear ads were video, the lowest of any category. Carousels were read from the Library card, which shows a carousel as more than one large image.

Share of ads that were video, by category
CategoryVideo
Pet products60% (30 of 50)
Supplements and vitamins54% (27 of 50)
Snacks and food53% (17 of 32)
Skincare38% (19 of 50)
Haircare34% (11 of 32)
Home and kitchen31% (9 of 29)
Apparel and activewear16% (8 of 50)

How many ads led with an offer?

Fewer than you might expect, even in mid-November. 74% (222 of 300) of the ads mentioned no discount, free shipping or bundle in their text. Discounts appeared in 22% (66 of 300), bundles or sets in 6% (18 of 300) and free shipping in just 1% (4 of 300).

One caution: the study's rules count a discount, a sale, free shipping or a bundle as an offer. A free gift with purchase is not counted, so a few ads that give something away sit in the no-offer group.

Share of ads mentioning a discount or sale, by category
CategoryDiscount or sale
Supplements and vitamins38% (19 of 50)
Apparel and activewear38% (19 of 50)
Pet products20% (10 of 50)
Home and kitchen17% (5 of 29)
Haircare13% (4 of 32)
Snacks and food13% (4 of 32)
Skincare10% (5 of 50)

How many ads run more than one version?

50% (149 of 300) of the ads carried Meta's "This ad has multiple versions" label, which Meta shows when an ad rotates more than one creative or text.

The study could not count how many versions each ad ran: that number sits in a part of the Library card it did not read. And because it took at most five ads per brand, it understates how many ads each brand runs in total.

Share of ads labeled "multiple versions", by category
CategoryLabeled
Pet products60% (30 of 50)
Supplements and vitamins58% (29 of 50)
Snacks and food56% (18 of 32)
Home and kitchen55% (16 of 29)
Haircare44% (14 of 32)
Apparel and activewear42% (21 of 50)
Skincare38% (19 of 50)

How do the opening lines start?

Each ad's first line was sorted into one of seven types by a fixed set of keyword rules, applied the same way to every ad. Most lines matched none of them: they are plain statements of the product or the brand.

Bold claims and questions were the most common named openings, and the problem-led opening was rarer than either.

Share of ads by the type of their first line
First lineShare
Other54% (162 of 300)
Bold claim11% (34 of 300)
Question10% (31 of 300)
Problem or pain point8% (25 of 300)
Testimonial or quote5% (15 of 300)
Number or statistic4% (13 of 300)
Offer-led4% (12 of 300)
Curiosity or tease3% (8 of 300)
  • questionTired of trying product after product with no real relief?Sven's Island, Library ID 26435354126147211
  • bold claimMust-haves for glowing skin✨ Discover science-backed, clinically-tested, & award-winning formulas.Alastin Skincare, Library ID 994791200034778
  • problem or pain pointMelasma and dark spots that had been there for years.Meditherapy.Global, Library ID 1054635993660613
  • testimonial or quoteOne of my high school students asked what skincare I used.DRMTLGY, Library ID 1590835822691704
  • number or statisticMake the switch to a cat litter that's 10x better controlling your stinky litter box odors than clay!Catalyst Pet Natural Litter, Library ID 1348723359686427
  • offer-ledGet a FREE full-sized 2oz Bottle of our famous Face Oil when you buy our night and day serum.Beauty From Bees, Library ID 676321291971283
  • curiosity or teasePOV: Your entire skincare routine fits in your pocketSaeskyn, Library ID 2032403464184564

Which buttons, placements and text lengths?

"Shop now" was the button on 62% (186 of 300) of the ads, and 26% (78 of 300) showed no link or button on the Library card. 98% (294 of 300) ran on both Facebook and Instagram, most of them on Messenger, Threads or the Audience Network too.

The median ad's text ran to 35 words. 47% (141 of 300) used at least one emoji, 10% (29 of 300) mentioned reviews, ratings, customers or press, and 11% (34 of 300) used urgency such as "limited" or "ends soon".

What should a brand take from this?

Make more versions, not one hero ad. 50% of the ads were rotating several versions, and brands kept ads live for months. Our guide to how many ads to test on Meta sets the number for your budget.

Do not skip still images. They outnumbered video overall and in 4 of the 7 categories with enough ads to report; our static ad examples show what makes one read at a glance.

You do not need a discount to sell. Most ads carried none, even weeks before Black Friday. If price is the worry, our ads that handle price objections show how to answer it without cutting it, and our Black Friday ad examples cover the season when you do.

Write the first line to stop the scroll. Most openings were plain statements; a question, a problem or a specific claim gives a reader a reason to stay, which is what hook rate measures.

How was the data collected?

Every ad came from Meta's public Ad Library website on 10 November 2026, set to all ads, the United States and active ads only, read in the order the Library showed them. No spy tool, third-party database or API was used.

We searched three plain keywords in each of eight product categories, kept only product brands selling their own products online, and took at most five ads per brand, in the Library's order, so no ad was picked by preference. Marketplaces, retailers selling many brands, creator and partnership pages, advertorials, services and political ads were skipped. That left 300 ads from 164 brands.

4 of the 8 categories reached the 50 ads aimed for; the others stopped at the true number found. A category with fewer than 20 ads is counted in the totals but given no figures of its own.

Searches run in each category
Skincareserum, moisturizer, skincare
Supplements and vitaminssupplement, vitamins, gummies
Haircareshampoo, hair growth, conditioner
Pet productsdog treats, cat litter, dog supplement
Coffee, tea and drinkscoffee, tea, energy drink
Apparel and activewearleggings, activewear, t-shirt
Home and kitchencookware, kitchen, bedding
Snacks and foodsnacks, protein bar, chips

What are the limits?

  • The Ad Library shows no spend, reach outside the EU, clicks or sales, so nothing here measures performance.
  • Days running is a proxy at best: it shows a brand kept an ad live, not that the ad worked, and the Library's ordering by impressions favors long runners.
  • This is a structured sample from keyword searches, not a random sample of every ad on Meta.
  • It is one reading, taken in mid-November before Black Friday; offers and urgency may differ at other times of year.
  • Opening lines and offers were classified by fixed keyword rules from the ad text: reproducible, but blunt.

Questions about the study

Do long-running Facebook ads perform better?

This data cannot say. The Ad Library shows how long an ad has run but not what it spent or sold, and it lists the ads with the most impressions first. A long run suggests a brand was happy to keep paying for an ad, which is a hint, not proof.

Can I use these figures?

Yes. Quote them with a link to this page as the source. The dataset behind every figure can be downloaded from this page as a CSV, with each ad's Library ID so anyone can check it.

Will the study be repeated?

Yes, each quarter, with the same method, so the figures can be compared over time and across the selling seasons.

How to cite this study

Vistalis, "Ecommerce Facebook ad statistics: 300 live ads in Meta's Ad Library", 10 November 2026, https://www.vistalis.net/research/ecommerce-facebook-ad-statistics

Download the data (CSV)

Source

Meta, Ad Library, read on 10 November 2026 (United States, all ads, active ads only).

Keep reading: How many ads to test · Teardowns · Compare

Last updated

Want eight ads built on what the live ads show?

Start with a Sprint

8 ads in 7 days, each testing one idea.

Book your Sprint.

The Sprint is $650, one-off. Enter your brand's website, pay securely with Stripe, then send us a ten-minute brief.

How we use your details: our Privacy Policy.