What Just Launched in eCommerce Ads, October 2026: Montage Staged 1,860 of the Month's New Ads
Scope: every Meta ad the eCommerceBenchmark Ad Library first saw in October 2026, across every industry and country it tracks. That is 6,858 ads from 340 brands. A launch is the day the library first saw the ad, and it finds ads late, so every count here is a floor. No split by country.
Picture a month of brand-new ads, stacked in one pile.
In October 2026 that pile held 6,858 new Meta ads. They came from 340 brands. Not the ads that had been running all year: the ones the library had never seen before.
Most were staged the same handful of ways. Montage staged 1,860 of them. A discount rode on 1,207. The static, video and carousel split is another piece in this series; video led it with 3,075 launches.
One more thing, and it is the odd part. Every one of the month's longest-running launches we can show has been on air 41 days. Every one was first seen on 1 October. That is a fact about when the library looked, not about which ads held up.
- How they staged it is already decided for most teams. Montage staged 1,860 of October's launches. Cinematic b-roll staged another 1,460. If your next batch is both, you are shipping into the two busiest lanes of the month.
- How they opened it barely varied. Visual reveal opened 1,424 launches. A bold claim opened 1,371. Newness opened 1,145, so three moves cover the front of the month.
- What they offered was mostly one thing. Discount rode on 1,207 launches. Nothing here says it sold anything. It says what brands put on the table.
Here is what brands made, how they opened it, what they argued and what they put on the table.
Montage staged 1,860 of October's launches
A creative concept is defined as how an ad is staged on screen, such as a montage of cuts or a close-up of the product in use. Montage staged 1,860 of the 6,858 ads the library first saw in October, ahead of every other concept it labels. Cinematic b-roll came next at 1,460. Product demo close-up staged 830, an offer-first banner 714, and a selfie talking-head 300.
Count of ads the library first saw in October 2026, by how each one was staged, largest first. Source: eCommerceBenchmark Ad Library, 2026.
Two lanes carry most of the month's staging. Read that as a map of what is crowded, not as a scoreboard.
So that is the staging. What did these ads do in their first seconds?
Visual reveal opened 1,424 of them
A hook is defined as the opening move an ad makes in its first seconds or its first line, before any argument. Visual reveal was the opening move on 1,424 of the month's launches. A bold claim opened 1,371. Newness opened 1,145.
Count of ads the library first saw in October 2026, by the opening move they make, largest first. Source: eCommerceBenchmark Ad Library, 2026.
After those three, the drop is steep. A direct question opened 638. A price anchor opened 517. Relatability opened 465.
The argument underneath was narrower still. Solution-aware, the pitch that names a problem and puts the product in front of it, was the angle on 2,478 launches. Identity took 1,958. Mechanism, the how-it-works case, took 754. Social proof carried 420, and an objection crusher 332. Named enemy, the ad that picks a fight with something, turned up on 7.
Opening and argument are set. What were brands putting on the table?
Discount was the offer most launches carried
Discount was the offer on 1,207 of October's launches, the only offer past a thousand. Bundle came next at 343, and every other offer the library labels sat lower. An offer here is what the ad puts in front of the shopper: money off, more product for the money, something free with the order.
So a shopper working through one category in October met the same kind of deal over and over. That count says what brands made. It does not say what any of it earned.
Who was doing most of this launching?
Fashion & Apparel launched 2,123 of them
Fashion & Apparel put out 2,123 of the month's launches, ahead of every other industry. Health & Wellness followed with 1,531. Beauty & Personal Care launched 1,305. Home & Furniture launched 848, and Food & Drinks 543.
| Industry | Ads first seen in October |
|---|---|
| Fashion & Apparel | 2,123 |
| Health & Wellness | 1,531 |
| Beauty & Personal Care | 1,305 |
| Home & Furniture | 848 |
| Food & Drinks | 543 |
Three industries make up the bulk of the pile. If you sell in one of them, the lanes above are the lanes your shopper scrolls past all day.
Now the part that looks like a result and is not.
October's launches that ran longest
The longest-running October launches we can show are all tied at exactly 41 days on air. Every one of them was first seen on 1 October. That tie is not a durability finding: it is the library's read window showing through the data. A long run only means a brand kept the ad live, not that the ad sold anything.
Of the month's 6,858 launches, 5,580 were on air when the library was read.
victoriabeckham.com leads with the product: "Sculptural eyewear for modern classic style". elvie.com opens on the situation: "Pregnant and thinking about pumping? Me too!" heights.com opens in a customer's own words: "I just couldn't engage in the conversation I was so tired". gozney.com names the result: "Unbox your inner chef". oatsovernight.com puts the product first: "Get Pumpkin Pie while you can. 20g protein." mykitsch.com opens on regret: "I regret not getting Kitsch's FREE Welcome Kit sooner!" Each has been on air 41 days, and that's time on air, not sales.
Pick one of those openings and test it. Then watch what you do with the counts.
Three tests from October's launch mix
- A montage against your control. Run one montage build next to your current best ad, same offer and same audience. Montage staged 1,860 of October's launches.
- An opening that holds the payoff back. Rebuild your top ad with the product shown first and the claim second. That opening move was used on 1,424 of the month's launches.
- A no-deal cut of your discount ad. Same creative, offer line swapped for the product claim. Discount was the offer on 1,207 launches.
The eCommerceBenchmark Ad Library sorts your ads into the same hooks, angles, concepts and offers used in this report and shows your mix next to the market, so you can see which shapes you have never shipped.
Audit my creatives, freeThose three tests only pay off if you read the counts the right way. Here is how they usually go wrong.
The mistake: copying what launched most because it launched most
A big launch count is a record of what brands made, not a result. Treating 1,860 montage launches as a verdict puts your next batch in the busiest lane of the month with nothing behind the move.
Do that at the same time as everyone else, and a shopper in one category meets the same cut-driven staging, the same opening and the same money-off line on every ad in the feed. They stop reading any of them. Your ad is left to win on staging alone. The useful move runs the other way: find the lane October left thin, and put it next to your own control.
Your test card
Here is the first test from this article, written the way a media buyer would set it up. Fill in your own numbers before you launch.
| Hypothesis | A montage build, added next to my current control with the same offer and audience, will hold my cost per purchase without taking budget from the control. |
| Change only this | The staging: a cut-driven montage instead of the single-scene control. |
| Keep the same | Offer, opening line, audience, placements, landing page, budget and bid setup. |
| Funnel stage | Say first whether this ad is for prospecting (cold) or retargeting (warm). A deal in the first line is often right for warm audiences and wrong for cold ones, so test it where the ad will really run. |
| Setup | One control and one variant, same audience, same placement, even budget split. Do not edit either ad while the test runs. |
| How long and how much | Until each version has about 50 purchases, or two full weeks, whichever comes later. This is general testing practice (Meta's own learning-phase guideline is about 50 results per ad set per week), not a library finding. |
| Success metric | Cost per purchase, not clicks. |
| Stop rule | Pause a variant that has spent about 3 times your target cost per purchase with no purchase. If it is close, keep it running. |
| What counts as a win | The variant beats the control on the metric by an amount you would act on, and still does in a second week. |
Key takeaways
- 6,858 new Meta ads were first seen in October 2026 [eCommerceBenchmark Ad Library, 2026], a pile big enough to read as a market rather than as your own feed.
- 340 brands launched at least one of them, and that is the sample behind every count on this page.
- Montage staged 1,860 launches, more than any other concept, so it is the month's most crowded staging lane.
- Visual reveal opened 1,424 launches, with a bold claim and newness close behind, so three moves cover the front of the month.
- Solution-aware was the argument on 2,478 launches, ahead of identity, so most of the month argued the same way.
- Discount rode on 1,207 launches, the only offer past a thousand: strip it out of one ad to see what it is carrying.
- Fashion & Apparel launched 2,123 of the month's ads, ahead of every other industry in the cut.
Questions
Sort your last month of ads into the same four boxes this piece used: how each one was staged, how it opened, what it argued and what it offered. Then look for the box you left empty. Montage staged 1,860 of October's launches, and a discount rode on 1,207 of them. Those lanes are full across the market. The empty box is the one worth a test before your next budget cycle, judged on your own numbers and nobody else's launch count.