How long does a skincare ad last before it stops working?

Between 19 and 21 days, according to the only two analyses that have measured it.

Both studies land inside the shaded band. The grey bar is how long a shoot takes, and it is still going when the ad it was made for has already stopped working.

Web Tonic put the median skincare ad lifespan at 19 days in 2026, and found that 31.6 percent of creative is replaced inside the first two weeks. BeautyMatter, in Beauty Ad Intelligence 2026, analysed 500 beauty ads from 50 brands across 17 countries and put the median at 21 days.

Two independent sources, two days apart. That is unusual agreement for a number this widely quoted.

What the number actually measures, because it matters

Both figures measure how long an ad stays live before it is paused or swapped out. Neither measures the moment it stopped working.

That distinction gets lost almost everywhere this statistic appears, and it is worth keeping, because the two are not the same thing. An advertiser might swap an ad at day 19 because performance slid at day 12. Another might leave one running for a month because nobody got to it. What the data shows is behaviour, not decay.

It is still the most useful number in beauty paid social. It tells you the working life of a piece of creative is measured in weeks, not quarters, and that the people who are good at this already behave as though that is true.

One more caveat, since I am being precise about the rest. BeautyMatter's 21 days is the median for top performing ads. Web Tonic's 19 days is the median across all skincare ads. Those are not identical populations. They agree closely. They are not measuring the same thing.

What the data shows is behaviour, not decay.

The number that follows from it

Both sources report the same thing about the advertisers doing well: they keep 8 to 12 new creative assets in rotation per month.

That figure is the practical version of the 19 day one. If the working life of an ad is about three weeks, and you want something ready before the current set tires, you need a pipeline rather than a shoot.

Most beauty brands at 1 to 5 million in revenue produce 2 to 4 a month. That gap is the whole problem.

Both studies report the same number. The gap between the two bars is the problem, and it is about how much you can make, not how good it is.

What a shoot costs against what it gives you

A product shoot in the United States runs 2,000 to 5,000 US dollars, takes two to four weeks, and gives you one set of images. No video. No second angle to test. Nothing written down about which frame did the work.

At a 19 to 21 day working life, that set is into its second cycle before the invoice is paid.

That set is into its second cycle before the invoice is paid.

What I would do about it

I build the next set before the current one tires, and each set is built from what performed in the last one.

Every asset goes into a record: what it was, which angle, which hook, which format. Each month I read your numbers against that record. What scaled gets varied. What died gets dropped. What has not been tried yet moves up.

By month three you can see which angles hold for your brand and which die.

Send me your product page and I'll build you three ads, plus a page on what I'd test first for your brand. Back in 24 hours, no card, no call.

Sources

Web Tonic, 2026. Median skincare ad lifespan, 19 days. 31.6 percent of creative replaced within two weeks. https://www.webtonic.io/blog/beauty-skincare-ad-creative-statistics

BeautyMatter, Beauty Ad Intelligence 2026. 500 beauty advertisements from 50 brands across 17 countries. Median lifespan 21 days for top performing ads. https://beautymatter.com/articles/what-beautys-best-ads-reveal