Before you put "20% off" into a Facebook ad, check what the market already does. We sorted 861 live Meta ads from 18 advertisers by the offer in their copy: how many carry a discount, which kinds advertisers keep paying for, and how deep they go.
The advice splits two ways: discounts create urgency, or discounts cheapen you. One Shopify merchant asked a community thread whether to keep running 20% off sales or just lower the price. Almost none of the pages that rank for this question count what advertisers actually run.
Share of each offer group still running 90 days after launch. Measured 7 September 2026.
The short answer
Most advertisers don't discount in the ad, and a discount alone doesn't make an ad last. The kind of discount does.
- Only 82 of 861 live ads (9.5%) state a discount in their copy. Another 64 (7.4%) offer a perk with no price cut, like free shipping. 715 (83.0%) offer nothing but the product.
- 10 of the 18 advertisers ran no discount ads at all, including Dior, Red Bull and Liquid Death.
- Pooled, discount ads look longer-lived, with a median of 35.5 days against 24. Brand by brand, that holds at five advertisers and flips at three.
- Standing offers to new customers account for 10 of the 11 discount ads that ran past 90 days. Time-boxed sales: 0 of 20.
What we measured
We hold 3,120 stored Meta ad records. Deduplicated by advertiser and ad copy, kept only if live when captured, and with two Pages belonging to ad-research software companies set aside, that leaves 861 creatives across 18 advertiser Pages. Everything is pinned to the last capture, 7 September 2026.
Age runs from launch to the last time we saw the ad live, so every age is a floor. Read "ran 90 days" as at least 90 days.
We read the primary text and the line of link text under the creative, then sorted each ad:
- Discount: a stated price cut. Percent or dollars off, a sale, a discount code, bundle savings, or a new-customer deal applied as money off, like HelloFresh's "10 free meals", which its fine print calls a discount on the first box.
- Perk only: free shipping, a free gift, a free trial or bonus points, with no price cut.
- No offer: neither.
We checked every discount and perk ad by hand, including ads in German, Dutch, French and Arabic.
Two limits. An offer that lives only inside the image or video is invisible to a text count. And longevity is a proxy, because Meta publishes no spend or results for commercial advertisers (we checked 2,108 records for exactly this). It shows what the payer chose to keep alive, the most honest public signal there is.
Discounts are the exception
| Brand | Live ads | Discount ads | Perk-only ads |
|---|---|---|---|
| instant.one | 111 | 4 | 0 |
| Top Athlete | 103 | 15 | 5 |
| Dior | 97 | 0 | 2 |
| True Protein | 80 | 0 | 19 |
| pay.com.au | 75 | 0 | 19 |
| Red Bull | 64 | 0 | 0 |
| HelloFresh | 56 | 32 | 0 |
| Organic Performance | 41 | 9 | 4 |
| Liquid Death | 38 | 0 | 0 |
| Gymshark | 37 | 9 | 1 |
| Adobe | 33 | 3 | 0 |
| Dunkin' | 26 | 0 | 5 |
| ToeSox | 26 | 6 | 0 |
| ToeSox Australia | 25 | 4 | 0 |
| Oakley | 21 | 0 | 4 |
| Death to Stock | 11 | 0 | 5 |
| Caterpy | 9 | 0 | 0 |
| The Great Aussie Bush Camp | 8 | 0 | 0 |
Ten of the 18 never state a discount. Dior ran 97 live ads without one. Red Bull, Liquid Death, Caterpy and The Great Aussie Bush Camp offered nothing at all in their copy.
True Protein is the interesting one. Zero price cuts across 80 live ads, yet 19 of them (23.8%) lead with a perk: a free shaker, a free gift with a first order, free shipping over a threshold. pay.com.au does the same with bonus points and 30-day free trials in 19 of its 75. Both make an offer without touching the price.
At the other end, HelloFresh has 32 of its 56 live ads (57.1%) stating a discount. That one advertiser supplies 39.0% of every discount ad in the dataset.
Why "discount ads last longer" is the wrong conclusion
Pool everything and discounts look like a winner. The 82 discount ads have a median age of 35.5 days and the other 779 have 24. That is 48% longer, exactly the number a vendor blog would headline.
The fair test compares ads inside one account, where product and budget are roughly constant. Eight advertisers ran at least three ads in each group:
| Brand | Discount ads | Median days | Other ads | Median days | Difference |
|---|---|---|---|---|---|
| instant.one | 4 | 93 | 107 | 11 | +82 |
| Gymshark | 9 | 67 | 28 | 25.5 | +41.5 |
| ToeSox Australia | 4 | 74 | 21 | 62 | +12 |
| Top Athlete | 15 | 34 | 88 | 24 | +10 |
| Organic Performance | 9 | 9 | 32 | 7 | +2 |
| Adobe | 3 | 8 | 30 | 23 | -15 |
| ToeSox | 6 | 3.5 | 20 | 34 | -30.5 |
| HelloFresh | 32 | 24 | 24 | 55 | -31 |
Five longer, three shorter. The biggest positive gap rests on four instant.one ads set against 107 short-lived ones. The biggest negative gap belongs to the heaviest discounter: HelloFresh's discount ads have a median of 24 days, against 55 for its ads that don't.
That carries its own warning. Copy is only one place an offer can live, and our HelloFresh ad teardown found an offer in the headline of every HelloFresh ad still running past 60 days. A discount is not a lever you can pull for longevity.
The kind of discount is what matters
Split the 82 discount ads by what the discount is for:
| Offer in the copy | Ads | Median days | Ran 90+ days |
|---|---|---|---|
| Entry offer (first order, first box, first month, student) | 39 | 31 | 10 (25.6%) |
| Always-on price cut (bundle savings, dollars off a product) | 23 | 44 | 1 (4.3%) |
| Time-boxed sale ("limited time", "sale ends") | 20 | 45 | 0 (0.0%) |
| Perk only | 64 | 24 | 7 (10.9%) |
| No offer in the copy | 715 | 24 | 59 (8.3%) |
Of the 11 discount ads that ran past 90 days, 10 are entry offers, from HelloFresh (5), Gymshark (2), instant.one (2) and Adobe (1). The oldest is Gymshark's 10% off your first order via an app code, live for at least 147 days. Next is HelloFresh's 10 free meals plus free breakfast for life, at 131.
Entry offers have the lowest median of the three discount types, 31 days, because HelloFresh keeps launching new ones: 16 of its 23 entry-offer ads went live after 1 July. The difference is in the ceiling. An entry offer is a permanent part of how an advertiser wins new customers, so its best versions run for months. A sale is an event, and events end.
Time-boxed sales never got past 90 days. The longest lasted 81. "Limited time" is loose, though: 15 of the 20 ran at least 30 days, and one supplement ad whose copy says "This week only" was still live at least 56 days after launch.
Small groups, few advertisers: treat this as a pattern to test, not a law.
Where the offer goes, and how big it is
- Primary text. 74 of the 82 discount ads (90.2%) state the offer in the primary text. 23 (28.0%) state it in the link text under the creative, and only 8 leave it to that line alone.
- 20% is typical. 32 ads state a percentage. The median is 20% off, with two clusters: 15% (8 ads) and 50% (8 ads).
- Dollars sit around $36. 18 ads state a dollar saving instead, with a median of $36.15 and a range from $13 to $400.
- Perks ride along. 21 of the 82 (25.6%) stack a freebie on top, usually free shipping, a free gift or both.
What the research says
Shirsho Biswas's 2020 doctoral research at the University of Chicago's Booth School tested this on a food delivery app. Discount ads drew more searches than brand ads, but fewer searchers bought, and advertising on price reduced brand preference, a "discount spotlighting" effect. That matches the market: most advertisers keep discounts out of the copy, and the ones that last target people who have never bought.
What to do with this
- Don't discount by default. 83.0% of live ads state no offer at all.
- If you discount, make it an entry offer. A first-order code or first-box deal is the only discount that shows up among the long-runners here. Plan a sale's replacement before you launch it.
- Try a perk before a price cut. True Protein runs zero price cuts and leads 23.8% of its ads with a perk.
- Check your category. If nobody in your feed discounts, a discount is loud. If everybody does, it is wallpaper.
For the baseline, see how long Facebook ads actually run. For why age alone isn't proof, see how to tell if a competitor's Facebook ad is working.
FAQ
Do discount ads work on Facebook? Public data can't show a conversion lift, because Meta withholds results for commercial advertisers. It shows what advertisers keep paying for: across 861 live Meta ads measured on 7 September 2026, 9.5% stated a discount, and discount ads outlasted other ads at five advertisers and fell short at three. The long-running discounts were standing new-customer offers.
What percentage of Facebook ads include a discount? 9.5%: 82 of 861 live ads stated a price cut in their copy. Another 7.4% offered a perk such as free shipping, and 83.0% offered neither.
Do discounts cheapen a brand? There is evidence they can. A 2020 University of Chicago field experiment found discount ads drove more search but lower conversion among searchers than brand ads, and reduced brand preference. In our data, Dior ran 97 live ads and none stated a discount.
How big a discount do brands advertise on Facebook? Of the 32 ads stating a percentage, the median was 20% off. The 18 stating a dollar saving had a median of $36.15.
Check your category before you discount
Your competitors have already answered this in public. Paste any brand into SOCIALFUEL to see their live ads across Meta, Google and TikTok, long-runners first, with the hook, angle, offer and audience decoded. Free to start, no credit card.
Weighing up tools? We compare the options in best ad spy tools.
About the author
Joel Helou is the founder of SOCIALFUEL, a growth agency that manages paid social and search campaigns for ecommerce and DTC brands, and the creator of socialfuel.io. He has spent the past decade buying media for brands across apparel, supplements and consumer goods, and now builds the tooling his own team uses to decode why competitors' ads work. This study was run on data captured through that platform.