The best Facebook ad examples come with a measured result: a €10 price cut that took CPA from €170.34 to €93.51, a founder filming in their car that cut CPA 62%, and a static-only campaign spending $253k at a $32.98 cost per purchase.
Below are 20 Facebook ad examples from our database, each with the real number and who got it.
How we picked: every example is a Meta result in our archive, chosen to cover ecommerce, apps, SaaS, lead gen and info products across formats.
Numbers come from the practitioner's post or the agency's public case study; vendor-reported means a tool telling the story of its own feature.
Quick answer: what the best Facebook ad examples have in common
Most winners changed the message, not the targeting.
A new offer, angle or messenger (a founder, customer or creator) moved CPA more than audience tweaks.
The second pattern is native-looking creative: selfie videos, statics, fake tweets and B-roll with a voiceover beat studio production.
Summary table of every example
| # | Example | What they did | Result | Type |
|---|---|---|---|---|
| 1 | Ecommerce store | €10 price cut, same ads | CPA €170.34 → €93.51 | Ecommerce |
| 2 | Tired account | ~20 angles for 2 avatars | CPA $101.29 → $60.50 | Ecommerce |
| 3 | Static-only campaign at about $18k a day | Statics only at ~$18k/day | $32.98 CPP on $253k | Ecommerce |
| 4 | Video ad with a "makeup teacher" hook | Teacher-style video hook | CPA $47.13, 45.9% hook rate | Ecommerce (beauty) |
| 5 | B-roll mashup made from old footage | Old footage + new voiceover | CPA $58.61 on $107k | Ecommerce |
| 6 | Founder talking in their car | Unscripted founder selfie video | −62% CPA | Ecommerce |
| 7 | Hurom | UGC on health outcomes, no discount | −65% CPA, +300% ROAS | Ecommerce |
| 8 | Sebastian Cruz Couture | Founder's-story ads, prospecting | 2.6x ROAS, $310 CPA | Ecommerce |
| 9 | Topo Designs | Catalog ads + retargeting | Facebook ROAS 916%, CPA −67% | Ecommerce |
| 10 | Twillory | Bundle and review catalog overlays | ROAS 1.96 → 4.49 | Ecommerce |
| 11 | K Republiq | Consolidated to ASC, carousels, UGC | ROAS 1.62 → 3.1 | Ecommerce (beauty) |
| 12 | Magic Story | Micro-persona UGC + partnership ads | ~−60% cost per subscriber, +800% incremental revenue | Ecommerce |
| 13 | DIFF Eyewear | Influencer selfie testimonials as ads | 200,000+ purchases | Ecommerce |
| 14 | RUX | Pre-sale lead list for BFCM | ROAS 2.64x → 6.47x | Ecommerce |
| 15 | Mindfulness app | Pain-point angles (burnout, ADHD) | CPA €57.15 → €23.85 | App |
| 16 | ChatPDF | Weekly concepts, 3 variants per winner | ROAS +280%, CAC −45% | SaaS |
| 17 | Online course | Webinar sign-up ads | ~$2/lead on Meta | Info product |
| 18 | Ambrook | Product-first demos + comparison ad | CPQL −62% | Lead gen (B2B SaaS) |
| 19 | Tie | Native-looking statics | 5.1x pipeline ROAS | Lead gen (B2B SaaS) |
| 20 | Honeydew Care | Box + value-prop static | $5 CPL, CPL −28% | Lead gen (telehealth) |
The 20 Facebook ad examples
For every Meta test we have, browse the full results database for this platform.
1. Ecommerce store: a €10 price cut, nothing else changed
What they did: Dropped the product price by €10 and kept the same ads, audience and budget.
Why it works: The price is part of the ad. The same spend bought more orders.
Result: Cost per purchase fell from €170.34 to €93.51 and ROAS rose from 2.5x to 4.5x in 10 days, per @SamPiliero. Result page.
Copy this: Test the offer before making new ads.
2. Tired account: 20 new angles for 2 avatars
What they did: Launched about 20 new angles written for two buyer avatars into broad prospecting.
Why it works: Fatigue usually means the message is worn out, not the targeting.
Result: Cost per purchase dropped from $101.29 to $60.50 (−40.27%) month over month, per @SamPiliero. Result page.
Copy this: Write 10 reasons to buy for each of your two main buyer types.
3. Static-only campaign at about $18k a day
What they did: Ran a Meta campaign built only from static image ads.
Why it works: Statics are cheap, so you can test more messages each week.
Result: $253k spent at a $32.98 cost per purchase, per @runtraffic. Result page.
Copy this: Put your best claim on a clean static and give it real budget.
4. Video ad with a "makeup teacher" hook
What they did: Opened a direct-response video with a teacher-style hook.
Why it works: A lesson earns attention without looking like an ad.
Result: A $47.13 CPA and a 45.9% hook rate, per @alexgoughcooper. Result page.
Copy this: Write hooks where the speaker teaches something in the first line.
5. B-roll mashup made from old footage
What they did: Cut existing footage into a B-roll mashup with a new voiceover and demo. No new shoot.
Why it works: The voiceover sells, so any clean product footage works.
Result: A top spender in the account: $107k at a $58.61 CPA, per @alexgoughcooper. Result page.
Copy this: Recut old shoots under a fresh voiceover script.
6. Founder talking in their car
What they did: On an account spending £90k a month, the founder explained, unscripted, why they started the company.
Why it works: The founder was explaining, not selling, so it felt like a message instead of an ad.
Result: CPA dropped 62% versus the brand's polished UGC and studio ads, reported by MHI Media (Kamal Razzak). Result page.
Copy this: Film your founder on a phone answering "Why did you build this?"
7. Hurom: health-benefit creator videos instead of discounts
What they did: Replaced always-on-sale ads with creator videos about gut health, skin and energy.
Why it works: Discounts had taught buyers to wait.
Health outcomes gave them a reason to pay full price.
Result: CPA fell 65% and ROAS rose 300%, per inBeat. Result page.
Copy this: Brief creators on one personal outcome and run it at full price.
8. Sebastian Cruz Couture: founder's-story ads
What they did: Moved from pure remarketing to new-customer statics, product videos and founder's-story ads.
Why it works: High-ticket buyers need a reason to trust a small brand.
Result: 2.6x ROAS at a $310 CPA on a $900 average order, per MuteSix. Result page.
Copy this: Judge CPA against your order value, not cheap-product benchmarks.
9. Topo Designs: dynamic product ads plus retargeting
What they did: Widened audiences and added dynamic product ads and retargeting for visitors and engagers.
Why it works: 90% of budget had gone to narrow prospecting with no retargeting.
Result: Facebook ROAS hit 916% (dynamic product ads 3,920%) and CPA fell 67% while spend doubled, per Hawke Media. Result page.
Copy this: Set up catalog retargeting before adding prospecting spend.
10. Twillory: catalog templates with a bundle and a review
What they did: Tested a bundle-deal template and a testimonial template against the plain catalog.
Why it works: The overlay puts the offer or proof right on the image.
Result: Control ROAS 1.96 vs 4.49 (bundle) and 4.41 (review) over ten days at equal spend. Vendor-reported by Marpipe. Result page.
Copy this: Add one overlay to your catalog and test it against the plain feed.
11. K Republiq: one Advantage+ campaign, UGC and carousels
What they did: Merged three competing campaigns into one Advantage+ shopping campaign, with cost caps, HD carousels and a simpler landing page.
Why it works: Three campaigns had been bidding against each other.
Result: Weekly ROAS went from 1.62 to 3.1 and CPA from $40 to $27 in under a month, per Flighted. Result page.
Copy this: If several campaigns target the same people, merge them.
12. Magic Story: UGC for micro-personas
What they did: Tested Reels, UGC and partnership ads for micro-personas like grandparent gifters.
Why it works: Each persona has its own reason to buy.
Result: Incremental revenue grew 800%+ and cost per subscriber fell roughly 60% in 60 days as monthly spend went from $20k to $150k+ (Flighted, Q4). Result page.
Copy this: Brief one creator video per buyer type.
13. DIFF Eyewear: celebrity selfie testimonials
What they did: Ran celebrity influencers' iPhone testimonials as paid ads from the influencers' own pages.
Why it works: The ads looked like posts from someone the viewer already follows.
Result: 200,000+ purchases on Facebook and 30M+ people reached at profit, per Common Thread Collective (no CPA given). Result page.
Copy this: Ask creators for ad access to their handle, not just one post.
14. RUX: a Black Friday lead list built early
What they did: Collected about 600 leads with the promise of an exclusive deal, then targeted that list during the sale.
Why it works: Early hand-raisers were primed to buy on day one.
Result: BFCM ROAS rose from 2.64x (six-month average) to 6.47x and CPA fell from $120.62 to $61.65, per Pilothouse. Seasonal comparison. Result page.
Copy this: Run an early-access lead ad three weeks before a sale.
15. Mindfulness app: burnout and ADHD angles
What they did: Swapped "better version of yourself" for procrastination, ADHD and burnout angles in X-style posts, UGC and AI animation.
Why it works: A named pain beats a vague aspiration.
Result: Meta CPA fell from €57.15 to €23.85 on the burnout funnel, per AdQuantum's award entry (the baseline may predate the agency). Result page.
Copy this: Rewrite your top ad around the problem users describe.
16. ChatPDF: weekly USP-to-hook concept tests
What they did: Mapped each benefit to a hook, tested new concepts weekly and made three variants of each winner.
Why it works: Weekly concepts fight fatigue.
Result: Meta ROAS rose 280% year over year and CAC fell 45%, per Admiral Media. Result page.
Copy this: Ship one new concept and three variants every week.
17. Online course: webinar ads on Meta vs TikTok
What they did: Ran webinar sign-up ads on Meta and TikTok in a first campaign.
Why it works: Meta found registrants far cheaper.
Result: Leads cost about $2 on Meta vs about $9 on TikTok, per @SamPiliero. Results vary. Result page.
Copy this: Get a Meta cost-per-lead baseline before splitting budget.
18. Ambrook: product-first demos for B2B software
What they did: Moved to product-forward ads with faster hooks, added a "tax prep" angle and an Ambrook vs. QuickBooks ad.
Why it works: B2B buyers want to see the tool.
Result: Cost per qualified lead fell 62% while spend grew 181% month over month, per Flighted. Result page.
Copy this: Show your product screen in the first three seconds.
19. Tie: fake-tweet and Apple Notes statics
What they did: Tested native-looking statics (a fake tweet, an Apple Notes screenshot) plus UGC, optimizing for qualified demos.
Why it works: They read like content, not B2B banners.
Result: 5.1x pipeline ROAS, profitable in month one, per Flighted. Result page.
Copy this: Mock up your main claim as a tweet and test it against branded creative.
20. Honeydew Care: a value-prop static for telehealth
What they did: Rebuilt the account around Advantage+ and ran a "Box + Value Props" static.
Why it works: One image shows what you get and why.
Result: That static hit a $5 CPL in testing, and CPL fell 28% on Advantage+, per Flighted. Result page.
Copy this: Photograph what arrives and list three reasons on the image.
Patterns across the winners
Ecommerce: fix the offer and the message first
The price cut (#1) and Hurom's switch from discounts to health outcomes (#7) both moved CPA by 45-65% without a new audience. See more offer tests on Meta.
Real people beat production
The founder car video (#6), DIFF's celebrity selfies (#13) and Magic Story's creator ads (#12) all looked like posts, not ads.
Our founder-led ad results and UGC ad results show the same thing.
Statics still scale
Static images carried $253k (#3), a $5 CPL (#20) and a 5.1x pipeline ROAS in B2B (#19). Browse all static ad results.
Apps and SaaS: name the pain, ship concepts weekly
The mindfulness app (#15) won with specific pains, and ChatPDF (#16) with a weekly testing rhythm.
For video, the first seconds decide it: track hook rate as well as CPA, and see our hook tests.
Lead gen: show the product, optimize for quality
Ambrook (#18) and Tie (#19) both moved from generic leads to qualified leads and showed the product early.
You can find more ad examples across TikTok, Google and YouTube.
Facebook ad examples that lost (what not to copy)
These are the versions that got beaten.
- Spending in countries that never buy. A UK merch seller's Rest-of-World campaign spent £5,891 at a 2.03 ROAS, but more than £700 of it went to countries that produced no sales (@SamPiliero). Check the country breakdown.
- Native creator video that cost more than an AI dub. In Inshallah's 10-day test in the same ad set, a native French UGC video got installs at €2.52, while an English UGC video translated with AI got €2.14, 15% cheaper (Admiral Media).
- Clothes on models. Taylor Stitch found that on-model images lost to "environmental lay flats", which had an 18% lower CPA in multivariate tests (vendor-reported by Marpipe).
- Click-to-Messenger ads. Qmomo's usual click-to-Messenger ads lost a conversion lift test to a leads campaign with automated coupon codes in Messenger, which had a 56% lower cost per online conversion (vendor-reported by Meta).
How to test these Facebook ad ideas in your own account
- Pick one variable. Offer, angle, messenger or format. Copy one example at a time.
- Keep everything else fixed. Same budget, audience and landing page, like the €10 price test.
- Give it enough spend. Run it until each version has enough purchases or leads to compare, not one good day. Here is how to set up a creative test.
- Judge on the metric that pays you. CPA or ROAS for ecommerce, qualified-lead cost for B2B, not CTR alone.
- Feed the loop. Look at what competitors are running in the Facebook Ad Library or with ad spy tools and swipe libraries, then write your own versions.
Foreplay and Atria save ads into a swipe file and turn them into briefs (we earn a commission if you sign up through our links).
They pay off if you make new creative every week; if you launch a few ads a month, the free Ad Library is enough.
If you would rather have someone run these tests for you, compare Facebook ads agencies or get matched with a media buyer.