Popunder vs Meta Ads: when impulse beats audience, and when it doesn't

My name is Marco. I worked at PropellerAds from 2018 to 2023, running tier-1 iGaming, dating, and finance. I am writing this comparison because the question gets asked in the wrong way every time, and the dishonest answer is uncomfortable so the industry has settled on the dishonest one. Let me give you the honest one.

Popunder is wrong for considered-purchase offers. Meta Ads is wrong for B2B and for a long list of regulated verticals. Where they actually compete is the middle layer — B2C impulse offers, lifestyle products, sweepstakes, dating, deposit-bonus iGaming where local rules allow it on social. Most of that middle layer goes one way more often than the industry pretends.

The opinion this post is built around

Popunder converts on impulse-friction offers where Meta doesn’t, because Meta’s audience targeting is excellent at finding the right user and bad at catching that user in a state of low-friction click. Popunder is the inverse: the targeting is crude (country, vertical, time-of-day, device) and the friction is near zero (the user sees your landing page in a fresh tab, no swipe required). For offers where the conversion mechanic is “user sees a price low enough to click on impulse,” popunder wins on cost-per-converted-user. For offers where the conversion mechanic is “user sees creative that resembles social content from someone they trust,” Meta wins.

The industry won’t say this clearly because Meta sells the second story (audience graph, look-alikes, AI-driven optimisation) and popunder networks sell a vaguer version of the first (volume, low CPM, scale). Both stories are partial truths.

Mechanics, briefly

A popunder opens a new browser tab or window behind the user. They see it when they switch tabs or close the original. The unit is a full landing page. No creative compete inside the slot. No swipe. No “view 3 seconds to count as an impression.” The viewability rate is effectively 100% because the user has to interact with the tab to dismiss it.

A Meta ad runs in feed, stories, reels, marketplace, or the audience network. The unit is a 1080x1350 image or 15-second video competing with the post that came before it and the post that comes after. Viewability is high inside feed and lower inside reels. Click rate depends on creative quality and audience temperature.

The two formats target different parts of the user’s attention. Popunder takes a user who is already on a publisher site and shows them a full landing page in a clean context. Meta takes a user who is scrolling social content and asks them to interrupt that scroll. Same human, different cognitive state.

CPM math, head to head

The cost gap is real. The conversion-rate gap is also real. The question is which gap is bigger for your specific offer.

Where popunder wins

Regulated verticals where Meta doesn’t accept the ad. iGaming in Brazil pre-2025 was unrunnable on Meta. iGaming in Italy is whitelisted only for ADM-licensed operators — if you are an affiliate sending traffic to an operator, you cannot run Meta. Sweepstakes is forbidden across most of Meta’s enforcement zones. CBD, gambling-adjacent, supplement, anything in a gray vertical — Meta will accept it on Monday and ban your account on Friday.

Popunder accepts these verticals. Adsterra, PropellerAds, RichAds, Mondiad, Adsy.tech — all of them. The compliance bar is lower because popunder publishers serve traffic they can’t moderate inside a feed; the offer is on your landing page, not on theirs.

Impulse-friction offers where the conversion happens in under 30 seconds. Push-button utility installs. Deposit-bonus iGaming with a no-deposit free spin. Dating offers with a free first message. The user sees the offer in a fresh tab, decides in three seconds, clicks. Popunder beats Meta on cost because Meta is paying for the audience-graph match that doesn’t matter once the user has clicked.

LATAM dating offers in 2022 ran at $0.80 CPM popunder against $7 CPM Meta in the same GEO. Conversion rate gap was 0.6% (popunder) against 1.4% (Meta). Cost per registration: popunder $13, Meta $50. Meta’s matched users were better but not 4x better.

Where Meta wins

B2C lifestyle offers that need creative storytelling. Skincare. Fashion. Specialty food. Anything where the user needs to see a beautiful image of the product in a social context before they want it. Meta’s creative format and feed placement do this. Popunder can’t. A popunder serves a landing page; it doesn’t serve a 6-second reel of someone applying the moisturiser. The format mismatch is structural.

Demographic-narrow B2C. If your offer converts well for women aged 28–42 in Lombardy who are interested in cycling, Meta finds them. Popunder targets country + vertical + device. The narrower your audience, the more Meta’s audience graph is worth.

Lookalike scaling. Once you have a converted-user list of 5,000+, Meta’s lookalike algorithm finds adjacent users who behave similarly. Popunder has no equivalent. Network-side AI optimisation exists (PropellerAds called theirs SmartCPM, Adsterra has CPA goal bidding, Adsy.tech has rule-based optimisation), but it is publisher-side optimisation, not audience-side. The closest popunder analogue to a Meta lookalike is “broad targeting, optimise on conversions, let the network find the publishers that work,” which is not the same thing.

What changed in 2021 and is still changing

iOS 14.5 launched App Tracking Transparency in April 2021. Meta’s targeting accuracy dropped because they could no longer attribute conversions cleanly through the iOS app graph. Meta moved to Aggregated Event Measurement and the SKAdNetwork API. Reported cost-per-acquisition on iOS Meta campaigns rose 20–40% in late 2021 and stayed elevated.

Popunder was unaffected. The format runs on browser tabs, not on app graphs. The popunder-vs-Meta cost gap widened slightly in 2022–2023 because Meta got more expensive, not because popunder got cheaper.

GDPR enforcement has been the inverse story. Popunder networks doing third-party-cookie attribution had to rebuild around server-side tracking. Adsy.tech’s RTB is in-house, which simplifies things on their side; Adsterra and PropellerAds rebuilt their conversion APIs in 2022–2024. Meta operates under GDPR but has the engineering team to absorb the cost of compliance. Smaller popunder networks struggled more. Anyone telling you GDPR didn’t move popunder economics in 2022 is either new to the format or selling something.

The B2B trap

Half of the “popunder for B2B” pages I have read in the last two years are written by people who do not run B2B campaigns. Popunder is wrong for B2B. Meta Ads is also wrong for most B2B. The right format for B2B is Google Search (intent capture), LinkedIn (audience precision), and increasingly content distribution platforms with intent signals. If you are reading this because you have a B2B SaaS at $499/month and want to know which of these two is right, the answer is neither.

The reason popunder is wrong for B2B is the same reason it is wrong for any considered-purchase offer. Pop a unit out of the user’s tab and you remove the publisher context. For impulse-friction offers, that is the conversion mechanic. For B2B offers that need three case studies, social proof, a free trial sign-up flow, and a sales conversation, you need to be inside the user’s context, not on a fresh tab.

The reason Meta is wrong for most B2B is that audience graphs are weak at job-title accuracy. Meta knows you are 38, in Milan, interested in cycling and adtech. Meta does not reliably know you are a Director of Marketing at a 200-person SaaS company with iGaming clients. LinkedIn knows that. Use LinkedIn.

What the panel doesn’t tell you

Both formats have hidden CPM-to-actual gaps. Popunder’s gap is auction mechanics — the panel often shows the bid ceiling, not the clearing price. Meta’s gap is bid optimisation — you set a target CPA and Meta auctions against itself across your audience until it finds the clearing CPM that delivers that CPA, which is usually different from the CPM in the ads-manager dashboard.

Both formats also have tracking-to-CRM reconciliation gaps. On popunder, that gap shows up as fake conversions from publishers serving fraudulent traffic. On Meta, it shows up as iOS attribution loss and AEM aggregation noise. Both gaps are typically 15–40%. Both require server-side conversion validation if you want a reliable cost-per-acquisition number.

In Q2 2022 I watched a dating campaign convert at 12% in Voluum and 0% in the advertiser’s CRM. The publisher was firing the conversion pixel from a server farm in Vietnam before the user actually saw the offer. Six weeks of optimisation, six weeks of advertiser questioning my judgment, before we caught it. The lesson is general: tracking platforms see HTTP events, not humans. Your CRM sees humans. If you don’t have a reconciliation job between them, you don’t have tracking, you have estimates.

When you should run both

If you sell a B2C impulse offer with a free trial in a market where Meta accepts it, you should run both. Meta captures the targeted audience at higher cost. Popunder captures cold volume at lower cost. They serve different layers of the funnel and they don’t cannibalise.

The mix depends on the offer’s LTV math. High-LTV impulse offers (iGaming, lifestyle subscription at $20+/month) tolerate Meta’s higher cost because the unit economics work. Low-LTV impulse offers (utility installs, one-shot affiliate offers under $10 commission) usually can’t afford Meta and have to run popunder to find scale.

When you should run neither

If your offer needs five seconds of evaluation before the click, popunder is wrong. If your offer is B2B, Meta is wrong. If both are true (B2B SaaS with a considered purchase), run Google Search + LinkedIn + content distribution and skip both of these.

What to do next

If you want a popunder test for a B2C impulse offer, $50 on Adsy.tech (because of the $0.50 CPM minimum on the rate card) is the cheapest way to find out whether your offer fits the format. If the conversion rate is in the range I gave for your vertical, scale up. If it isn’t — either your offer isn’t impulse-friction, or your landing page is slow, or the publishers the network gave you are bot-adjacent. The order I troubleshoot in those three is the reverse: check publisher quality first, landing page second, offer fit last.

If your offer is B2B or considered-purchase, don’t waste the $50 on popunder. Spend it on Google Search with tight keyword match.

See my honest 2026 popunder network ranking, the format glossary, or if you meant to compare against Google instead of Meta, the Google Ads version of this post.

FAQ

Is popunder cheaper than Meta Ads?

On CPM, almost always — popunder runs at $0.40–$12 depending on vertical and GEO, Meta runs at $4–$22. On cost-per-converted-user, it depends on offer fit. Impulse-friction offers usually convert cheaper on popunder. Lifestyle and demographic-narrow B2C usually convert cheaper on Meta.

Why does Meta keep banning my account?

Compliance enforcement. Meta operates a strict policy on iGaming, sweepstakes, dating outside whitelisted markets, supplements, financial offers, CBD, and a long list of gray verticals. Popunder networks accept these verticals because publishers serve traffic without curating ad content the way Meta curates feeds. If your offer is in a gray vertical, popunder is the structural fit.

Does popunder traffic work for e-commerce?

For price-driven impulse e-commerce (deal sites, daily-deal listings, sub-$50 AOV with strong offer copy), yes. For brand-driven creative-led e-commerce (boutique fashion, specialty food, anything where the photo is the conversion mechanic), no. The format mismatch is structural: popunder serves a landing page, not a creative.

Should I run popunder if I have a strong Meta funnel?

Probably yes. Meta gives you targeted audience volume. Popunder gives you cold volume at lower cost. They serve different funnel layers. The mix depends on offer LTV. Test for $50 and check whether the cost-per-converted-user math works for your specific unit economics.

Sources. Popunder CPM ranges from Marco DeLuca’s PropellerAds book 2018–2023 plus buyer-side testing 2024–2026. Meta CPM ranges from Meta’s published Marketing API reports, third-party benchmark studies (Hootsuite 2025 social benchmark report), and Marco’s first-party Italian iGaming campaign data Q3 2023. EDPB summary of EU popunder ePrivacy guidance, EDPB 2024.