Popunder Ads for CPA Marketing: The 2026 Playbook

Popunder traffic is cheap, high-volume, and criminally underused on modern CPA offers. Here's the exact setup, targeting stack, bid strategy, and blacklist discipline that keeps popunder campaigns profitable in 2026.

Popunder ads are the format most CPA affiliates dismiss too early and then quietly go back to. They're cheap, they're global, and they still convert on the offers that matter — sweeps, utilities, dating, adult, iGaming, low-friction subscriptions. This is the playbook we walk publishers through when they ask how to run popunder ads on CPA offers on the [Swift Digital Ads network](/publishers).

What popunder traffic actually is in 2026

A popunder opens the advertiser's URL in a new tab that stays behind the user's active window. The user finishes what they're doing on the publisher page, then finds your page when they close or switch tabs. That delayed reveal is why popunders keep working: they're not intrusive at the moment of click, so ad blockers, browsers, and users tolerate them at scale.

The main global sources in 2026 are PropellerAds, Adcash, PopAds, RichAds, Clickadu, HilltopAds, TrafficStars, ExoClick, EroAdvertising, and TrafficJunky for adult. Between them they move well over 30 billion daily impressions, and every one of them will front-load junk zones on a new account. Filtering those junk zones is 70% of the job.

Setting up a popunder CPA campaign — the exact stack

1. Pick the offer. Start with a sweeps, utility, dating, or iGaming registration offer priced between $0.80 and $6.00. Anything below $0.80 CPA can't survive the CPM economics; anything above $6 needs a stronger intent source than popunders.

2. Pick one GEO and one device. Never run mixed. A US-desktop iGaming campaign is a completely different auction from US-mobile-Android — different zones, different CPMs, different conversion rate.

3. Build a prelander. This is non-negotiable. Direct-linking a CPA offer to raw popunder traffic converts at 20-30% of what a prelander does. Use a fast, mobile-first prelander: 1 above-the-fold hook, 1 CTA, no navigation, sub-1-second LCP.

4. Set up a tracker (Voluum, RedTrack, Bemob, Binom). You need zone-level, sub-source-level data or you can't blacklist properly.

5. Start with a low CPM and Smart CPM bidding on whichever network you pick. On PropellerAds and RichAds this is the built-in "CPA Goal" mode. On PopAds and Adcash you'll manually adjust the average bid every 2–3 hours during the learning window.

6. Cap frequency at 1 pop per user per 24h. More than that eats budget on the same low-intent user twice.

Targeting stack that survives

• GEO: single country, or a Tier-1/Tier-2 bundle only when payout is identical.
• Device + OS: always split. Mobile Android, mobile iOS, desktop Windows, desktop macOS are four different campaigns.
• Browser: on mobile CPA offers, split Chrome vs in-app-browser (Facebook, TikTok, Instagram). In-app conversion rates differ by 3–5x on the same offer.
• Connection type: WiFi vs carrier only matters on iGaming and subscription offers where a payment step exists.
• Time of day: run 24h to gather data, then trim to the top 6–10 hours per GEO after 3–5 days.

Bid strategy — the part everyone gets wrong

Popunder networks are second-price auctions. You don't need to bid the maximum you can afford; you need to bid enough to beat the second-highest bidder on the zones you want. Start at 70–80% of the network's suggested Tier-1 CPM, then let the CPA-goal or Smart CPM engine adjust.

For manual networks (PopAds, Adcash on manual mode) the rule is: bid, run 300–500 pops, check zone-level EPC. If zones with EPC above your target CPA are still winning enough impressions, hold. If they're getting starved, raise bid 10% and repeat.

Blacklist discipline (the actual moat)

Every popunder campaign lives or dies on the blacklist. The rule:

• After 500 pops on a zone with zero conversions on a sub-$3 offer → blacklist.
• After 1,000 pops with zero conversions on a $3–$10 offer → blacklist.
• After 2,500 pops with zero conversions on a $10+ offer → blacklist.
• A zone with 3+ conversions but sub-target ROI → cap bid on that zone, don't blacklist yet.
• A zone with high volume but 100% bot postback pattern (all registrations, no downstream event) → blacklist immediately.

Mature popunder campaigns typically blacklist 30–70% of the network's zones. That's not a sign the network is bad; it's the format working as designed.

Prelander design that lifts CVR

The prelander is the entire creative in popunders. What consistently outperforms across sweeps, dating, and iGaming in our data:

• One-question quiz or spinner for sweeps and iGaming — lifts CVR 2–4x vs static prelander.
• Chat-style prelander for dating — 1.5–2.5x lift.
• "Your device is at risk" scan prelander for utility/VPN — legal on most networks, still the highest-converting utility angle in 2026.
• Localized landing copy (real translated, not machine) — 20–40% lift on Tier 2/3.
• LCP under 1 second — every 500ms of extra load time costs you roughly 10% CVR on popunder traffic because the user has already switched away.

Common pitfalls that torch popunder budgets

1. Running new offers on a fresh network account. Networks front-load poor traffic to new accounts. Warm the account with a proven offer first.

2. Direct-linking the CPA offer. Kills CVR and gets your account flagged.

3. Mixed GEO/device/OS. Impossible to optimize; the winning slice gets averaged out by the losing slice.

4. No frequency cap. Same user hitting your offer 5 times a day = wasted budget and irritated users → advertiser complaints → offer paused.

5. No postback loop. If you can't see FTDs or funded events, you're optimizing on registrations, which any bot can fake.

6. Blacklisting too early. 100 pops isn't a signal on a $5 CPA offer. Give zones the volume the math above requires.

When popunder traffic beats push and native for CPA

• Low CPA payout ($0.80–$3) offers where push CPCs would eat all the margin.
• Tier 2/3 GEO campaigns where native inventory is thin.
• Global sweeps and utility offers that need volume more than intent.
• Adult and iGaming in regulated GEOs where push and native reject the creative.

Push wins on higher-intent verticals (finance, insurance, subscription upgrades) because the format demands attention. Native wins on content-driven angles. Popunders own the cheap-volume, low-friction end of the market — and that end is bigger than most affiliates realize.

Compliance and account safety

Every major network in 2026 enforces:

• No auto-download, no auto-redirect chains, no fake system alerts that impersonate the OS.
• Adult-only offers isolated to adult networks (ExoClick, TrafficJunky, EroAdvertising).
• No misleading "you've won" claims without a real sweepstakes offer behind them.
• Working unsubscribe on any collected email flow.

Break these and your account gets banned, not just the campaign. Popunder accounts are hard to reopen once flagged.

Running popunder traffic on Swift Digital Ads offers

We accept popunder traffic on most sweeps, utility, dating, adult, and iGaming offers on the [publisher network](/publishers). Every offer page lists allowed traffic sources up front so you don't waste creative time on offers that reject the format. High-volume popunder buyers get postback access, weekly payouts on low minimums, and dedicated managers who share converting zones across affiliates running the same offer.

If you're already running push or native on our CPA offers and haven't tested popunders, that's usually the fastest way to add 20–40% incremental volume at the same margin. Start with one offer, one GEO, one device, and hold the blacklist discipline above — the numbers work.

Related reading: [CPA vs CPL vs RevShare in 2026](/blog/cpa-vs-cpl-vs-revshare-2026), [Search Arbitrage for CPA Marketing](/blog/search-arbitrage-cpa-marketing), [Scaling iGaming with Search Arbitrage](/blog/scaling-igaming-with-search-arbitrage).