Push vs. popunder: which one should your budget go to first?
Both formats deliver volume. They do not deliver the same user. A practical way to choose before you split a budget.
The short answer
If the offer needs one strong hook and a fast decision, start with popunder. If it needs repetition, re-engagement or a softer nudge, start with push. Most buyers eventually run both, but they should not start with both.
What actually differs
Popunder owns the full page. The user finishes an article, closes a tab, and the landing page is already there. Attention is broad and shallow: one screen to make the offer obvious.
Push arrives on the lock screen or in the notification tray. It is smaller, easier to ignore and much cheaper per impression. It works when the message is specific and the audience has shown interest in the category.
- Popunder: sweepstakes, betting, utilities and direct-response offers with an instant promise.
- Push: re-engagement, iGaming, app installs and offers that benefit from repeated exposure.
The mistake we see most
Advertisers judge both formats on click-through rate. That number is almost meaningless alone. A popunder campaign at 0.4% CTR can beat a push campaign at 3% CTR if the downstream event is stronger.
Set the optimisation event before launch — first deposit, funded account, purchase, verified registration — and compare formats on cost per event, not cost per click.
A simple testing plan
Run one format at a time for the first week with the same offer and comparable GEOs. Keep creative constant so the format is the only variable. After seven days, move 70% of budget to the winner and use the remaining 30% to test the second format.
No elaborate matrix, no twelve-way split. Formats are a blunt instrument; creative and source quality decide the outcome.
More from the buying desk
The in-page push playbook for iOS-heavy markets
No permission prompt, no OS dependency, full reach. How to structure in-page push when classic push lists…
Read articleHow to read a postback: a tracking checklist for buyers
Most “traffic quality” problems are tracking problems. Here is what we verify before a campaign goes live.
Read article
A creative testing framework that survives small budgets
You do not need fifty variants. You need two variables, a clear event and enough volume to tell them apart.
Read article