Profile on LinkedIn. Editorial correspondence goes to the editorial inbox.
Who I am
I have spent nine years buying performance traffic, most of it on formats other buyers treat as a last resort. I came to pop from the direct-response side, where the brief never changed: find volume that converts at a price the offer can carry, and find it before the competition does. Pop turned out to be the cheapest place to learn, because a few hundred dollars produces enough data in a day to tell you whether a GEO and device combination deserves another week. I work with iGaming, dating, sweepstakes and utilities offers, which are the verticals where pop still earns its keep in the United States. I also read every platform from the publisher side, because payout thresholds, hold periods and site-acceptance rules decide whether a monetization plan survives contact with reality, and those terms are documented far less consistently than deposit floors.
What I actually know
Pop mechanics at the browser level. I can tell you why a pop fires on one site and not another, why the same tag behaves differently on mobile Safari than on desktop Chrome, and why transient activation expiring mid-session quietly costs a publisher volume. None of that is academic. A publisher whose pop rate collapses after a browser update usually has a gesture problem, and treating it as a demand problem wastes a week.
Auction routing. I read bid requests as OpenRTB objects instead of treating the auction as a black box, and that is what separates direct inventory from resold inventory in practice. My working test is source ID stability across a week: placements that hold their identifiers are real sites I can whitelist, identifiers that churn are a feed rotating underneath me.
Fraud and filtering. Pop is the cheapest inventory in performance advertising to fabricate, because a bot needs no dwell time and no click for the impression to register. I know which filters a buyer can switch on without asking a manager, which ones cost legitimate volume when enabled, and why proxy and IPv6 exclusion belong in a mid-flight test against a clean baseline, not in the settings on day one.
Publisher economics. I track deposit floors and payout floors as two separate numbers, because a platform generous on the buying side is often the slowest to pay. Thresholds between $5 and $50, hold periods from NET7 to Net30, and per-method minimums that jump to $200 on crypto are the details deciding whether a small site can use a platform at all.
Pricing units. Floors, clearing prices and publisher payouts get quoted interchangeably across this industry, and a rate stated per view sits a thousandfold below the same rate stated per thousand views. Most of what I untangle for people turns out to be unit confusion, not a genuine disagreement about price.
How I work
Everything begins in the documentation.
Before a figure reaches a page I have opened the platform's terms, help center or FAQ, read the clause it comes from and recorded the date. When a document contradicts what the comparison lists say, I quote the document and state plainly that the two disagree. When a platform publishes nothing on a point, I write that down as a gap and leave it there, which is why several sentences on the main page exist purely to record an absence. The alternative would be to borrow a number from a directory and present it as research, and that is the one shortcut this project cannot take.
My campaign work shapes what I think is worth checking. It does not license me to publish numbers I never measured.
Frequency caps, source ID behaviour, dayparting and postback setup are things I handle weekly, and the sections covering them are specific for that reason.
What I do not claim
I have not run a controlled side-by-side spend across all six platforms covered on this site, so nothing published here is presented as my own measured result. I do not reconstruct payout times, conversion rates or CPM figures from memory, and I do not describe a test I did not run. Where market benchmarks appear, the label says aggregated market figures, because that is what they are and no platform quotes them as its own rates. I also hold no professional licence in finance or law, and nothing I write is advice of that kind. The limits of what these pages are for are set out in the terms of use.
Corrections and contact
If you have a document that contradicts something I published, that is the message I most want to receive, and it is welcome from the platforms themselves.
Send the page address and the source to the editorial inbox; other routes are listed on the contact page. The standards I hold myself to are written out in the editorial policy, the background to the project is on the about page, and the current worked example is Popunder Ad Networks. What happens to your message once it reaches me is described in the privacy policy: it is used to answer you and to record what was corrected, and nothing else.
Content last verified 5 August 2026.