We help publishers, apps, and content platforms increase yield with cleaner placement systems, sharper viewability, and reporting that operators can actually use.
Hong Kong based | Publisher-first optimization | POPENTECH
We map every placement, request path, and demand dependency to find where revenue is leaking before new code or layouts go live.
Floor rules, refresh logic, and demand sequencing are tuned against your traffic mix so RPM improves without destabilizing fill.
We turn noisy dashboards into weekly operator views with the few metrics that actually drive action: eCPM, fill, viewability, and session RPM.
We redesign slot structure for scroll depth, readability, and loading behavior so monetization feels built into the experience, not layered on top.
Above-the-fold pressure, sticky behavior, and lazy-load timing are rebalanced to improve attention without damaging page quality.
Each test ships with a hypothesis, guardrails, and a review cadence, so optimization becomes a repeatable operating loop instead of random adjustments.
One ad ops lead, one analyst, one review cadence. No handoffs between accounts, no rotating teams.
We pull four numbers every Monday -- RPM, fill, viewability, session length -- and only ship tests that improve at least two.
Every layout test ships with a bounce-rate and time-on-page guardrail. If engagement drops, the experiment rolls back automatically.
We tier demand partners by region, device, and connection type -- not by one flat global ranking.
Every change starts as a written hypothesis with expected impact. Results are logged so lessons accumulate across quarters.
Before weekly reports, we run anomaly checks. If RPM drops in a region overnight, you hear from us before the numbers hit the dashboard.
For two years we just accepted that weekends underperformed. Popentech mapped weekend demand to a separate floor schedule and fixed it in the first week of the audit.
They sat with our product team and rebuilt the interstitial cadence around level-completion timing. ARPDAU moved up, and our day-7 retention actually ticked slightly higher.
The difference is that they write down every test. Nothing disappears into a Slack thread. Four months in, we can actually trace why our eCPM behaves the way it does.
We rebuilt mediation three times with previous vendors. Popentech did a pass on waterfalls per region and kept the same partners -- revenue moved without a single SDK change.
The weekly operator view is the only report we send to the exec team now. Four numbers, one sentence, and a recommended next test. It stopped a lot of pointless meetings.
They warned us upfront that viewability would trade off slightly against session RPM. We chose the balance, they shipped it, and we reviewed the delta. No surprises.
Demand patterns by day-of-week are consistent enough to tune separate floors. Here's how we structure them, and what we use as guardrails...
Read NoteWhen rewarded placement is anchored to a natural in-game step rather than a timer, retention and ARPDAU can move in the same direction...
Read NoteA quick audit of which demand paths de-rate first, what we do as a first response, and the four reporting lines we watch during consent transitions...
Read NoteWe map every placement, waterfall path, and reporting gap in week one. Output is a ranked list of 6-10 adjustments with expected direction of impact.
We agree on engagement metrics that cannot regress, plus the RPM/fill thresholds that trigger an automatic rollback for any experiment.
Every change ships as a named hypothesis with a target review date. Layouts, floors, and refresh rules are tested in parallel on separate cohorts.
A one-screen summary of what moved last week, what ships this week, and one recommended next test. No 40-tab dashboards, no noise.
The audit takes about a week. If you already have engineering access to ad slots, the first floor/refresh adjustments usually ship the same week as the audit readout. Layout work depends on your release cycle.
Web on Google Ad Manager, mobile apps on AdMob / AppLovin MAX / Unity LevelPlay / ironSource, and custom mediation setups. If you run a custom stack, tell us what you use during the intro call and we'll flag whether it's a fit.
We don't. What we do promise is a written baseline, named hypotheses with review dates, and automatic rollback rules if any test hurts engagement. The direction of change has been consistent for past clients, but we won't quote a number we can't commit to.
Two models are most common: a flat monthly retainer for ongoing pod work, or a revenue-share arrangement on the incremental uplift above your agreed baseline. We'll recommend whichever fits your traffic size.
Before anything ships, we pick 2 engagement guardrails -- typically bounce rate and time-on-session. Every experiment is monitored against them. If either guardrail is hit, the variant rolls back on the next review day.
That's part of the audit. For publishers it's usually a native/display mix with one sticky unit. For games it's usually rewarded + interstitial cadence. We don't have a one-format recommendation; the starting point is your current inventory and retention profile.
Yes, that's the default path. We almost never ask clients to switch partners during the first 60 days. Most of the early lift comes from re-sequencing existing waterfalls, not ripping out integrations.
GDPR, CCPA, and ATT regions are treated as separate demand paths with explicit floor rules. All setups ship with a consent-handling checklist and separate reporting for opted-in vs. opted-out traffic.
One Monday email with four numbers (RPM, fill, viewability, session length), a one-sentence summary, and one recommended next test. Monthly reviews go deeper. The 40-tab dashboards stay in the background.
You own all accounts and contracts. We work inside your existing Ad Manager / MAX / LevelPlay accounts with limited access. If you need help creating new accounts, we walk you through that during onboarding.
POPULAR ENTERTAINMENT NETWORK TECHNOLOGY CO., LIMITED is a Hong Kong-based ad monetization studio. We work with publishers, app teams, and content platforms who want revenue work treated the same way they treat product work -- with written baselines, hypotheses, guardrails, and a review cadence that builds up knowledge instead of re-learning it every quarter.
Most agencies optimize for short-term spikes and impressive-looking case-study numbers. We optimize for stable, explainable lift that survives a traffic mix change, a new consent rule, or a platform policy update. If something about your revenue is still a mystery after three months, we consider that a gap in our work.
Headquartered in Mongkok, Hong Kong, with client teams across APAC. We work with publishers and apps globally -- the operating loop is the same regardless of where your traffic comes from. What changes is the regional demand sequencing, which is the whole point of the audit.
ROOM 602, 6/F, KAI YUE COMMERCIAL BUILDING, NO.2C ARGYLE STREET, MONGKOK, KOWLOON, HONG KONG
+1 (888) 462-3628
Mon - Fri: 9:00 AM - 6:00 PM (HKT)
24/7 emergency support for clients