25gamers.com

This is no BS gaming podcast 2.5 gamers – 🧠 CloudX just hit 10-20% Revenue Lifts in 90 Days: The First AI-Native Ad stack for publishers

Sharing actionable insights, dropping knowledge from our day-to-day User Acquisition, Game Design, and Ad monetization jobs. We are definitely not discussing the latest industry news, but having so much fun! Let’s not forget this is a 4 am conference discussion vibe, so let’s not take it too seriously.

The Team:

⁠Jakub Remia⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠r- Game design consultant⁠

Felix Braberg⁠ – Ad monetization consultant

⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Matej Lančarič – User Acquisition & Creatives consultant

Join our slack channel here: SLACK CHANNEL

Summary

The first AI-native programmatic stack for mobile publishers is real, and the early numbers are absurd. 10-20% revenue lifts across every publisher they’ve launched.

Felix Braberg sits down with Dan Sack β€” Co-founder of CloudX, formerly MoPub, Twitter, Max, and AppLovin β€” for round two of their conversation, three months after CloudX went GA in February. The company now works with dozens of publishers, has five bidders fully live, a dozen more integrating, and a 35-person team. The pitch is “monetization as code” β€” a publisher platform where every setup lives in code and can be operated by humans OR AI agents.

What’s actually working today is the infrastructure
layer: transparent fees, no competing ad network,
verified auctions, and integration options that include
first look, parallel, and standalone. The agentic
capabilities are coming in Q3 β€” and that’s the part
that could reshape how publishers operate.

πŸ“Œ KEY TAKEAWAYS

β€” CloudX is delivering consistent 10-20% revenue lifts
across every publisher launched, regardless of
integration type (parallel, first look, banners, full
screen). Dan admits even he didn’t expect lifts of
this magnitude this quickly.

β€” “Monetization as code” means your entire setup lives
in code and can be operated by humans OR AI agents.
Think Claude Code or Cowork but for ad operations β€”
agents that monitor, recommend, and execute on your
behalf.

β€” Three integration types: first look (CloudX gets the
call first with a floor, clears or passes), parallel
(both platforms auction simultaneously in silos,
winners compared), and standalone (CloudX called only
for specific apps/inventory). Standalone is launching
in the next few weeks.

β€” Banners are NOT dead. CloudX is seeing real lifts
there too β€” partly via dynamic price flooring (which
many SDK networks underprioritize) and partly by
bringing massive web/CTV banner demand into in-app
through partnerships with buyers who historically
ignored the channel.

β€” Demand stack today: Meta, Liftoff fully live. Next
wave (weeks away): Moloco, Pangle, Digital Turbine,
MobileFuse, Verve. Big omni-channel buyers set for
Q3. Twenty-plus bidders integrating.

β€” Multi-round auctioning is the killer feature. Bidding
dynamic flooring + waterfall optimization in a
single auction system, no hacks needed. Sophisticated
publishers have been hacking this for years; CloudX
productized it.

β€” The AI agentic vision: dashboard becomes a Claude-Code-style
conversation β€” “what’s happening, what should I know,
what’s the recommendation, do it.” Automation in Q3,
intelligence layer following. The platform you log
into shouldn’t be a dashboard you operate, it should
be an agent you direct.

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

πŸŽ™οΈ HOST
Felix Braberg β€” Ad Monetization consultant

🎀 GUEST
Dan Sack β€” Co-founder, CloudX (formerly MoPub, Twitter, Max,
AppLovin)
/ dansack

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

πŸ’¬ Reach out to Dan: dan@cloudx.io
πŸ’¬ Comment below: would you try a programmatic stack that runs on AI agents?
πŸ‘ Like the video if you learned something
πŸ”” Subscribe for more deep-dive analysis

#mobilegaming #admonetization #cloudx #mediation
#programmaticads #aiads #f2p

Chapters



00:00 Cold open β€” dating app publisher up 14% revenue
01:19 Round two intros + Dan’s torn ACL update
03:30 What CloudX has shipped since February GA
04:39 Monetization as code: human + AI agent operated
07:28 Felix’s example: “kill DT for IAP whale cohorts”
08:43 First look vs parallel vs standalone explained
14:27 The real reason CloudX delivers 10-20% lifts
24:14 Felix’s pickle: 3 weeks of manual tuning, no AI yet
32:19 The Q3 automation + intelligence layer roadmap