If your Japan UA dashboard is telling you installs are up, you might be celebrating the wrong number.
Across the hypercasual genre, campaigns that generate 20% of all installs in Japan are producing just 1% of all sessions in the same market. In other words: users are downloading the game, opening it once (or not at all), and disappearing. You paid for the install. Japan didn’t pay you back with a player.
This gap — between “install” and “actually played” — is the single most misunderstood dynamic in Japan mobile UA right now. Here’s what the data shows, and why it changes how UA should be bought in this market.
Contents
The Data: Installs Happened. Nobody Played.
Pull genre-level data from adjoe, Liftoff, and PocketGamer’s 2025 reporting, and the hypercasual picture in Japan looks like this:
- Installs: 20% of all installs in Japan
- Sessions: 1% of all sessions in Japan
That’s not a soft engagement dip — it’s a near-total disconnect between acquisition volume and actual play. A campaign can look like a top performer on a CPI dashboard and still be delivering almost no real users to the game.
The CPI Trap: More Spend, Less Engagement
It would be one thing if this install-session gap were at least getting cheaper to accept. Unfortunately, it is not true for most cases.
According to Adjust’s Mobile App Trends 2026 report, gaming CPI is up 30% year-over-year. Cost to acquire is climbing. Actual engagement is falling. And in Japan specifically, that gap between spend and engagement is wider than in almost any other market.
The math doesn’t work in UA’s favor here: installs without sessions generate zero ad revenue. Teams optimizing purely for CPI are paying more, for less, in the one market where the gap matters most.
Japanese Users Are Different: They Go Deep, or They Don’t Go at All
This isn’t a story about Japanese users being harder to engage in general — it’s about how they engage once they commit.
Average session length by genre in Japan:
- Action: 43.8 minutes — the highest average session length globally
- RPG: 40 minutes — story-driven, long-term committed players
- Strategy: 31 minutes — deep engagement, high IAP potential
- Puzzle: 1% of the entire installs, but it constitues 21% of total sessions in Japan
That last status shows that puzzle barely registers on the install side, yet it commands over a fifth of all session volume in the market. Japanese users aren’t shallow — they’re selective. When they commit, they go deep. When a game doesn’t earn that commitment, they simply don’t open it a second time.
Install vs. Session: The Question That Changes Everything
Most UA strategy in Japan is still built around one question: “How many users installed my game?” That question optimizes for cost per install — and, as the data above shows, it misses Japan almost entirely.
The right question is different: “How many users actually played?”
That framing optimizes for real engagement, and it aligns with how monetization actually works in this market — through session depth and sustained play, not download counts.
Why CPE (Reward) Wins in Japan
If installs aren’t the right thing to pay for, what is? Rocket A’s answer is CPE — Cost Per Engagement, structured as reward-based campaigns.
- You pay only when users engage. CPE (Reward) campaigns charge based on a defined in-game action being completed — not simply on install.
- Reward-driven users go deeper. Users who opt in for a reward arrive with higher intent. They play longer, and they convert at higher rates.
- It’s native to Japan’s Poikatsu culture. Roughly 81.3% of Japanese consumers practice reward-based (poikatsu) engagement daily. This isn’t a behavior UA teams need to teach Japanese users — it’s already part of how they use their phones.
- Direct media access means more reward reaches the user. Without an intermediary margin sitting between budget and user, more of the campaign’s spend goes toward the incentive itself — driving stronger participation.
Conclusion: Stop Paying for Installs Nobody Plays
The Japan mobile UA data is consistent across every angle in this report: installs are cheap to generate and increasingly expensive to buy, while the users who actually matter — the ones generating sessions, IAP, and ad revenue — are being systematically missed by install-first campaign structures.
- Install-first UA optimizes for a number that, in Japan, correlates poorly with real usage.
- Engagement-first UA (CPE/Reward) pays for the behavior that actually drives revenue, and maps naturally onto an existing local consumer habit.
For teams running UA in Japan, this is a good moment to ask which question your current campaigns are actually optimized to answer.
Interested in running a CPE (Reward) campaign in Japan?
Visit Rocket A 👉
