The numbers are in, and they are not moving in a friendly direction.
iOS CPI hit $5.84 globally in Q1 2026, up 19% year over year. Android climbed 8% to $1.92. Finance apps are running CPIs around $8.70. Competitive gaming categories in North America are seeing CPIs between $4.50 and $6.00, with strategy and RPG titles up 18 to 25% year over year. Total global mobile app install spend reached $94 billion in 2026, up from $81 billion in 2025.
If your UA budget stayed flat this year, you are effectively running a smaller program than twelve months ago. And if your budget grew but your results did not keep pace, the math above is probably why.
The structural forces behind CPI inflation are worth understanding because they change how you respond.
More advertisers are competing for the same inventory. Every category that matures brings more funded apps chasing the same users, which drives up auction prices across every channel. Gaming is the most obvious example, but fintech, health, and subscription apps have all seen the same pattern as those categories scale.
iOS attribution is still broken, and it is making things worse. With ATT opt-in stalled at around 27%, more than 70% of iOS attribution decisions in 2026 are running on probabilistic or SKAdNetwork data rather than deterministic signals. When advertisers cannot measure with confidence, they overbid to compensate. That overbidding inflates CPIs for everyone in the auction, including the teams that have invested in better measurement.
Creative fatigue is accelerating spend waste. The median creative refresh cycle dropped from nine days in 2025 to three days in 2026. Teams running stale creative are getting lower performance and paying more for it, because fatigued creative drives up effective CPI without anyone on the team necessarily connecting the two.
Here is where it gets more uncomfortable.
iOS CPI rose 19% year over year. iOS ARPU grew only 7% in the same period. That gap is compressing LTV-to-CPI margins, particularly in mid-tier categories like utilities, casual gaming, and photography apps. Users are getting more expensive to acquire, while the revenue they generate is not keeping pace.
For teams still optimizing primarily toward install volume, this is a quiet crisis. The CPI line goes up, ROAS numbers soften, and the instinct is to optimize harder toward cost. That optimization usually produces cheaper installs that convert and retain worse, which compounds the LTV problem rather than solving it.
The teams navigating this well have shifted their orientation entirely toward user quality. That shift changes almost everything about how a UA program runs, from the channels prioritized to the creative built to the metrics used to evaluate success.
A $1.00 install from a user who opens the app once and never returns produces nothing. A $6.00 install from a user who subscribes, comes back daily, and refers three friends produces a business.
The more useful question is which users are actually worth acquiring at current prices and how to build a program around finding more of them.
Subscription apps post Day-30 retention near 14%, more than 2.5 times the cross-category mean. That retention premium is what justifies higher acquisition costs in that category. The teams that understand their own retention and monetization curves at the cohort level are making better bidding decisions than the teams chasing low CPIs across the board.
This is also where creative becomes a direct financial variable.
High-performing creative assets can reduce effective acquisition costs by 30 to 45% compared to generic approaches. On a $500,000 quarterly UA budget, the difference between strong creative and average creative is potentially $150,000 to $225,000 in wasted spend or recaptured efficiency. Teams treating creative as a fixed cost are leaving that on the table every quarter.
Given the structural reality of rising CPIs, the teams finding ways to grow profitably are working a few specific levers.
Creative quality and refresh rate. With the median creative lifecycle at three days, programs shipping diverse, well-tested creative consistently are getting meaningfully more out of the same media spend. Systematic testing across hooks, formats, and emotional frames produces directional learning that compounds over time. Teams producing volume without structure spend more to stand still.
Cohort-level measurement. Understanding which user segments pay back fastest, retain longest, and generate the most downstream revenue produces better bidding decisions. That insight does not come from platform-reported ROAS. It comes from understanding your own data at a level most teams have not invested in building.
Channel diversification. When iOS CPIs are running 3x Android in the same category, teams with disciplined channel mixes find real efficiency advantages. CTV, rewarded UA, and retargeting all play different roles in a well-structured program, and the interaction effects between them can meaningfully improve blended acquisition economics.
Retention as a UA input. Programs with the strongest payback metrics can afford to bid more aggressively because their downstream numbers justify it. Retention and monetization improvements are UA budget multipliers, full stop.
CPI inflation is a structural feature of the current market. The teams building programs around user quality, creative performance, and cohort-level economics are compounding their advantage while everyone else fights over the same inventory at higher prices.
If you want to talk through where your current program sits and what the real levers are for improving acquisition economics, the Work Dog team works through exactly these kinds of problems with clients every day. Reach out, and let’s get into it.
