CPI, CPA, and ROAS answer different questions. A campaign can buy inexpensive installs and still lose money, or start with a high CPI and become valuable when players return and spend. The useful approach is to connect acquisition cost with player quality and revenue.
By Dorn Bros Entertainment · How we prepare our guides
What CPI tells you
Cost per install divides advertising spend by attributed installs. It is a fast acquisition benchmark, but it does not describe retention, purchase behavior, or advertising revenue after installation.
- Use one attribution window consistently.
- Compare like-for-like countries, platforms, and campaign goals.
- Separate paid installs from organic installs when evaluating ad spend.
How CPA differs from CPI
Cost per action measures a chosen action after the click or install, such as registration, completing a tutorial, or making a purchase. Define the action before comparing campaigns so the number keeps the same meaning.
- Choose an action tied to real game value.
- Track enough volume before making large changes.
- Check whether low-cost actions lead to retained players.
Using ROAS with retention
Return on ad spend compares attributed revenue with advertising spend. Early ROAS can be incomplete because players may generate revenue over time. Review it alongside day-one, day-seven, and later retention when those measurements are available.
- Use the same revenue window for every campaign in a comparison.
- Include both in-app purchase and ad revenue when your reporting supports it.
- Scale gradually when results are stable rather than reacting to one day.
Worked acquisition example
Consider an illustrative cohort with $120 in advertising spend and 240 attributed installs. CPI is $0.50. By day seven, the cohort has produced $72 in advertising revenue and $30 in purchase revenue, for $102 total attributed revenue. Day-seven ROAS is therefore 85%, leaving an $18 acquisition-cost gap before development, support, and other business expenses.
The example does not prove that the campaign will lose money. Later cohort revenue could close the gap, but that decision should be based on measured retention and later revenue rather than an assumption. Compare the same cohort again at a consistent later window before increasing spend.
- CPI: $120 ÷ 240 = $0.50.
- Day-seven revenue: $72 + $30 = $102.
- Day-seven ROAS: $102 ÷ $120 × 100 = 85%.
A practical review order
Start with tracking quality, then confirm spend and install volume, compare CPI, review the chosen post-install action, and finally connect cohort revenue to ROAS. This sequence prevents a cheap install from being mistaken for a profitable player.
Use the CPI Calculator
Run your own numbers with a free browser-based tool, then explore a Dorn Bros release.
Frequently asked questions
Is a lower CPI always better?
No. A lower CPI is useful only when the acquired players also engage, return, or generate enough value.
What is a good mobile game ROAS?
There is no universal target. It depends on margin, revenue timing, country mix, platform, and how long you can wait to recover acquisition cost.
How often should campaign metrics be reviewed?
Review tracking daily, but make strategic decisions on enough conversion and cohort data to avoid reacting to normal short-term variation.