The starting point
Scaling installs without understanding revenue accelerates the problem.
CPI is useful for comparing channels and creative, but it does not tell you whether the right user arrived. An app can lower cost per install and still make the business worse if trial rate falls, paid conversion drops or churn consumes the new MRR.
The goal is not to buy the largest number of installs. It is to acquire users who activate, pay and stay long enough to recover CAC.
One connected view
The funnel must reach MRR, not stop at the app store.
Impression and click
Measure the ability of the creative and message to attract the right person.
Install and activation
Show whether the ad’s promise remains true in the store and during first use.
Trial and subscription
Reveal whether the product demonstrates value before asking for payment.
Retention and expansion
Define how much each cohort can afford in acquisition without breaking the economics.
Decision dashboard
The metrics need to speak to one another.
The minimum set
- Impression-to-install conversion by creative and channel.
- Activation, install-to-trial and trial-to-paid by cohort.
- D7, D30 and D90 retention, plus subscriber churn.
- CAC per paid subscriber and the months required to recover the investment.
- LTV by acquisition source, operating system and market.
Growth discipline
Media cannot fix onboarding that fails to deliver value.
The routine combines creative, store page, product events, onboarding and retention communication. If a person arrives with one expectation and finds another experience, the platform may still record a cheap conversion. The business records a loss.
Before increasing the budget
- Validate the events connecting campaign, install, trial and subscription.
- Separate iOS and Android when costs and behavior differ.
- Compare mature cohorts instead of judging retention after only a few days of use.
- Set an acceptable payback threshold for cash flow and margin.
Operational evidence
Three different product stages.
MRR grew 20-fold in six years, more than 80% of leads came from inbound and LTV/CAC stayed between 15 and 17.
From three founders to an operation of 35 to 40 people, ending in an acquisition by an investment fund.
More than one thousand installs in the first month for a new augmented-reality try-on category.
Common questions
What usually needs to be answered before scaling.
Is a low CPI a good sign?
Only when acquired cohorts also activate, pay and stay. A cheap CPI with low conversion and high churn is still expensive.
Is it too early for paid media?
If the product has not validated activation, retention or monetization, ad spend can amplify uncertainty. First, measure the core behavior.
Should iOS and Android stay together?
Not always. Costs, conversion and LTV can vary significantly. Separation makes sense when it changes the budget decision.
When can we increase spend?
When attribution is reliable, the funnel sustains conversion and payback remains within the operation’s limit.