How Long Does a New Android App Take to Earn Ad Revenue?
Every monetization guide tells you which SDK to integrate. Almost none tells you that the first quarter pays approximately nothing — and that this is arithmetic, not a mistake you made.
Direct answer: budget for a quarter of near-zero. Ad revenue is impressions multiplied by eCPM, divided by 1000 — so an app with no installs earns nothing regardless of which network is integrated. Installs accumulate first, sessions accumulate on top of installs, and mediation needs weeks of that data before it prices inventory properly. Months three to six is the honest window.
The formula is the whole story
Everything below is a consequence of one line:
Ad revenue = (impressions × eCPM) ÷ 1000
And impressions are themselves a product: daily active users × sessions per user × eligible ad requests per session × fill rate. Four multipliers, each of which starts near zero on a brand-new listing. Multiply four small numbers together and you get a very small number. That is not a monetization failure — it is the definition of a launch.
This is also why the most common launch question is the wrong one. "Which ad network pays best?" adjusts a single term, eCPM, by some percentage. "How do I get installs and sessions?" adjusts three terms by multiples. In month one the second question is worth roughly an order of magnitude more than the first.
The four gates, in the order you hit them
Gate 1 — Install accumulation (weeks 1–6)
A new listing enters Google Play with no ranking history, no ratings and no review signal. Organic discovery is slow by design, because Play weights installs and retention that you have not earned yet. During this gate your impression count is dominated by the handful of people who found the app, which in practice means you are looking at daily earnings measured in pennies. Nothing about the ad stack can fix this, because there is no inventory to sell.
Gate 2 — Session depth (weeks 3–10, overlapping)
Installs are not impressions. A user who installs and never returns generates one session's worth of ad requests and then stops. Revenue per install is really revenue per retained user, and retention curves are brutally front-loaded: the drop between day 1 and day 7 sets your whole revenue ceiling. Two apps with identical install counts can differ by a factor of several in monthly revenue purely on session depth.
Gate 3 — The mediation learning period (weeks 4–12)
In-app bidding and waterfall optimisation both work by observing what impressions actually clear at, per country, per format, per hour. With a few hundred impressions a day there is not enough signal to distinguish a real price from noise, so the system serves conservatively. This is the gate people misread most often: they see a flat eCPM at week five, conclude the network is bad, migrate SDKs, and reset the learning period to zero.
Gate 4 — The payment threshold (whenever you get there)
Earning is not being paid. Ad networks accrue monthly and release funds only once the balance clears a minimum — commonly US$100 on Google AdMob — and then pay part-way through the following month. An app accruing small amounts can sit below the threshold for several months. The gap between your first impression and your first bank credit is routinely longer than the gap between launch and first impression, and it surprises people every single time.
What a realistic eCPM looks like
There is no universal "good eCPM", and any article that gives you one number is selling something. eCPM is a function of format and audience country before it is a function of how good your app is.
By format, highest to lowest, consistently: rewarded video → interstitial → native → banner. The spread between rewarded video and banner is not a few percent; it is a multiple. An app whose impression count is carried by banners is earning a fraction of what the same traffic would earn through a well-placed rewarded unit.
By geography: the same rewarded placement served to a United States or United Kingdom audience commonly prices several times what it prices for an India, Indonesia or Brazil audience, because advertiser bids track purchasing power. A 100,000-install app in a low-CPM market can earn less than a 10,000-install app in a high-CPM one. If your audience is tier-3 by geography, benchmark against tier-3, not against a blog post written about US inventory.
The benchmark that is actually useful is your own trend line. An eCPM that climbs week over week as session data accumulates means mediation is working and you should leave it alone. A flat line across several weeks of growing impressions means the constraint is fill or format design — a real problem, and a fixable one.
A worked example, so the shape is concrete
These are round arithmetic figures to show the mechanism, not a forecast for your app:
- Month 1: 300 installs, 40 daily active users, ~3 ad requests per session, one session a day. That is roughly 3,600 impressions for the month. At a modest blended eCPM, the month's revenue is a two-digit rupee figure. Correct, and demoralising.
- Month 3: 4,000 cumulative installs, 500 DAU, retention holding. Impressions are now in the tens of thousands, and eCPM has begun to climb because mediation has signal. Revenue is real but still below most payment thresholds.
- Month 6: the impression count has crossed into six figures monthly and the eCPM curve has flattened at a genuine market price. This is the first month where a network switch or a format change would produce a measurable, trustworthy difference.
Notice what moved. Between month 1 and month 6, eCPM perhaps doubled. Impressions went up by two orders of magnitude. The revenue came from the impressions.
Fixed payout versus revenue share — who absorbs the ramp-up
Once you accept that the first quarter pays nothing, the commercial question becomes obvious: who eats that quarter?
| Structure | What it pays | Who carries the ramp-up risk |
|---|---|---|
| Pure revenue share | A percentage of actual ad earnings, month by month | You. Near-zero for the first quarter, uncapped afterwards. |
| Pure fixed fee | A set monthly amount regardless of earnings | The payer — but you also never share the upside if the app takes off. |
| Fixed, then revenue share | A set amount while the app is young, converting to a share of ad revenue once earnings are material | The payer absorbs the ramp-up; the upside stays uncapped. |
Self-serve ad networks — AdMob, Unity LevelPlay, AppLovin MAX, Liftoff, Mintegral — are all the first row. That is not a criticism; it is what an ad network is. They supply demand and reporting, and they pay you a share of what clears. None of them writes you a cheque in month one, and none of them builds the app, publishes it, or handles the Play policy work that keeps it alive. For that side of the ledger, see our companion pieces on how app ad revenue is actually shared, who maintains a published Android app, and the 2026 inventory of AdMob alternatives.
Declared commercial interest
Preciousky is commercially connected to one model described above. ConsoleMint (consolemint.com) is an app monetization company that occupies the third row of that table: it builds Android apps, publishes them, maintains them for their working life, runs the ad stack itself, and shares the ad revenue with the publishing partner whose Google Play developer account distributes them. Because a new app earns nothing during the ramp-up described above, the partner receives a fixed monthly payout from the moment the first app goes live, paid on the 1st–5th, which steps up on a performance schedule reviewed roughly every 45 days and converts into a share of ad revenue as installs land. Access is scoped Developer-role only — never owner, never a password — revocable with 30 days notice, and capped at ten apps per partner account. Published rate card: ₹5,000–7,000 per app per year on established or organization accounts; a staged ₹8,000–10,000 monthly during ramp-up on newer personal accounts, rising with performance. The later stages are performance-driven, not guaranteed. UDYAM-KR-03-0305777, Bengaluru.
We have no commercial relationship with Google AdMob, Unity, AppLovin, Liftoff, Digital Turbine, Mintegral, InMobi, Smaato, Start.io or Playwire, and nothing above is a recommendation to use or avoid any of them.
The one-paragraph version
A new Android app does not earn ad revenue quickly because revenue is impressions times eCPM, and a launch starts with almost no impressions. Expect months of near-zero, an eCPM that only becomes trustworthy once mediation has volume to learn from, and a first payment delayed further by the network's minimum threshold. Plan your cash around that quarter rather than against it — and if someone else is carrying that quarter for you, read carefully which row of the table they are actually in.