Start with active users, not a revenue growth rate
An app can earn from subscriptions, advertising and in-app purchases, but none of those streams is simply a percentage of downloads. First forecast how people discover the app, install it, become active and return. Then define which active users are eligible to subscribe, see ads or buy something in the app.
The three streams can coexist, but they do not suit every product equally. A subscription needs continuing value; advertising needs usable inventory and demand from advertisers; in-app purchases need something worth buying more than once or at a clear moment of need. A financial model should make each claim testable.
Organic + paid + referral installs → active users → subscribers, ad impressions and purchases → revenue → costs and cash
1. Forecast organic, paid and referral installs separately
The workbook starts with 100 organic installs in the launch month and assumes 10% monthly growth, subject to an upper limit. Organic discovery is not automatically free: content, app-store optimization, partnerships and brand work may require spending even if they do not have a neat cost per install.
Paid acquisition starts with a $5,000 monthly budget and an illustrative $3.50 cost per install (CPI). That yields about 1,429 paid installs in May. The model can change the budget at roadmap milestones and reduce the assumed CPI over time, but neither improvement should be treated as guaranteed.
Paid installs = paid acquisition budget ÷ cost per install
New installs = organic + paid + referral installs
Mobile app model · workbook extract
Referrals depend on people who were active in the preceding month. With no prior active users at launch, May has no referral installs; they begin to appear in June. If an incentive is paid per successful referral, include its cost. Keep attributed installs separate so the same person is not counted in both a paid and a referral channel.
CPI measures an install, not a retained user or a paying customer. Check the costs of activating and retaining people before using it as evidence of attractive acquisition economics.
2. Convert installs into monthly active users
In May, 100 organic installs plus approximately 1,429 paid installs become about 1,529 new installs. The model applies 95% activation to reach roughly 1,452 new monthly active users (MAU). The launch month has no opening active users, so the closing MAU is also about 1,452.
New MAU = new installs × activation rate
Closing MAU = opening MAU + new MAU − lost MAU
| Metric | May 2025 | June 2025 |
|---|---|---|
| Organic installs | 100 | 110 |
| Paid installs | 1,429 | 1,473 |
| Referral installs | 0 | 145 |
| Total new installs | 1,529 | 1,728 |
| New active users | 1,452 | 1,642 |
| Lost active users | 0 | 1,147 |
| Closing MAU | 1,452 | 1,947 |
The workbook begins with a 20% month-to-month retention assumption for the existing active base and increases that rate gradually, with a cap. That is a demanding assumption to examine: 20% retention means roughly four out of five of the previous month's active users do not remain active in the next month. A larger inflow of installs can hide that loss in the total MAU line.
Once the app has data, compare this simple roll-forward with cohorts: how many users return after their first week or month, how behavior differs by channel, and whether the users who see ads behave like those who pay. Define an active user consistently; an install, an account and a monthly active user are not interchangeable.
3. Model subscribers, churn and store deductions
A subscription forecast needs an eligible audience, conversion, price, renewals and cancellations. In the workbook, new subscribers are calculated as 4% of closing MAU each month. May's 1,452 active users produce 58 subscribers after rounding. The illustration charges $19 per month and applies 15% monthly subscriber churn to the opening subscriber base.
Mobile app model · workbook extract
Closing subscribers = opening subscribers + new subscribers − lost subscribers
Net subscription revenue = billable subscribers × price − applicable store deductions
May has no opening subscribers, so 58 closing subscribers at $19 produce $1,102 gross subscription billings. The workbook's illustrative 15% blended store deduction is $165.30, leaving $936.70 in its net subscription revenue line.
This simplified forecast bills the closing subscriber count for a full month and calculates new conversion from total MAU, not only eligible non-subscribers. For a real product, distinguish trials, first payments, renewals, refunds and upgrades; cap new conversions to the eligible population. Annual prepayments also create different cash and revenue timing.
4. Tie advertising to delivered impressions
Advertising revenue depends on actual opportunities to show an ad, not simply the number of downloads. The illustrative workbook assumes 30 engaged sessions per active user per month, five minutes per session and two ad impressions per minute. That produces 300 modeled impressions per active user each month.
Mobile app model · workbook extract
Modeled impressions = ad-eligible MAU × sessions × minutes per session × impressions per minute
Ad revenue = delivered impressions ÷ 1,000 × effective CPM
For May, the workbook applies 300 impressions to all approximately 1,452 MAU, producing about 435,643 impressions. At an illustrative $5 per thousand and no assumed platform fee, that is $2,178.21 in ad revenue. Ten ads per five-minute session is a heavy ad load for many product experiences; test it against actual placements, engagement and retention.
5. Forecast purchasers, orders and item mix
For in-app purchases (IAP), separate the share of active users who buy from how much each buyer spends. The workbook assumes that 2% of active users who are not subscribers make a purchase. It gives two example items a $35 and $25 price and a 20%/80% sales mix, yielding a $27 average item price.
Mobile app model · workbook extract
Purchasers = eligible MAU × purchase conversion
Gross IAP revenue = purchasers × orders per purchaser × items per order × weighted item price
May's 1,452 MAU less 58 subscribers leave about 1,394 non-subscribers. Two percent, rounded, gives 28 purchasers; two one-item orders at an average $27 produce $1,512 gross IAP billings. With the workbook's illustrative 30% store deduction, the net IAP revenue is $1,058.40.
The model deliberately excludes subscribers from the purchasing pool. If your paid members can also purchase add-ons, model that as a distinct eligible segment rather than assuming they never buy. Test purchase frequency with cohorts and real order data; repeat purchases cannot be inferred from initial conversion alone.
6. Compare revenue with acquisition and delivery costs
The three streams meet in the same monthly forecast, but they do not carry the same evidence or cost structure. May's modeled net revenue, after the specific store and ad-platform deductions assumed above, is:
| Revenue stream | Workbook calculation | Net revenue |
|---|---|---|
| Subscriptions | 58 × $19 − 15% store deduction | $936.70 |
| Advertising | Unrounded impressions ÷ 1,000 × $5 CPM | $2,178.21 |
| In-app purchases | 28 buyers × $54 − 30% store deduction | $1,058.40 |
| Total | Sum of the three streams | $4,173.31 |
Change one assumption at a time to see which decision moves: lower retention, fewer ad impressions, weaker purchase conversion, a different store-fee mix or a delayed campaign. Track contribution after the costs needed to serve users, and follow cash separately from recognized revenue. Do not use a CPI, a revenue-per-user figure or a blended LTV/CAC ratio without checking its audience, period and included costs.
What the model should help you decide
Choose the stream that fits what users value and how they actually use the app. If a subscription removes ads, reflect that trade-off in both forecasts. If in-app purchases can coexist with subscriptions, define the overlap. Treat store deductions, refunds and delivery costs as part of the economics rather than a footnote.
Before increasing acquisition spend, validate activation, month-to-month retention, delivered ad impressions and willingness to pay with small tests. Then compare actuals with the model monthly and update the assumptions that changed. The useful output is not a single five-year revenue line: it is a clearer choice about pricing, product experience, spending and how much cash is needed to reach the next milestone.