An investment decision on a mobile app is not built on the size of the market but on the economics of a single unit. Sensor Tower’s State of Mobile 2026 report puts consumer in-app spending at $167 billion in 2025, up 10% on the year before, and Apple says the App Store ecosystem facilitated $1.4 trillion in developer billings and sales in the same year. Against that, RevenueCat’s State of Subscription Apps 2026 reports that only 17.3% of subscription apps reach $1,000 in monthly revenue within two years of launch, and only 4.6% reach $10,000. The decision this article addresses sits between those two numbers: is the app an investment whose return can be calculated in advance, or a marketing expense dressed as one?

The market in numbers: large, and concentrated

The app market has not stopped growing, but the nature of spending in it has changed. According to Sensor Tower, consumer spending in non-game apps overtook spending in games for the first time in 2025, reaching roughly $85 billion, up 21%, driven by generative-AI services, while users spent 5.3 trillion hours inside apps that year. The Analysis Group study Apple published in June 2026 breaks the $1.4 trillion into $1.1 trillion of physical goods and services bought through apps, $149 billion of digital goods, and $151 billion of in-app advertising, with more than 850 million weekly users; it notes that more than 90% of that value paid Apple no commission, because it is physical commerce and advertising that passes through the app rather than being sold inside it.

On the user side, Saudi Arabia is one of the most app-ready markets in the world. The Saudi Internet Report 2025, issued by the Communications, Space and Technology Commission in July 2026, reports that 61.3% of users spend seven hours or more online every day, that average mobile data consumption is 53 GB per person per month, three times the global average, and that 45.2% of users already use AI tools, while internet penetration in the report dashboard stands at 99.6%. On the supply side, the General Authority for Statistics puts the digital economy’s share of GDP at 16.0% in 2024, with computer programming activities alone earning SAR 31.1 billion in revenue that year.

The app market, one line per indicator
IndicatorValueSource
Global consumer in-app spending, 2025$167 billion (+10%)Sensor Tower
Non-game app spending, 2025about $85 billion (+21%)Sensor Tower
App Store ecosystem, 2025$1.4 trillion, of which $149 billion digital goodsApple / Analysis Group
Subscription apps reaching $1,000 a month within two years17.3%RevenueCat
Internet penetration in Saudi Arabia, 202599.6%CST
Digital economy share of Saudi GDP, 202416.0%GASTAT

Growth in the app market accrues to the few, not to everyone. RevenueCat’s report, built on more than 115,000 apps and $16 billion in revenue, finds that apps launched before 2020 still generate 69% of all subscription revenue despite a sevenfold surge in new launches since 2022, and that the top quartile of apps grew monthly revenue 80% in a year while the bottom quartile shrank 33%. An investor who reads the market figure alone sees an ocean; one who reads the distribution sees that most of it flows to a handful of boats.

Three kinds of app investment, three return equations

Most app-investment mistakes begin by confusing three different things that carry the same name. The app a restaurant, clinic, or retail chain builds for its customers is not a product for sale; it is a distribution channel that replaces an intermediary, and its return is measured by what it saves and lifts in the existing business. An app launched on its own to be subscribed to or to carry advertising is a standalone product, and its return is measured by customer lifetime value against acquisition cost. Acquiring an existing app, or a stake in one, is a financial investment, and its return is measured by current cash flow and the exit multiple. Each type has its equation, and the wrong equation gives the wrong decision even when the numbers are right.

Three types, three yardsticks
TypeWhere the return comes fromThe right measureThe misleading measure
Channel app for an existing businessIntermediary commissions saved, higher purchase frequency, lower cost to serveIncremental annual margin against build and running costDownload count
Standalone product appSubscriptions, in-app purchases, advertisingCustomer lifetime value ÷ acquisition costStore ratings
Acquisition of an existing appRecurring cash flow and a multiple at exitRecurring-revenue multiple with retention qualityMonthly download growth

Unit economics: five numbers that govern the return

  1. User acquisition cost: what is paid for one paid install, and its ratio to organic installs. An app that relies on paid advertising alone buys every customer at market price.
  2. Trial-to-paid conversion: RevenueCat reports a median day-35 conversion of 10.7% for apps with a hard paywall against 2.1% for freemium apps, a fivefold difference.
  3. Retention: the same report puts one-year retention at roughly 27% to 28% of subscribers, so recurring revenue must replace three quarters of the base every year before it can be called growth.
  4. Store commission: 15% on the first $1 million of annual proceeds under the App Store Small Business Program and under Google Play’s service fees, and 30% above that in the Saudi market, so the app loses a sixth of its revenue before it reaches the bank.
  5. Annual running cost: servers, AI interfaces, support, and the mandatory updates that follow every new operating system; a number that appears in no build proposal and in every invoice after it.

The five numbers collapse into one: revenue per install. RevenueCat reports that a hard-paywall app earns a median $3.09 per install within sixty days against $0.38 for a freemium app. If the cost of a paid install in the Saudi market exceeds that figure, and in most categories it does, the app needs an organic base, a higher price, or better-than-median retention, or all three, before it becomes an investment.

A worked example: a subscription app in disguised numbers

The figures in this example are illustrative and disguised, composed from more than one file; they are not any single client’s numbers, nor the Group’s own apps. A subscription app aimed at the Saudi market costs SAR 420,000 to build and SAR 12,000 a month to run, and targets 50,000 installs a year. Two scenarios were calculated with identical build and running assumptions, differing only in four operating numbers: the paid share of installs and its cost, the trial-start rate, the conversion rate, and price with retention. The first sits at the market median as RevenueCat publishes it; the second at the edge of the top quartile. Commission is 15% in both.

The two scenarios, assumptions and results (Saudi riyals)
VariableScenario one: market medianScenario two: top quartile
Paid installs and cost per install60% at SAR 640% at SAR 4
Trial start, then conversion to paid25%, then 10%30%, then 15%
Annual price and one-year retentionSAR 149 and 27%SAR 249 and 35%
Subscribers at the end of year one1,2502,250
Year-one net after build and marketingloss of 586,000loss of 168,000
Year-two netloss of 123,000profit of 419,000
Cumulative position at the end of year threeloss of 820,000profit of 728,000

The gap between a SAR 820,000 loss and a SAR 728,000 profit in this example lies neither in the build nor in the market, but in four operating numbers that can be measured before a single line of code is written. Scenario one never breaks even because net revenue per install is about SAR 3.2 while a paid install costs SAR 6, so every marketing campaign deepens the loss. Scenario two breaks even during year two because net revenue per install is SAR 9.5 against a cost of SAR 4. Nobody reaches the top quartile by wishing; it takes a deliberate paywall, a price that reflects value rather than fear, and a first session that brings the user back on day two.

The channel app: an entirely different calculation

For the owner of an existing business the equation is usually simpler and faster. The figures in this example are illustrative and disguised, composed from more than one file; they are not any single client’s numbers, nor the Group’s own apps. A six-branch restaurant chain sells SAR 9 million a year through third-party delivery apps at a 25% commission, that is SAR 2.25 million a year in commissions. Its own app costs SAR 350,000 to build, SAR 120,000 a year to run, and SAR 150,000 a year to market. If it moves just a third of those sales to its own channel it saves SAR 750,000 a year, recovering its cost in the first year and earning about SAR 480,000 in every year after from saved commissions alone, before counting purchase frequency and customer data. Here the app is measured neither by downloads nor by ratings, but by the share of sales that moved, and that share can be tested with a single campaign before the full build.

Five signs the app is an expense, not an investment

  • There is no baseline: nobody knows what a customer costs today or how much they buy, so nobody can know what improved.
  • The stated return is “presence” and “customer experience”: descriptions that cannot be divided by a cost.
  • The marketing plan starts after launch: meaning acquisition cost was never in the feasibility study.
  • The budget ends at delivery: no line for a full year of running costs or for updates to both operating systems.
  • The decision is taken because a competitor launched an app: a reason to study the market, not to build a product.

What is settled before the first line of code

  • The investment type and its equation: channel, product, or acquisition, and the single number it will be judged by after twelve months.
  • A numeric baseline: current customer cost, purchase frequency, and commissions paid to intermediaries, before any design is touched.
  • A demand test without an app: one page, a small campaign, and a waiting list measure the real cost of attention in the local market within three weeks.
  • A decision gate at ninety days after launch: actual retention and conversion are compared with the assumptions, and the choice to expand, adjust, or stop is made without sentiment.

IAD Business Services Group looks at an app from two practices at once: technology and digital transformation, where it is built, and private investment, where it is held to account. The Group also operates its own apps published on the App Store, so the numbers it asks clients for before a build are the same numbers it watches on its own dashboards every morning: cost per install, conversion rate, thirty-day retention, and net revenue after commission. Separating those who build from those who keep score is what stops an app from becoming a handsome project nobody has the courage to shut down.

Conclusion

Investing in apps is a real investment in a real market: $167 billion spent inside apps every year, an ecosystem worth $1.4 trillion, and a Saudi market where 61% of users spend seven hours a day on their phones. But its distribution is sharp: 17.3% of subscription apps reach $1,000 a month, and older apps take two thirds of the revenue. What separates those who get there from those who do not is rarely the idea or the quality of the code; it is four operating numbers tested before the build and audited after it. Those who hold these numbers invest. Those who do not, spend.

An app is not measured by how many people downloaded it, but by what is left in the account after commission, marketing, and running costs. That number is calculated before the build, or never.

Sources

Technology & Digital Solutions