The International Monetization Gap: When Growth Costs More Than It Earns

I was visiting a major game studio in Finland when one of their executives said something that has stuck with me since: their growth was killing their profit.

The numbers looked good from the outside. Daily active users were climbing fast, and most of the increase was coming from India and Brazil. Two enormous markets, real players, genuine engagement, the kind of chart a board likes to see.

Underneath it, server and support costs had skyrocketed, and almost none of those new players were paying for anything. Every additional thousand users made the company bigger and slightly less profitable. They had reached the point where the fastest way to improve margin would have been to stop growing.

By 2018 support alone was consuming close to 10% of revenue. Put plainly, money from the people who paid was being spent servicing the people who did not. Their eventual response was to restrict customer support to paying users. They applied it globally, but the decision was driven by getting costs under control in Tier 3 geographies. By December of that year support was under 3% of revenue.

That conversation is the clearest illustration I have of what I would call the international monetization gap, and most studios have some version of it without having measured it.

What Is the International Monetization Gap?

The gap is the distance between where your costs come from and where your revenue comes from.

Cost per user is close to geography-neutral. A player in Sao Paulo consumes roughly the same bandwidth, the same server cycles, the same storage, and files roughly as many support tickets as a player in Stockholm. Some of those costs are marginally cheaper to serve in some regions and marginally more expensive in others, but the variation is small.

Revenue per user is not remotely geography-neutral. In-app purchase conversion and average spend in Tier 1 markets are multiples of what they are in Tier 2 and Tier 3, driven by disposable income, payment method availability, and price sensitivity. Advertising revenue skews the same direction for the same reasons, since advertisers bid more for users they believe will spend.

Put those two facts together and you get the uncomfortable arithmetic the Finnish studio ran into. Growth in a low-monetizing market adds cost in a straight line and revenue in a trickle. Past a certain mix, each new cohort makes the business worse.

Why the Gap Stays Invisible

Because almost nobody reports margin by country.

The standard dashboard shows daily active users, total revenue, and a blended ARPDAU. All three can look healthy while the underlying mix is deteriorating. If your Tier 1 base is flat and your Tier 3 base is doubling, blended ARPDAU falls, and the usual interpretation of falling ARPDAU is that monetization needs work rather than that geography mix has shifted.

The number that would have caught it earlier is contribution margin per daily active user, broken out by country. Revenue from that country, minus the infrastructure and support cost of serving it, divided by users. It is unglamorous and most teams do not have it, which is why the problem tends to be discovered by a finance review rather than by a growth team.

Why Restricting Support Is a Rational Move

It is worth being fair to the decision the studio made, because the logic is sound.

Support is one of the few costs in a game that scales with users and cannot be engineered away. You can optimize server spend, compress assets, and renegotiate CDN rates. You cannot make a human answer a ticket for free. If a large share of your ticket volume comes from users who will never generate revenue, restricting support to payers converts a variable cost into a benefit for the people actually funding the business.

Plenty of large products do exactly this, and it is defensible. Non-paying users were not cut off entirely. They kept automated support, which in practice meant a chatbot with no route through to a person.

Taking support from roughly a tenth of revenue to under a third of that inside a year is a serious result, and no growth initiative in the same period moved margin that far.

The Problem With Solving It on the Cost Side

Cutting cost per non-paying user treats those users as a liability to be minimized. That is one theory of the business. The other is that they are unmonetized revenue, and the difference matters because the two theories point in opposite directions.

Three things make the cost-side theory weaker than it looks.

The non-payers are part of what the payers are buying. In any game with other people in it, the free population is the content. Someone spending heavily is paying to compete, to be seen, to fill a clan, to have a match start in under ten seconds. Thin that population out and you degrade the product for the people funding it. That cost never appears as a support line. It appears weeks later as longer matchmaking times and quieter leaderboards, in a different part of the funnel, attributed to something else.

Payer conversion is a moment, not a trait. Almost nobody spends on their first day. They spend after weeks of investment, or when a sale lands, or when money arrives, which in many of these markets is lumpy and seasonal rather than a monthly salary. A policy that degrades the experience of everyone who has not paid yet is aimed at a group that contains most of next year’s payers. Splitting an audience into payers and non-payers treats a timestamp as a personality.

The revenue cost is real but nobody owns it. A ticket that goes unanswered does not simply save the cost of answering it. It produces a user who leaves and, often enough, a public review. Store ratings feed install conversion for everybody, including the payers you were protecting. The saving lands in a support budget where someone is measured on it. The cost lands in acquisition, where no one is looking for it. That asymmetry is why cost programmes tend to look more successful than they are.

There is also a structural asymmetry between the two approaches. A cost lever has a floor. You can take support from a tenth of revenue to almost nothing exactly once, which the studio did, and that roughly seven points of margin is now banked and gone. There is no second time. Monetizing the same cohort has no equivalent ceiling, and it scales with the growth that created the problem rather than being consumed by it.

The studio was not being careless. They made a cost decision and a revenue decision in separate meetings, which is how most companies do it, and the cost decision had a clear owner and a measurable result while the revenue decision had neither.

Does an Offerwall Close the Gap?

It narrows it. It does not close it, and any vendor telling you otherwise is selling.

The case for it is straightforward. An offerwall earns from users who will not spend cash, which is precisely the cohort creating the gap, and is the argument we set out in our case for monetizing the non-paying majority. In a market where payer conversion is very low, in-app purchase optimization has almost nothing to work with, so a mechanic that does not depend on the user reaching for a card is the main lever available.

The honest limit is that advertiser demand concentrates in the same wealthy markets that already monetize well. Offerwall rates are highest in the United States, the United Kingdom and Canada, and lower in the markets where the gap is widest. So the format helps most where you need it least. That is a real constraint and worth planning around rather than discovering.

What that means in practice is that the goal in Tier 3 is not to match Tier 1 revenue per user. It is to move contribution margin per user from negative to positive. That is a much lower bar, and it is often achievable.

Is Your Cheapest Price Actually Cheap?

Probably not, and this is the part of the gap most studios have never examined.

Almost every game treats roughly a dollar as the entry price. It is the smallest coin pack, the starter bundle, the thing you offer someone who has never spent before. In the United States a dollar is a rounding error. In India, Brazil, Indonesia or Nigeria, converted at the market exchange rate, it is a real purchase. Measured against local wages rather than against the dollar, your cheapest item can be the equivalent of a considered discretionary buy rather than an impulse one.

This reframes the payer conversion problem. The usual reading of a low payer rate in these markets is that the users will not pay. A more accurate reading is that the cheapest thing you sell is expensive, so the only people converting are the ones for whom it is not.

The distinction that matters is between currency conversion and purchasing power. Taking a dollar price and running it through the day’s exchange rate produces a number in the local currency, and studios call that localized pricing. It is not. It holds the dollar value constant and lets the real cost to the buyer float wherever local incomes happen to sit. Pricing to purchasing power parity asks a different question: what price represents the same relative sacrifice for this buyer as a dollar does for a buyer in a wealthy market? The answer is often considerably less than the exchange rate suggests.

The tooling excuse expired in 2023. Apple’s pricing overhaul took developers from around 100 price points to more than 900, starting as low as $0.29 in the United States, with granular increments and the ability to pick one storefront as a base and generate prices across the other 174 storefronts and 44 currencies. Subscriptions got this in December 2022 and everything else followed in spring 2023. Sub-dollar price points exist. Most catalogues simply were not revisited.

Two practical consequences. First, if you have never set a market-specific floor, your entry price in your fastest-growing regions is probably wrong by a wide margin, and that single number gates every other monetization decision. Second, the same logic applies to rewarded currency. A single global conversion ratio is either too stingy to convert in wealthy markets or too generous in cheap ones, which is the same failure in a different currency.

None of this closes the gap on its own. A player who genuinely has no discretionary spend is not reached by a better price point, and that is exactly the cohort a rewarded channel is for. But pricing to purchasing power moves some share of your audience from structurally unable to pay into plausibly able to, and it is cheap to test.

What Actually Works in Lower-Monetizing Markets

• Survey supply. Surveys reach users who will not install anything and require no spend. They pay less per completion than an install with a deep engagement event, but they convert a population that would otherwise contribute nothing, and availability is broad.

• Local and regional advertisers. Global brand campaigns are Tier 1 weighted almost by definition. Catalogue depth in your specific countries matters far more than a network’s total offer count, and it is the first thing to ask about when evaluating a provider.

• Lower-friction conversion events. A deep engagement event that works in a market with fast phones and cheap data may be unreachable elsewhere. Matching event difficulty to device and network conditions changes completion rates substantially.

• Reward pricing set per market. A single global currency conversion ratio is either too stingy to convert in wealthy markets or too generous in cheap ones. Anchoring the ratio to local purchasing power is the same discipline as localized IAP pricing, and it is covered in our guide to anchoring the ratio to local pricing.

• Retail gift cards and a brand web store. Moving the transaction off the app store avoids the platform commission, and a gift card on a shelf reaches people without a card on file. Supercell runs its own store with gift cards redeemed against a Supercell ID, and Pokemon GO cards sit in the same retail racks. The honest caveat is that this is a brand privilege rather than a tactic. It requires enough recognition to earn shelf space, which most studios do not have. The April 2025 Epic v. Apple ruling made linking to a web store from inside an iOS app considerably easier, so the mechanic is more available than it was.

• Events and tournaments. Competitive events concentrate engagement into a window and monetize through participation rather than through individual purchasing power, which suits markets where many people play and few people spend. They also give local partners something to attach a sponsorship to.

• Deep local partnerships and offline tie-ins. The strongest version of this is a distribution partner who already bills your users. Telefonica’s Brazilian operator Vivo ran a Fortnite promotion in August 2025 giving its Vivo Valoriza loyalty customers an exclusive outfit, redeemed with a code in the Vivo app and unlocked through a quest in the game. The promotion is the visible part; the substantive part is the relationship underneath it, since Telefonica had already rolled out direct carrier billing for Epic and preinstalled the Epic Games Store on Android devices across Latin America in 2024. Carrier billing matters enormously in markets where card penetration is low, because it removes the payment barrier rather than working around it.

• Cost reduction that does not touch trust. Asset compression, regional CDN routing, and lower-bandwidth modes reduce the cost side without degrading the experience of the users you are trying to monetize.

What to Measure Instead

If you take one thing from the Finnish studio’s experience, make it this: report by country, and report margin rather than revenue.

• Contribution margin per daily active user, by country. Revenue minus infrastructure and support, per user.

• Growth in users versus growth in gross margin, plotted together. If they diverge, your mix is shifting.

• Support tickets per thousand users, by country, so the cost you are trying to control is actually visible.

• Rewarded revenue per user by country, so you can see which markets crossed into positive contribution and which have not.

The point of the exercise is not to stop growing in large markets. It is to know which of them pay for themselves, so the decision to keep growing is a decision rather than an accident.

The Takeaway

The Finnish studio’s problem was not that Indian and Brazilian players were bad users. It was that the company had a cost model that scaled with every user and a revenue model that only worked on a few of them, and no reporting that would have shown the two diverging until finance noticed.

Cutting support for non-payers closes the gap from one side. Monetizing them closes it from the other, and only one of those options gets better as you grow.

If you want to look at what rewarded revenue per user actually looks like in your markets rather than a blended global figure, see how RevU works for publishers or talk to our team.