4X games buy whales and pay for tourists

Every 4X marketing plan contains a cross-subsidy that nobody writes down. The studio spends on installs it knows will never pay, deliberately, because the payers it is really buying won't stay in a game where the map is empty. Tourists aren't waste. They're the product the whales are purchasing, and most 4X UA teams under-invest in that layer while the networks get better and better at finding payers, which quietly makes the problem worse. Two ledgers, each managed against a different goal.

A seasoned buyer will object that every dollar spent on a non-payer is a dollar of ROAS left on the table, and that the whole point of value-based bidding is to stop paying for people who don't pay. Fair enough for a single-player puzzle game. It's wrong for a game whose core loop is other people.

The loop needs bodies before it needs wallets

A 4X server is a social economy. The high-spending player buys speed-ups and troops to win fights, and a fight only means something when somebody sits on the other side of it: rivals to humiliate, alliances to lead, a leaderboard that other people are climbing too. Strip out the low-spending population and the whale pays to beat nobody.

This is why the genre's per-install economics look strange from outside. Sensor Tower's State of Gaming 2026 in February put strategy alongside puzzle as the categories leading revenue growth while total installs fell about seven percent, and Adjust's March report showed strategy sessions up sharply year on year. Revenue is concentrating on a small number of players spending more, in games that still buy large volumes of installs. No contradiction there. The volume is the environment that keeps the concentration going.

UA Ledger's "Inside the creative machine behind 4X acquisition" in February covered how these teams industrialise creative. What gets less attention is who all that creative recruits: two different people, for two different jobs.

Two ledgers, not one campaign

So a 4X budget does two jobs, and the accounting should say so.

The population ledger buys installs for one reason: full servers, staffed alliances, contested leaderboards. Its metrics are day-one and day-seven retention, alliance join rate, server fill rate. Cost per install should stay low and the creative can run broad. ROAS has no business anywhere near it.

The payer ledger buys the small number of players who will carry the revenue. Return on ad spend over a long window is the metric, cost per install will be high, and the creative has to get specific about the power fantasy and the competitive stakes.

As an illustrative example only, a team spending a hundred units a week might run seventy through population campaigns at a low CPI target and thirty through value-optimised campaigns at a CPI several times higher. Blended ROAS on that plan looks worse than a payer-only plan for the first month and better by the third, because the payers in a populated server stay. A payer-only plan looks efficient right up until the servers thin out and the payers leave for a rival with a fuller map.

The second-order effect: better targeting empties the map

Most commentary on the genre misses the trade-off sitting underneath all of this. Every network is getting better at finding likely payers, and 4X advertisers rank among the most enthusiastic adopters of value-based and ROAS-target bidding. AppLovin's automated tooling and Moloco's ROAS range controls do it. So do the equivalents on Meta and Google. All of them promise to spend less on people who won't buy.

For a single-player game that's a pure gain. For a 4X game it's a slow leak. As the payer campaigns get more precise, the incidental tourist inflow that used to arrive alongside the payers shrinks. Servers fill more slowly, so alliances have fewer members to recruit, and the payers who do arrive find a quieter game where their spend, which depends on rivalry, softens. The dashboard reports rising ROAS on the campaign and falling revenue on the server, and nobody reads those two numbers in the same meeting.

The correction isn't to stop using value bidding. It's to fund the population ledger explicitly, because the networks will no longer fund it by accident.

A decision rule for the split

Server health should set the ratio between the two ledgers, not finance. A rule that has held up in practice:

• When new-server fill time is lengthening or alliance join rate is falling, move budget from the payer ledger to the population ledger, even if blended ROAS dips.
• When servers are filling ahead of the merge schedule and payer retention is stable, move budget back toward the payer ledger.

There's a floor under all of it. Never let the population ledger drop below the point where a new server fills inside the game's first-week retention window, because a payer who lands in an empty server churns before the alliance mechanic gets a chance to hold them.

The signal telling you which way to move is the ratio of payers to active players per server, tracked by cohort week. A rising ratio doesn't mean monetisation is improving. It means the population around each payer is thinning.

Consolidation in the genre makes this more urgent, not less. PocketGamer.biz and Aream reported in early April that Q1 games M&A reached about $7.7B across 52 deals, the strongest quarter since the post-pandemic boom, and strategy publishers are among the acquirers. A studio bought for its 4X revenue will face pressure to show ROAS discipline, and the easiest discipline to show is cutting the installs that don't pay. That's the cut that guts the asset.

What to put in front of finance

The argument that wins the budget meeting isn't that tourists are cheap. It's that the payer's lifetime value is a function of server population, and that population is a cost of goods for this genre in the same way server infrastructure is. Present the population ledger as the cost of keeping the product playable and the payer ledger as growth spend, and the ROAS conversation moves to the right ledger.

That payers-to-players ratio belongs on the weekly report next to ROAS. When it climbs, the map is emptying, and the fix sits on the population side long before any of it shows up in revenue.

Related archive reading

These articles provide related context and remain subject to their stated review status.

• Idle games and the economics of cheap installs
• Ad-revenue and IAP games need different UA maths