Subway Surfers: the business behind “zero UA”

An audience the ad budget did not buy.

In a 20 October 2026 interview with PocketGamer.biz, SYBO CEO Mathias Gredal Nørvig describes Subway Surfers’ growth as almost entirely unpaid, apart from occasional acquisition tests. He says advertising generates nearly 90% of the original game’s revenue and connects its economics to its continuing virality. He also cautions that its early opportunity is hard to repeat.

These are company-reported observations across the game, not iOS benchmarks. The interview gives no revenue measurement window, platform split or controlled estimate of what caused growth. It is a retrospective, reviewed here on 22 October; the exact interview date is not stated.

For a small iOS studio, the useful question is practical: what would make someone enjoy watching the game, then want to play it themselves—and what would it cost to sustain that interest?

Similar at a glance. Different reasons to continue.

Both official iPhone galleries put a runner, rails and immediate obstacles on screen. The original listing emphasizes fast swipes and dodging trains. City’s listing also describes character and board upgrades, district unlocks and a separate City Tour mode.

Two official iPhone gallery examples

Can a viewer see the next decision?

Open full image ↗Subway Surfers

The coin trail and approaching trains make the immediate route visible. Our design observation: the viewer has something concrete to anticipate.

Open full image ↗Subway Surfers City

Pickups and a barrier create another visible choice. This moment shows the run; it does not show the upgrade or purchase flow around it.

In the same interview, Nørvig contrasts the original’s long-running flow with City’s shorter runs and progression, and says City leans more on in-app purchases. Our interpretation: a familiar visual identity can contain a different reason to return, with different acquisition economics to validate.

Give organic growth its own cost sheet.

Separate media spend from the work that attracts an audience. Count the time spent recording clips, making content, handling community replies and maintaining the game. A low advertising invoice alone does not tell you whether the operation pays for itself.

Keep each acquisition cohort separate. Revenue per install from an established organic audience does not automatically describe people reached by a new paid campaign. Follow that campaign’s net ad and purchase proceeds, its variable costs and its own observation window before increasing spend.

If you allocate content-production costs across acquired users, document the allocation rule and the uncertainty in attribution. Compare it with your LTV and acquisition cost on the same basis; do not divide all organic installs by one successful post’s production cost.

Test one moment people want to finish.

Hypothetical iOS experiment: a small obstacle game has one satisfying last-second dodge. Record an honest, short gameplay clip that begins just before that decision. Show it without explanation to people unfamiliar with the game. Ask what they think happens next, then let them try the same challenge.

Watch where their expectation matches the controls and where it breaks. If the clip promises an action the first session cannot deliver, fix that gap before making more videos.

Next, run a small distribution pilot with a trackable store link and a fixed production budget. Keep views, store visits, attributable first installs and completed first runs separate. For the resulting install cohort, report how many start another run on day 7, using all first installs as the denominator and allowing the whole cohort to reach that age.

That pilot can reveal interest and experience gaps. It cannot prove incremental installs or a retention lift by itself. A randomized comparison or a suitable holdout is needed for those claims. Expand production only when the next batch earns enough useful play and eventual proceeds to justify the effort; keep any paid-acquisition test separate.