MyAIDevelopment · Product · Operations notes

Game Monetization

Learning goal: Tell apart the money structures of premium, F2P and ad-supported games, know the key Steam and mobile numbers and Korean rules, and keep a solo developer's game scope under control.

Games are the highest-variance asset in a solo studio's portfolio: in document 02's terms, the aggressive end of the barbell. This document covers how to shrink that bet to a size you can calculate.

As of: 2026-09-29.

Key Concepts

Model Who pays Key metrics Fit for a solo developer
Premium (paid) Each buyer, once Wishlists, conversion, price High: low operating load, a proven path on PC
F2P + IAP A small share of spending users Retention, ARPDAU, payer conversion Low: content operations and user acquisition costs are large
Ad-supported (rewarded ads and others) Advertisers DAU, ad impressions, eCPM Medium: assumes huge install volume
Hybrid Ads + IAP ARPDAU Medium: complex to design

Principles

How Money Works on Steam

Item Detail Source
Revenue share Per game lifetime revenue: Valve 30% up to $10M, 25% from $10M to $50M, 20% above $50M Valve announcement, applies to revenue after 2018-10-01
Steam Direct fee $100 per app, non-refundable. Recouped in payment once adjusted gross revenue passes $1,000 Steamworks documentation
Revenue distribution Of about 41,000 games released over the prior three years, half grossed $500 or less Game World Observer, reported 2023-10-06

The last row is how document 01's power law shows up in games. The median game does not earn a living.

Wishlists and Launch Visibility

flowchart LR
    P[Store page goes live] --> W[Wishlists accumulate]
    D[Demo and Next Fest] --> W
    C[Content and community] --> W
    W --> L[Launch notification email]
    L --> S[First-week sales]
    S --> A{Algorithm widens exposure}
    A -->|strong response| T[More sales]
    A -->|weak response| F[Tail revenue from discount events]

Mobile Game Numbers

Applied: The Example Studio

The example studio (assumption) has three game prototypes: GM1 roguelike puzzle (PC), GM2 hyper-casual (mobile), GM3 narrative adventure (PC).

If GM1 sells on Steam at $9.99 (all assumptions)

Assumed average net unit price after VAT, refunds, regional prices and discounts: $7
After Valve's 30% → 7 × 0.7 = $4.90 → KRW 6,860 at KRW 1,400
Launch wishlists 1,500 (the D90 double-down bar in 09 and 10) × first week 15% (heuristic) = 225 units
First-week revenue = 225 × 4.90 = $1,102.5 ≈ KRW 1.544 million (KRW 1,543,500)
To earn KRW 3.5 million in the first week: 3,500,000 ÷ 6,860 = 510.2 → 511 units
Wishlists needed = 511 ÷ 0.15 = 3,406.7 → about 3,407

1,500 is the double-down bar for starting Next Fest preparation, not a launch target. Launching straight at 1,500 would bring first-week revenue of only about 44% of one month's burn (1.544 ÷ 3.5). So GM1's plan is to gather about 3,400 or more wishlists through Next Fest and launch after that.

If GM2 runs ad-supported (all assumptions)

Target DAU 2,000 × ARPDAU $0.04 = $80 a day
30 days = $2,400 ≈ KRW 3.36 million (gross, before ad network settlement)
Assume 3 active days per install on average → holding DAU 2,000 needs about 667 installs a day, 20,000 a month

Without a channel that brings 20,000 installs a month without ads, GM2 does not work on paper. GM1 has wishlists as a pre-launch leading indicator, so the bet size can be adjusted.

Bad example:

GM3, now with an open-world editor and multiplayer, has been in production for two years. It still has no store page.

Good example:

Built a 20-minute demo containing only GM1's core loop first, opened the store page, watched the pace of wishlist growth, then set the production scope.

Scope Control for a Solo Developer

Going Deeper

Duties When Selling Games in Korea (checked 2026-09-29)

Duty Detail What it means for a solo developer
Age rating A game must be rated before distribution (Game Industry Promotion Act Article 21). Platforms designated as self-rating operators, such as Google and Apple, handle it through a store questionnaire On mobile stores this often ends with the questionnaire
Steam and other non-designated platforms As of 2026-09, no announcement of Valve's designation as a self-rating operator could be confirmed (latest report 2024-07, designation under consideration) PC games rated All, 12+ or 15+ apply to the Game Contents Rating Board (GCRB), adults-only to the Game Rating and Administration Committee (GRAC). Fee = base amount per platform (PC KRW 360,000) × usage factor (networked 1.5, non-networked 1.0) × genre factor (group 1: 4.0, group 2: 2.0, group 3: 1.0) × localization factor (Korean 1.0, not Korean 1.5). Example: a Korean-language, non-networked group-3 PC game KRW 360,000; a Korean-language, networked group-1 game KRW 2,160,000. A game production business registration certificate is required
Loot box probability disclosure From 2024-03-22, a duty to display item types and probabilities (Enforcement Decree amendment) If you add gacha or random boxes, show the odds in the game and on the website
Damages for disclosure violations In force from 2025-08-01: up to three times the damages for intentional violations, and the company must prove it was not at fault Loot boxes carry large legal risk for one person

Common Misconceptions

Self-Check Questions

  1. GM2 (hyper-casual, mobile) must run as premium, F2P or ad-supported. For each model, write one key metric to track in the first month and the operating load the model demands, then pick the one that fits a solo studio with 160 hours a month.
  2. With a net unit price of $4.90 and 15% first-week conversion, how many wishlists does KRW 5 million in first-week revenue need? (KRW 1,400 per dollar)
  3. What does a mobile game with 22% D1 need to hold its DAU?
  4. Which rating path must a Korean solo developer releasing a PC game on Steam take?
  5. You decide to add random boxes to GM2. What must be ready in the game and on the website before launch, and what damages risk does a solo developer carry if the displayed odds are wrong?

References