Valve: The private 'Half-Life' team that has dominated PC gaming for 30 years through Steam's 30% cut
Founded: Gabe Newell, Mike Harrington · Valve Corporation
Key Fields
FIELD STAMPSOrigin
In 1996, Gabe Newell and Mike Harrington, who had spent 13 years at Microsoft working on three generations of Windows, walked away from the 'stable, high-paying' corporate life to bootstrap their own company with millions earned from Microsoft stock. As avid fans of 'DOOM', they bet that the PC would become the ultimate gaming console and purchased the license for the Quake engine to build their own game. Their success wasn't a stroke of genius, but the result of nearly two years of work and a complete overhaul of 'Half-Life'. Gabe famously delayed the game multiple times, insisting that 'a delayed game is eventually good, but a rushed game is forever bad.' 'Half-Life' was a massive success upon its 1998 release, but the true key to their fortune came in 2003: Steam, a download tool created to automate patches for 'Counter-Strike', eventually evolved into the tax-collecting machine for the entire PC gaming industry.
Milestones
Turning Points
- In 1997, choosing to rebuild 'Half-Life' despite the risk of bankruptcy, prioritizing quality over meeting a fiscal deadline.
- In 2003, forcing the Steam bundle despite public outcry, turning a patch tool into a toll booth for PC games.
- In 2005, opening Steam to third parties, shifting from a game developer to a platform operator, and changing the revenue model from product sales to rent collection.
- In 2018, responding to Epic's 12% commission by offering tiered cuts to top-tier publishers rather than a blanket price drop, protecting the platform's tax base.
- Refusing to go public or take external funding, keeping all Steam profits within the company to ensure long-termism without outside interference.
Failures & Pitfalls
- The initial version of 'Half-Life' was almost entirely scrapped after a year of work, causing the company to miss the 1997 Christmas window and nearly run out of cash.
- Steam's launch in 2003-2004 was plagued by instability and the forced binding of 'Half-Life 2', leading to massive player boycotts and a damaged reputation.
- The 2015 paid mod program with Bethesda was withdrawn within a week due to community rage, becoming a rare public admission of failure for Valve.
- Several VR games initiated around 2014 and numerous projects reported by Ars Technica (including multiple attempts at 'Half-Life 3') were cancelled and never released.
- The first-generation Steam Machine (2015, in partnership with Alienware and others) failed due to confusing specs and high prices, only redeemed later by the success of the Steam Deck.
关键成功要素
- Entering the market through 'patch distribution'—a dirty job no one else wanted—and turning it into essential infrastructure, the deepest of moats.
- Using top-tier internal IPs ('Half-Life 2', 'Counter-Strike') to drive traffic to the new platform; hardware-level cold starts rely on content hegemony.
- A small organization of 300+ people with no public listing, no external funding, and no middle management, achieving per-capita profits that tech giants can only envy.
- Maintaining the 30% commission as long as there were no viable alternatives, and only making precise concessions to top-tier partners when threatened, without compromising the core model.
- Ten years of hardware trial and error: evolving from the failed first-gen Steam Machine to the Steam Deck and the 2026 console, using software ecosystems to nurture hardware.
Lessons
- Great companies often start by solving their own product pain points—Steam was originally created just to update Valve's own games.
- Once a platform tax is established, it is the most powerful business model, but the power to tax has limits; overreaching into the community (paid mods) will lead to backlash.
- You can do things your way without relying on capital: Valve never went public, allowing them to tolerate a decade of hardware failures and the eternal delay of 'Half-Life 3'.
- When facing low-price competition, don't engage in a full-scale price war; tiered commissions that satisfy top-tier partners while leaving others with nowhere else to go is the art of defense.
- Delays are not a disgrace but a brand asset—the mantra 'a delayed game is eventually good, but a rushed game is forever bad' has been proven for 30 years.
- Organizational scale is not a prerequisite for competitiveness; the per-capita output of 300+ people can crush gaming giants with tens of thousands of employees.
Core Data
- 2024 Estimated Steam Platform Revenue:Approximately $9.9 billion (estimated, independent verification pending)
- Steam Commission Rate:30% (25% for revenue over $10 million, 20% for revenue over $50 million) (publicly available data, independent verification pending)
- Company Employee Count:Approximately 350 (publicly available data, independent verification pending)
- Half-Life Original Sales:Approximately 9.3 million copies (publicly available data, independent verification pending)
- Gabe Newell Estimated Net Worth:Approximately $9.5 billion (estimated, independent verification pending)
- Steam Concurrent User Peak:Surpassed 41 million in 2025 (publicly available data, independent verification pending)
- Revenue disclosed in Eyres lawsuit and Epic antitrust case:2021 EU fine of 1.6 million Euros against Valve (publicly available data, independent verification pending)
Competitors / Peers
In PC digital game distribution, the Epic Games Store competes with a 12% commission and free weekly games, but has suffered years of losses and holds only about 10% market share. GOG (owned by CD Projekt) focuses on DRM-free retro games and has a smaller footprint. Microsoft uses the Xbox Game Pass subscription and Battle.net to leverage its own IPs, with Blizzard Entertainment having been a peer competitor for years. On the console side, Sony's PlayStation and the Nintendo eShop are also 'walled gardens' with 30% commissions. Valve's barrier lies in the inventory assets, social connections, and achievement systems built over nearly two decades—players aren't just buying games; they are keeping their library on Steam, something Epic hasn't been able to displace despite spending billions.
- https://36kr.com/p/3371044799008772
- https://www.163.com/dy/article/L53SFSGU0552OI16.html
- https://zh.wikipedia.org/wiki/%E7%B6%AD%E7%88%BE%E7%A6%8F%E5%85%AC%E5%8F%B8
- https://www.jiemian.com/article/1837161.html
- https://www.ign.com.cn/steam-machine/60904/steam-machinewei-shi-yao-mai-1049mei-yuan-valvexiang-tan-jie-ge-gong-huo-yu-wei-lai-gui-hua
- https://www.71396.com/s-2-f3cb3f924c582c32/