businessbriefs
11:58in productionCh. 1 · What Valve Actually Is/ 11:58 · ceiling 15 min
Companies · Strategy

Valve Corporation

Gabe Newell & Mike Harrington · 1996

Valve didn’t build Steam to serve gamers—it built it to make decentralisation financially viable.

Valve is a rare case where platform ownership fully decouples creative output from financial sustainability. Its flat structure is not a virtue—it is a tax the company pays for avoiding managerial overhead, made bearable only by Steam’s dominance. It does not scale. It does not replicate. It survives.

CompaniesStrategyPlatformsVideo gamesSoftwarePlatforms across the network →
Chapters & takeaways6
  1. 1:00
    What Valve Actually Is

    Valve is not just a game studio—it is a vertically integrated platform operator founded by two ex-Microsoft engineers.

  2. 2:04
    The Flat Structure That Isn’t Agile

    No managers, no job titles, no fixed teams—just employees moving between projects based on interest and momentum.

  3. 3:22
    How Valve Turns Users Into Contractors

    Valve’s SDK releases didn’t democratise development—they outsourced QA, level design, and IP expansion at zero marginal cost.

  4. 5:04
    The Myth of Internal Genius

    Half-Life’s expansions were outsourced; Team Fortress was acquired from modders; Dota 2 emerged from community stewardship—not internal strategy.

  5. 6:12
    Hardware as Infrastructure, Not Product

    Hardware (Index, Steam Deck) serves Steam’s lock-in—not profit. There is no evidence it contributes meaningfully to margins.

  6. 7:52
    The Platform Subsidy

    Valve’s model only functions because Steam’s network effects absorb the organisational friction of flat management.

Worth your time?

Yes. Study the whole thing.

3.5/ 5
What works
  • Steam’s commission model creates durable cash flow.
  • Mod SDKs convert user labour into low-cost content expansion.
  • Decentralised iteration works for narrative FPS and puzzle design—when backed by massive playtest infrastructure.
What does not
  • Valve’s flat structure scales beyond ~400 people.
  • Its hardware division is profitable.
  • Its acquisition strategy consistently yields IP control.
Study it if
  • Students of platform economics.
  • Game studios evaluating distribution leverage.
  • Managers assessing trade-offs of hierarchy vs autonomy.
Skip it if
  • Startups seeking organisational templates.
  • Investors assessing repeatable growth models.
  • Policy makers treating Valve as proof of post-corporate viability.
The written brief1 min read

What the company or idea is

Valve is an American video game developer, publisher, hardware maker, and digital distribution company founded in 1996 by Gabe Newell and Mike Harrington. It operates without managers, titles, or hierarchy.

How it actually makes money

Valve makes money almost entirely through Steam: a digital storefront that takes a 30% commission on third-party game sales and sells Valve’s own games. It does not license its engine to external developers for revenue. Hardware (Index, Steam Deck) is sold near cost, with no public evidence of profitability.

What works

Steam’s dominance gives Valve pricing power, discovery control, and data leverage. Its iterative, playtest-driven development—applied to Half-Life, Portal, and Dota 2—produced genre-defining titles. Releasing SDKs (e.g., GoldSrc) turned users into unpaid R&D.

What does not

Valve’s flat structure causes chronic delays, project cancellations, and inconsistent output. It has no formal product roadmap, no release discipline, and no accountability mechanism—leading to multi-year gaps between major titles and repeated legal challenges over consumer rights and antitrust.

What to take from it

Valve proves that a platform can subsidise creative risk—but only after achieving near-monopoly scale in distribution. Its internal process works because Steam locks in users and developers, not because open allocation is inherently effective.

Is it worth your time

Yes—if you are studying platform economics, self-publishing leverage, or the trade-offs of flat management in creative industries. No—if you assume its structure is replicable or that its success validates decentralisation as a general business model.

Same desk · Companies4 of 325
10:54
ShopifyTobias Lütke · 2006Shopify is a Canadian multinational cloud e-commerce management platform for retail point-of-sale systems, founded in 2006 by Tobias Lütke, Daniel Weinand, and Scott Lake. In 2024, it processed US$292.3 billion in transactions, with 5 million customers. Its software is praised for ease of use and reasonable fee structure, and it is described as the 'go-to e-commerce platform for startups'. Shopify went public in 2015 and uses a two-class voting structure that grants disproportionate voting control to Lütke despite his minority economic stake.
11:48
NextEra Energy1925NextEra Energy is not a clean-tech startup. It is a century-old utility conglomerate whose dominance rests on regulatory capture, tax policy, and the strategic co-location of renewables and fossil generation—not their replacement.
12:18
ICA Gruppen1938ICA Gruppen is a Swedish food-and-health retailer cooperative, founded in 1938, operating ~1,300 stores in Sweden and the Baltic countries, and owning a bank, pharmacy chain, and real estate division.
11:06
Sinopharm (company)1998Sinopharm is a Chinese state-owned enterprise founded on November 26, 1998. It developed the BBIBP-CorV inactivated virus COVID-19 vaccine, which showed 78.1% efficacy against symptomatic disease and 100% against severe disease in peer-reviewed Phase III trials. In 2009, it merged with China National Biotec Group. Its subsidiary Wuhan Institute of Biological Products was fined in November 2017 for selling 400,520 ineffective DPT vaccines.
Up next in Business

Warner Bros.

Warner Brothers · 11:02

Warner Bros. is not a studio — it’s a trademarked revenue conduit that outlives every company that licenses it.

Play nextBack to Companies
11:02