You Don’t Need to Own the Stack

Technology companies often respond to strategic dependence the same way:

own more.

Own the distribution.

Own the identity system.

Own the store.

Own the social layer.

Own the infrastructure.

Own the content.

Own the community.

Own the customer relationship.

There is logic behind this.

Ownership can reduce dependency. It can preserve margins. It can provide access to data, control over product decisions, and leverage against other platforms.

But there is another way to create strategic power.

Instead of owning every layer, a company can build an environment in which independent layers work better together.

That is a different architecture.

And it requires a different definition of platform strength.

This distinction matters because modern gaming already contains extraordinary companies doing things well.

Epic does not need Xbox to teach it how to operate Fortnite.

Discord does not need Xbox to teach it how to build communities.

Roblox does not need Xbox to teach it how to operate a creator economy.

YouTube does not need Xbox to teach it how to distribute video.

Publishers do not need another company to redefine their brands, games, communities, or creative identities.

The opportunity is not to replace these systems.

It is to make them more valuable when they participate together.

A shared platform does not require every participant to become part of the platform owner.

That distinction is fundamental.

A publisher can retain its brand.

A developer can retain creative control.

A community can retain its identity.

A creator can retain an audience.

A service can retain its product.

A game can retain its economy.

Participation does not have to mean assimilation.

This is already visible in platform economics.

Industry platforms derive part of their value from products and services created by outside firms rather than from vertically integrating every complement themselves.

The point of the platform is not necessarily to replace the participant.

It can be to provide the conditions in which participation becomes more valuable.

That suggests an architectural principle:

The pieces remain recognizable.

The board provides common systems around them.

The resulting environment gains value because the pieces can interact without losing what made them valuable in the first place.

Steam Workshop provides a simple version of this idea.

Valve supplies common infrastructure for user-generated content: hosting, discovery, subscriptions, updates, community interaction, and distribution.

But individual games determine what Workshop content means inside their own products.

Valve does not need to redesign every game.

The game does not need to rebuild the entire distribution layer for modifications.

Shared infrastructure sits underneath local authority.

Steamworks more broadly follows the same pattern. Valve provides common systems around authentication, cloud storage, matchmaking, commerce, achievements, community features, inventory, Workshop integration, and other capabilities while third-party developers retain control of the games using them.

That is not full sovereignty. Every platform still imposes rules and technical constraints.

But the pattern matters.

That is more powerful than simply asking every developer to rebuild the same systems independently.

Discord provides another example.

Its Social SDK allows developers to integrate Discord-powered relationships and communication directly into games.

A player can connect a Discord account, access friends, publish rich presence, send invitations, and—with appropriate authorization—use text and voice communication from within an integrated game. Discord also supports provisional accounts so developers can use the same social infrastructure even for players who have not linked an existing Discord account.

That is strategically important because Discord does not have to disappear for the integration to work.

The game can retain its interface.

Discord retains its social graph.

The player gains continuity between them.

No participant has to recreate the other.

Discord currently reports higher median retention, more game-launch days, and longer sessions among linked players across a study of more than 15 integrated games, though those results are company-reported and should be treated accordingly.

The architectural lesson is more important than the percentages.

That is integration without absorption.

And it offers a very different model for Xbox.

Before asking publishers to participate in a larger board, it is worth taking their current behavior seriously.

Publishers build launchers, account systems, stores, subscriptions, communities, and proprietary services because these systems can provide valuable things.

Direct customer relationships.

Control over distribution.

Better economics.

First-party promotion.

Identity.

Data.

Subscriptions.

Account persistence.

Customer support.

Strategic independence.

Reduced dependence on another gatekeeper.

Those are not mistakes.

A publisher running its own platform may decide that the control gained is worth the cost of operating it.

The board therefore cannot succeed by telling publishers:

You do not need this.

It has to demonstrate something harder:

That changes the competitive question.

The board is not competing merely against another storefront.

It is competing against the value of vertical integration itself.

Running a proprietary platform requires more than building a launcher.

It can require:

identity,

authentication,

payments,

fraud prevention,

account recovery,

global commerce,

parental controls,

distribution,

updates,

social systems,

moderation,

customer support,

analytics,

compliance,

privacy controls,

infrastructure,

security,

and continuous maintenance.

Not every publisher operates all of these systems independently.

But the more of the customer relationship a publisher wants to control, the more platform responsibility it can assume.

That creates a form of duplicated cost across the industry.

Multiple companies solve versions of the same problems because each wants ownership of the surrounding layer.

Sometimes that duplication is worth it.

Sometimes it is not.

Bethesda provides a useful precedent.

In 2022, Bethesda retired the Bethesda.net Launcher and migrated users' game libraries to Steam. Players could transfer their Bethesda libraries and continue accessing their games through Valve's platform rather than through Bethesda's proprietary launcher.

That does not prove proprietary launchers are economically irrational.

Bethesda's decision occurred under its own circumstances.

But it demonstrates something important:

a publisher can decide that continuing to operate an independent distribution layer is no longer worth the value that layer provides.

That is the strategic threshold a board should attempt to cross.

Not through coercion.

Through comparative advantage.

This creates a useful concept:

launcher substitution value.

The board creates launcher substitution value when participating in the shared environment gives a publisher enough infrastructure, distribution, audience, economics, and control that maintaining a parallel proprietary layer produces diminishing additional benefit.

This is not simply a lower store fee.

It is a total operating equation.

A publisher has to compare:

the value of independence

against

the cost of operating independence

and then compare both against:

the value of participation on the board.

That includes direct economics.

But it also includes:

development expense,

support expense,

fraud exposure,

compliance expense,

technical maintenance,

discovery,

audience acquisition,

payments,

social infrastructure,

and the opportunity cost of engineering teams maintaining commodity systems rather than differentiated products.

The board becomes compelling when the answer becomes:

That is a much stronger proposition than "put your game in our store."

A platform cannot ask publishers to surrender direct economics while claiming to reduce their dependence.

The financial proposition has to be competitive.

Epic illustrates how high that bar has become.

As of 2026, Epic Games Store lets developers retain 100 percent of the first $1 million in annual net revenue per product, after which its standard 88/12 developer-platform split applies. Epic also markets global distribution, more than 80 payment methods, dozens of regional currencies, and access to a large existing user base.

That means a future board cannot rely on the old argument that 30 percent is the unavoidable price of distribution.

Publishers already have alternatives.

The board therefore needs a more sophisticated economic model.

One possibility is to separate infrastructure access from transaction fees.

Smaller developers may prefer a conventional revenue share because it minimizes fixed cost.

Larger publishers may prefer a subscription or negotiated infrastructure contract in exchange for lower marginal transaction fees.

A strategic participant may want direct control over portions of its commerce while paying the board for the systems it actually uses.

The point is not that one model is correct.

The point is that:

That is particularly important if the board wants major publishers and platform operators to participate deeply rather than reluctantly.

Shared infrastructure creates another source of value.

Governance.

Platforms already spend enormous amounts of money solving problems that do not differentiate their games:

fraud,

age-appropriate access,

parental controls,

identity,

moderation tooling,

payment safety,

privacy,

account recovery,

marketplace abuse,

security,

and regulatory compliance.

These responsibilities do not disappear when infrastructure becomes shared.

But some of the work can potentially be standardized.

The board can provide:

common age systems,

permission frameworks,

parental controls,

transaction standards,

fraud detection,

moderation interfaces,

appeal systems,

identity layers,

compliance APIs,

marketplace policies,

and shared trust infrastructure.

That changes governance from pure overhead into part of the board's value proposition.

The board does not remove responsibility from publishers.

It can make responsible operation easier.

That is a different claim.

And an important one.

Shared governance also creates a danger.

If the board becomes involved everywhere, does it become responsible for everything?

Not necessarily.

A more useful principle is:

If the board operates the payment layer, it owns responsibility for the payment systems it controls.

If it operates shared identity, it owns responsibility for that identity layer.

If it defines platform-wide marketplace rules, it owns responsibility for those rules.

But a developer that controls its game's internal content, economy, design, or community decisions retains responsibility for those systems.

The boundaries will never be perfectly clean.

Platforms influence participants through interfaces, incentives, rules, access, and technical standards. A systematic review of 103 platform-governance studies describes governance precisely as the mechanisms platform owners use to influence complementors and users and sustain an ecosystem.

So the board cannot pretend to be neutral infrastructure while exercising meaningful control.

But neither must every responsibility collapse upward into the board.

This leads to a stronger architecture:

The board governs the relationships it creates.

The pieces govern the domains they control.

A well-designed board should create incentives through architecture.

Not merely policy documents.

If safe payments are easier to implement than unsafe ones, more participants will use them.

If parental controls are available through common infrastructure, developers do not need to rebuild them from zero.

If identity permissions are standardized, developers have less reason to invent inconsistent access systems.

If fraud protection is available at the board level, smaller publishers gain capabilities that may otherwise be expensive to operate independently.

The goal is not to eliminate bad behavior through architecture.

That is unrealistic.

The goal is to make responsible participation the path of least resistance.

That creates value for players.

It creates value for publishers.

And it protects the board itself.

The stronger the shared environment becomes, the more its trustworthiness becomes part of the economic product.

Platform owners often treat interoperability as something they are forced to provide.

That may be too narrow.

Interoperability can also be a growth mechanism.

The European Union's Digital Markets Act now requires designated gatekeepers to provide certain forms of access, data portability, and interoperability. The Commission explicitly connects those requirements to network effects, user lock-in, business-user dependence, and the competitive power of large platforms.

The policy debate around those rules is separate from the architectural point.

The important signal is that data portability, access, and interoperability have become central questions in modern platform design.

A future board can treat those capabilities as strategic advantages rather than regulatory burdens.

A participant should be able to know:

what data it controls,

what data the board controls,

how identity travels,

how relationships are represented,

how commerce works,

what happens if it leaves,

and what remains portable.

That matters because participants will hesitate to build deeply on infrastructure they believe can later trap them.

Openness can therefore create trust.

But complete openness can also undermine coherence, safety, and operating discipline.

The design problem is not:

open or closed.

It is:

That is the real governance question.

The board's success creates its own danger.

The more participants depend on shared infrastructure, the more leverage the board acquires.

That leverage can be abused.

Fees can rise.

Rules can change.

Discovery can favor first-party products.

Data can become difficult to access.

Participants can become dependent on interfaces they do not control.

Portability can weaken.

Appeals can become opaque.

The platform can become indispensable and then begin extracting more of the value it originally helped create.

That is not a hypothetical weakness unique to Xbox.

It is a recurring problem of platform power.

Research on platform governance consistently treats value creation and value appropriation as separate governance problems. A platform has to help participants create value while also deciding how that value is divided.

Those interests will not always align automatically.

So a board built around participant sovereignty needs protections against the board itself.

Potential principles include:

transparent economics,

predictable rules,

clear data rights,

portability,

documented interfaces,

reasonable exit paths,

consistent enforcement,

bounded authority,

and meaningful dispute processes.

The architecture cannot depend on:

Trust us.

The architecture should make trust easier to justify.

This gives the board a simple test.

Does participation weaken the participant?

Or strengthen it?

If a publisher joins and loses its brand, direct relationships, economic flexibility, data access, and ability to operate independently, then the board has not created partnership.

It has created dependency.

If a creator gains distribution but becomes unable to reach the audience outside the platform, the board has accumulated power by weakening the creator.

If a community gains tools but loses ownership of its identity, the value exchange may be structurally unstable.

The healthier model is different.

A publisher joins because it gains access to capabilities that would be inefficient to reproduce alone.

A developer reaches more players.

A creator reaches more communities.

A service gains more useful integrations.

A player experiences greater continuity.

The participant retains the things that make it distinct.

And the board becomes more valuable because the participant became stronger.

That is the difference between lock-in and earned retention.

This changes how Xbox should compete.

Against PlayStation, the answer does not have to be:

own more exclusives.

Against Steam, the answer does not have to be:

build another PC store.

Those strategies compete over pieces.

The larger opportunity is to become the platform publishers and developers most want to operate on.

That means giving them:

strong economics,

shared infrastructure,

large audiences,

identity systems,

global commerce,

social integration,

discovery,

governance tooling,

reduced operating burden,

and meaningful control over their own products.

The board becomes the preferred operating environment.

Not because participants have nowhere else to go.

Because going somewhere else requires rebuilding things they no longer need to rebuild.

That is a much stronger competitive moat.

A publisher does not remain because leaving is impossible.

It remains because participation keeps creating more value than independence.

The other side of the architecture is cultural.

Publishers will not come simply because the infrastructure is elegant.

They will come because players are there.

Players will not remain simply because the economics work for publishers.

They will remain because the environment is useful to them.

This is where the first two essays converge.

Gaming culture already exists across:

games,

Discord,

Twitch,

YouTube,

mods,

communities,

creators,

events,

platform networks,

stores,

competitive scenes,

and countless independent spaces.

Xbox does not need to replace them.

It needs to create an environment in which engaging with them becomes more continuous.

The best place to discover what friends are playing.

The best place to encounter new communities.

The best place to move from watching something to participating in it.

The best place for identity and history to survive across games.

The best place to participate in gaming culture without requiring every part of that culture to become Microsoft-owned.

That is the other half of the flywheel.

Better participants attract players.

More players attract better participants.

More cultural activity makes the environment more valuable.

More valuable infrastructure makes participation easier.

The board compounds both sides.

The strategic outcome should not be forced consolidation.

It should be voluntary redundancy.

A company should stop operating a parallel platform layer when that layer no longer creates enough additional value to justify its cost.

Bethesda's migration away from its launcher and toward Steam demonstrates that publishers can reach that conclusion under real-world conditions.

The more ambitious question is what would cause much larger operators to make a similar calculation.

For Epic, Roblox, or another major ecosystem, the board would have to offer far more than distribution.

It would have to preserve:

identity,

economics,

brand,

community,

data rights,

and strategic autonomy.

While improving:

reach,

infrastructure efficiency,

discovery,

interoperability,

commerce,

and participation.

That is an enormous standard.

It should be.

Because the board only becomes important if strong participants choose it even when they could operate independently.

That is not acquisition.

It is not forced consolidation.

It is platform competition.

This is where Xbox Forever becomes a specific architectural proposal.

Not a giant application containing every function.

Not a replacement for Discord.

Not a replacement for Roblox.

Not a replacement for Epic.

Not a replacement for creators, communities, publishers, or developers.

A board.

A shared environment connecting them.

Publishers retain identity.

Developers retain products.

Communities retain culture.

Creators retain audiences.

Services retain specialization.

Xbox provides the connective infrastructure that makes participation across those systems more continuous.

Identity.

Presence.

Commerce.

Discovery.

Interoperability.

Trust.

Relationships.

History.

Governance standards.

Shared services.

Distribution.

And access across console, PC, cloud, mobile, handhelds, and future endpoints.

The platform succeeds when the things participating on it become more successful.

That creates the governing economic principle:

And it creates the governing architectural principle:

Everything else should remain as sovereign as the system can responsibly allow.

That is what makes the architecture scalable.

Xbox does not have to become every future service.

It has to become the place where future services want to participate.

Vertical integration will remain valuable.

Some things should be owned.

Some systems require deep control.

Some products benefit from complete integration.

The argument is not that ownership is obsolete.

It is that ownership is not the only source of platform power.

A company can own fewer pieces while coordinating a larger ecosystem.

It can capture value from infrastructure without capturing ownership of every participant.

It can become more important by making independent companies more capable.

It can reduce duplicated cost without eliminating independent identity.

It can create standards without demanding sameness.

It can connect cultures without owning culture.

That is a different kind of platform.

One that competes by becoming useful rather than unavoidable.

One where independence and participation do not have to be opposites.

One where the board becomes stronger because the pieces remain strong.