Why Eventually

Why Naver Will Eventually Disappear

Naver played an important role in the development of the Korean internet.

That should be acknowledged.

It helped millions of Koreans discover information online, built services suited to the Korean language, created an enormous domestic digital ecosystem, and became one of the companies that defined the early Korean internet.

But historical contribution does not guarantee permanent relevance.

In fact, I believe the opposite may now be happening.

The very structure that made Naver dominant in Korea may eventually become the reason it declines.

My argument is not simply that Naver is an old company, that a new competitor will appear, or that artificial intelligence will suddenly destroy every search engine.

The problem is deeper.

Naver represents a particular model of the internet:

a closed platform that attempts to capture information, users, merchants, media, advertising, payments, and transactions inside its own ecosystem.

That model was extraordinarily powerful during the portal era.

I believe it is becoming increasingly incompatible with the direction in which the internet is moving.

And over the long run, systems that continuously extract value from an ecosystem without allowing enough value to circulate back into that ecosystem eventually become vulnerable.

Naver may be approaching that point.

1. The Original Sin: Turning the Open Web Into a Walled Garden

The internet was built around a remarkably simple idea.

Documents could link to other documents.

A website did not need permission from a central authority to exist.

Information could move across domains.

Publishers could build their own audiences.

Users could travel through hyperlinks from one independent source to another.

This architecture created the open web.

Naver developed in a very different direction.

Instead of merely helping users navigate the web, it gradually built an environment in which enormous amounts of Korean internet activity happened inside Naver itself.

Blogs.

Cafes.

Knowledge Q&A.

Shopping.

Maps.

News.

Reviews.

Reservations.

Payments.

Advertising.

Local business discovery.

Instead of sending users outward, the economic incentive was increasingly to keep them inside.

From the perspective of a platform company, this makes perfect sense.

More time inside the platform means more data.

More data means better targeting.

More transactions mean more commissions.

More page views mean more advertising inventory.

More merchants inside the system create more dependency.

More dependency strengthens the platform.

The problem is that what is rational for the platform can be unhealthy for the broader information ecosystem.

A search engine should ideally help people discover the best information available.

A portal has a different incentive.

It benefits from helping people discover the best information that also keeps them inside the portal.

Those two objectives are not always the same.

That difference matters enormously.

2. Search Gradually Stops Being Search

When people use a search engine, they assume they are asking a relatively straightforward question:

What is the most useful information related to what I typed?

But once a company simultaneously operates the search engine, advertising marketplace, shopping platform, payment system, booking system, blogging platform, map service, review ecosystem, and merchant marketplace, that question becomes more complicated.

The platform is no longer merely organizing information.

It is organizing an economy in which it is also a participant.

Imagine walking into a shopping mall where the building owner also owns several stores, controls the directory, decides which stores appear first on every sign, operates the payment system, sells advertisements to the tenants, collects customer behavior data, and controls access to the largest stream of visitors.

Even if every individual decision could be defended technically, the structural conflict is obvious.

The intermediary has enormous power.

And once businesses become dependent on that intermediary, their behavior changes.

They stop asking:

How do we create the best product?

They begin asking:

How do we rank higher on Naver?

That distinction may sound small.

It is not.

It changes an entire economy.

3. When Businesses Optimize for Algorithms Instead of Customers

One of the most damaging consequences of portal dominance is the enormous amount of economic activity spent trying to manipulate visibility.

Businesses write content not because the content deserves to exist, but because the algorithm rewards it.

Bloggers repeat keywords.

Agencies manufacture posts.

Merchants purchase advertisements.

Restaurants ask customers for reviews.

Companies hire people whose job is essentially to understand how to appear higher inside a proprietary ranking system.

SEO exists everywhere, of course.

Google has created its own enormous SEO industry.

But Korea developed an especially concentrated version of this phenomenon because a very large portion of discovery, local commerce, blogging, shopping, and information consumption became concentrated inside a handful of portal-controlled surfaces.

The result is familiar to almost every Korean internet user.

You search for something simple.

Instead of immediately finding genuine expertise, you encounter promotional blog posts.

Sponsored content disguised as experience.

Affiliate-style recommendations.

Repetitive articles.

Keyword-stuffed pages.

Commercial reviews.

Content created primarily because somebody wanted exposure.

You begin scrolling.

And scrolling.

And scrolling.

The cost is difficult to measure because each individual search wastes only a few seconds or minutes.

But multiply that friction across tens of millions of people over many years.

The social cost becomes enormous.

Bad information systems do not merely annoy people.

They waste human attention.

And human attention is one of the most valuable resources in an advanced economy.

4. Information Pollution Is an Economic Problem

When search quality declines, people often treat it as a minor user-experience problem.

It is much more serious than that.

Imagine an entrepreneur researching a market.

A student studying a technical subject.

A patient looking for general information before visiting a doctor.

A consumer comparing products.

A software engineer debugging a problem.

A small business owner researching government policy.

Every one of these activities depends on information retrieval.

If high-quality information is buried beneath commercial optimization, the cost of obtaining truth increases.

That cost propagates through the economy.

People make worse decisions.

They spend longer researching.

They rely more heavily on intermediaries.

Experts become harder to discover.

Original sources receive less attention.

Low-quality summaries outperform high-quality primary material simply because they understand the ranking system better.

Eventually the system rewards not the person who knows the most, but the person who understands distribution mechanics the best.

That is a dangerous equilibrium.

A healthy information economy should reward knowledge creation.

A distorted information economy rewards visibility engineering.

5. The Portalization of Journalism

The same structural problem appears in news.

When a portal becomes the main gateway through which millions of people encounter journalism, newspapers are no longer competing only to produce journalism.

They are competing for placement, clicks, attention, and algorithmic distribution inside the portal.

This changes editorial incentives.

A reporter may want to write a careful investigation.

But an organization operating under intense traffic pressure also has to ask:

Will people click?

Will the headline perform?

Will the story trend?

Will the portal distribute it?

Will it generate enough page views?

This is not a uniquely Korean problem.

Facebook, Google, X, YouTube, and other global platforms have created similar tensions.

But Korea’s portal-centered media structure made the relationship unusually visible.

News organizations increasingly became suppliers of content to a distribution platform they did not control.

And whoever controls distribution possesses enormous power, even if that power is exercised indirectly.

This produces a fundamental imbalance.

The institution creating the journalism bears the cost.

The platform controlling discovery controls the audience.

When distribution becomes more valuable than production, journalism becomes economically weaker.

And weak journalism becomes more vulnerable to sensationalism.

Shorter news cycles.

More provocative headlines.

More celebrity coverage.

More political outrage.

More stories generated for clicks rather than public value.

The platform does not necessarily need to order journalists to produce bad journalism.

The incentive structure can do the work by itself.

That is precisely why structural power matters.

6. A Ranking Algorithm Is Never Merely Technical

People often speak about algorithms as though they were neutral machines.

They are not.

Every ranking system reflects decisions.

What signals matter?

Freshness?

Authority?

Engagement?

Commercial relevance?

User history?

Platform participation?

Popularity?

Payment?

Location?

Recency?

The moment a system decides what appears first, it influences attention.

And attention influences reality.

A restaurant appearing at the top of local search receives more customers.

A news article appearing prominently receives more readers.

A product receiving more visibility receives more purchases.

A blog post ranking first receives more authority.

The algorithm does not merely describe popularity.

It can create popularity.

That means concentrated control over discovery becomes a form of social power.

The larger the platform becomes, the more consequential that power becomes.

And the less transparent the system is, the harder it becomes for outsiders to understand why certain information is visible while other information disappears.

Again, this does not require a conspiracy.

It is a property of centralized systems.

If one company controls an enormous share of digital discovery, even ordinary product decisions can have extraordinary social consequences.

7. The Small-Business Dependency Machine

The same pattern extends into commerce.

A small restaurant wants customers.

It needs map visibility.

Reviews.

Reservations.

Search exposure.

Perhaps advertisements.

Perhaps payment integration.

Perhaps a storefront.

Perhaps delivery or commerce tools.

Each individual service may be useful.

That is precisely why the system becomes powerful.

The merchant gradually becomes dependent on the platform’s infrastructure.

At first the platform helps the business.

Eventually the business may discover that participating is no longer optional.

Once enough customers begin their purchasing journey inside the platform, merchants must follow the customers.

That creates platform power.

A small merchant negotiating with a dominant platform does not possess equal leverage.

The platform can change ranking rules.

Advertising systems.

Commission structures.

Product interfaces.

Exposure mechanisms.

Review policies.

Search layouts.

The merchant must adapt.

This is the economic logic of digital toll roads.

The road becomes essential.

And once it is essential, whoever controls the road can collect increasing amounts of value from everyone traveling through it.

8. Self-Preferencing Is a Structural Problem

This becomes even more problematic when the platform is not merely the road.

It also operates businesses traveling on that road.

Search.

Shopping.

Payments.

Reservations.

Maps.

Content.

Advertising.

Commerce.

The deeper the vertical integration becomes, the more difficult it becomes to separate the interests of users from the interests of the platform.

Even without assuming malicious intent, the incentives are obvious.

If a search engine sends a user to an independent website, the platform may receive relatively little economic value.

If the user instead clicks a platform-hosted product, pays through the platform, books through the platform, leaves a review on the platform, and returns through the platform, the company captures much more of the transaction.

Therefore the economic architecture naturally pushes toward enclosure.

And enclosure produces dependency.

This is why debates about self-preferencing cannot be reduced to individual search results.

The deeper question is:

Should the company controlling discovery also be allowed to become the dominant participant in everything being discovered?

That is not merely a competition-policy question.

It is a question about the architecture of digital society.

9. Naver’s Greatest Strength Became Its Greatest Weakness

For many years, Naver’s closed ecosystem was an enormous competitive advantage.

Korean-language content was concentrated there.

Users were there.

Businesses were there.

Bloggers were there.

Advertisers were there.

News was there.

The network reinforced itself.

Users attracted businesses.

Businesses generated content.

Content attracted more users.

More users attracted advertisers.

Advertising funded more services.

The flywheel worked.

But platforms often misunderstand something when they become dominant.

A moat can become a prison.

The same architecture that prevents users from leaving can prevent the company from adapting to a new paradigm.

The internet is now changing from a world organized around pages and portals to a world increasingly organized around agents, models, APIs, direct answers, and machine-mediated actions.

That transition threatens the basic economics of the traditional portal.

10. AI Changes the Fundamental Question

Traditional search asks:

Where can I find the answer?

AI increasingly asks:

What is the answer?

Traditional commerce asks:

Which website should I visit?

AI agents increasingly ask:

What should I buy, and can you complete the transaction for me?

Traditional travel search asks:

Which booking service should I use?

An agent can potentially compare options across multiple services and execute the booking.

Traditional local search asks:

Which restaurants appear in this portal?

An AI system can potentially aggregate information from numerous sources and generate a personalized recommendation.

This is a profound change.

The portal model depends heavily on intermediate navigation.

Search.

Click.

Browse.

Compare.

Click again.

Open another page.

See advertisements.

Return.

Search again.

An intelligent agent tries to eliminate those steps.

That is extremely convenient for users.

And potentially devastating for businesses whose economics depend on controlling those intermediate steps.

11. The Better AI Becomes, the Less Valuable the Portal Becomes

Imagine a future interaction.

You do not open Naver and search:

“Best accounting software for a Korean manufacturing company with 30 employees.”

You simply tell your AI:

Find three appropriate options, compare their pricing and integrations, check independent reviews, and recommend one based on our current accounting workflow.

The agent performs the research.

Perhaps it contacts vendors directly.

Perhaps it asks clarifying questions.

Perhaps it schedules demos.

Perhaps it negotiates pricing.

Perhaps it configures the software.

Where is the portal in that process?

It may not exist.

This is the strategic threat.

AI does not merely create another search competitor.

It can eliminate part of the behavior we currently call “search.”

And companies that historically profited from controlling information navigation may struggle when navigation itself becomes unnecessary.

12. The Open Internet Has a Second Chance

Ironically, AI may also weaken the walled-garden strategy.

AI systems are more useful when they can access broad, structured, interoperable information.

APIs matter.

Machine-readable data matters.

Open standards matter.

Independent databases matter.

Public documents matter.

Direct merchant information matters.

Structured product information matters.

A closed ecosystem optimized primarily for human users clicking through proprietary interfaces can become less attractive in a world where software agents need to interact directly with information.

The future internet may therefore become less portal-centric.

Not necessarily more decentralized in every respect.

AI itself can create new concentrations of power.

But the center of gravity can shift.

The dominant interface may no longer be a Korean portal homepage containing dozens of boxes.

It may be a conversational layer sitting above the internet.

And if that happens, owning the boxes matters much less.

13. Younger Users Already Understand This Intuitively

There is another danger for traditional portals.

Habits change.

Previous generations learned to begin the internet from a portal.

You opened Naver.

Then you decided where to go.

Younger users increasingly begin from the service that directly satisfies their intent.

YouTube for video.

Instagram or TikTok for discovery.

Google for certain information.

Coupang for products.

Maps for locations.

ChatGPT or other AI systems for questions and research.

Specialized communities for expertise.

They do not necessarily need a universal front door.

This fragmentation weakens the portal concept.

And AI pushes the fragmentation one step further.

Instead of choosing between ten specialized applications, an AI layer may orchestrate them automatically.

The portal loses another layer of relevance.

14. The Real Problem Is Trust

Technology alone will not determine Naver’s future.

Trust may matter more.

People tolerate closed platforms when those platforms provide overwhelming convenience.

But once users begin believing that search results are heavily commercialized, that content is optimized rather than authentic, that reviews are unreliable, or that useful information is buried beneath promotional material, the platform begins consuming trust.

This can continue for a surprisingly long time.

Network effects are powerful.

People use systems they dislike because everyone else uses them.

Businesses remain because customers remain.

Customers remain because businesses remain.

This creates the illusion of permanent dominance.

But network effects can reverse.

Once enough users leave, businesses follow.

Once businesses diversify, users have fewer reasons to stay.

Once creators stop treating the platform as essential, content quality weakens.

Then the flywheel begins rotating backward.

History contains many companies that looked immortal shortly before their decline.

Dominance is often strongest immediately before the underlying paradigm changes.

15. Korea Does Not Need Another Naver

The lesson should not be that Korea needs a different company to build the same structure.

Replacing Naver with another closed super-platform would reproduce the same problem.

What Korea needs is a healthier digital architecture.

Businesses should be able to own their customer relationships.

Creators should be able to own their audiences.

Publishers should be able to distribute information without surrendering their entire economic relationship to an intermediary.

Search systems should compete on usefulness.

Payment systems should interoperate.

Commerce should remain contestable.

Information should be linkable.

Data should be portable where appropriate.

Independent websites should matter again.

Platforms should create value by making the ecosystem more productive, not merely by positioning themselves between every participant in the ecosystem.

That distinction will become increasingly important in the AI era.

16. A Platform Should Be Judged by What Grows Around It

There is a simple way to evaluate a technology platform.

Ask:

Does the ecosystem around it become stronger as the platform becomes stronger?

A great platform creates new companies.

New developers.

New industries.

New creators.

New technologies.

New markets.

It lowers the cost of entrepreneurship.

It increases access to information.

It gives independent participants more leverage.

The best infrastructure allows others to become powerful.

A weaker model does the opposite.

As the platform becomes stronger, everyone around it becomes increasingly dependent on it.

Businesses pay more for visibility.

Creators optimize for algorithms.

Media companies chase traffic.

Users encounter more commercialized information.

Independent distribution weakens.

The platform captures more value while the surrounding ecosystem loses autonomy.

That may generate excellent financial results for years.

But it is not necessarily a healthy technological civilization.

17. This Is Why I Think Naver Can Disappear

When I say that I believe Naver will disappear, I do not necessarily mean that the legal entity will suddenly cease to exist.

Large companies rarely vanish overnight.

They shrink.

They transform.

They split.

They become infrastructure providers.

They retreat into profitable niches.

They remain financially significant while becoming culturally irrelevant.

Yahoo still exists.

AOL still exists in various forms.

Many once-dominant technology companies technically survived long after they stopped defining the future.

That is the distinction.

The question is not:

Will Naver’s corporate registration disappear?

The question is:

Will Naver remain the default architecture through which Koreans experience the internet?

I believe the answer will eventually be no.

The portal era is ending.

AI accelerates that ending.

Direct relationships between businesses and customers will become more important.

Agents will reduce the need for navigation.

Global services will increasingly localize.

Language barriers will collapse.

Information retrieval will become conversational.

Transactions will become automated.

Interfaces will become invisible.

In that environment, the strategic value of owning a giant centralized portal declines.

18. The Irony

There is an irony in all of this.

Naver became powerful because it adapted the internet to Korea.

Its decline may come because Korea no longer needs a Korean wall around the internet.

Machine translation is improving.

AI understands Korean.

Global software can increasingly operate across languages.

Korean companies can sell internationally.

Korean users can access global information without the same linguistic friction that existed twenty years ago.

The economic justification for a heavily enclosed domestic internet ecosystem therefore becomes weaker every year.

The wall that once protected Naver from the world can eventually prevent Naver from participating fully in the world.

19. Creative Destruction Is Not a Tragedy

If Naver loses its dominant position, I do not think that would necessarily be a tragedy.

Companies are not countries.

Companies are not permanent social institutions.

They are organizational technologies.

They exist because, for some period of time, they solve problems better than available alternatives.

When better systems emerge, capital, talent, users, and attention should move.

That is creative destruction.

Naver itself once benefited from that process.

It challenged earlier ways of navigating the internet.

Someday another architecture will challenge Naver.

That is healthy.

What would be unhealthy is preserving an outdated system simply because it became large.

20. The Future Belongs to Systems That Create More Value Than They Capture

Ultimately, this is not really an essay about Naver.

It is about a broader principle.

Every platform faces a choice.

It can create an ecosystem.

Or it can extract from one.

Of course every company must capture some value.

A company that captures nothing cannot survive.

But there is a difference between capturing a reasonable portion of newly created value and designing an entire ecosystem around dependency.

The greatest technology companies of the next era will understand this.

They will make creation cheaper.

They will make knowledge easier to access.

They will make transactions easier.

They will give individuals and small organizations capabilities that previously belonged only to large institutions.

They will reduce friction.

They will increase autonomy.

They will allow value to move.

And then they will take a small fraction of an enormous amount of newly created value.

That is a fundamentally different philosophy from building tollbooths everywhere.

The internet is moving again.

AI is accelerating the movement.

Search is changing.

Software is changing.

Media is changing.

Commerce is changing.

The cost of producing technology is collapsing.

Language barriers are collapsing.

Distribution is being reorganized.

The old Korean portal structure was built for a different technological era.

Naver was one of the greatest beneficiaries of that era.

It may also become one of its greatest casualties.

That is why I believe Naver, at least in the form we recognize today, will eventually disappear.

Not because it contributed nothing.

Not because every decision it made was wrong.

And not because a company of its size can collapse tomorrow.

But because no amount of historical dominance can permanently protect a business model when the architecture of technology changes underneath it.

The portal was once the gateway to the internet.

In the coming era, we may no longer need the gate.

And when people no longer need the gate,

the gatekeeper loses its power.