AEO vs Zero-Click Search: Ancaman atau Kanal Baru

Ada ketakutan yang sangat masuk akal di dunia publisher dan marketing:

“Kalau AI menjawab semuanya di halaman hasil, siapa yang masih klik website?”

Itulah inti debat zero-click.

User mendapat jawaban tanpa membuka source.

Dulu fenomena ini sudah ada melalui:

featured snippets.

knowledge panels.

local packs.

weather.

calculator.

Sekarang generative AI membuat jawabannya lebih panjang dan lebih conversational.

Threat terasa lebih besar.

Dan memang, untuk beberapa jenis query, kemungkinan user tidak perlu click bisa meningkat.

Tetapi menyebut zero-click hanya sebagai ancaman juga terlalu sederhana.

Untuk beberapa brand, answer surface bisa menjadi kanal exposure baru.

Jadi pertanyaannya bukan:

“Zero-click bagus atau buruk?”

Lebih useful:

“Untuk query mana zero-click merusak value, dan untuk query mana visibility tanpa click tetap bernilai?”

Publisher dan brand punya jawaban berbeda.

Publisher monetisasi traffic.

Ad impression.

Subscription.

Pageview.

Zero-click bisa langsung menekan business model.

Brand service mungkin lebih fleksibel.

Kalau AI menyebut:

“Agency X cocok untuk enterprise…”

user mungkin tidak click saat itu.

Tetapi brand masuk consideration.

Later direct search.

Sales mention.

No attribution.

Still possible value.

Measurement jadi lebih sulit.

Ini bukan alasan menganggap click tidak penting.

Click tetap valuable.

Hanya journey tidak selalu linear.

Informational query paling rentan

“Apa itu AEO?”

AI bisa memberi definisi.

Selesai.

User tidak perlu article 2.000 kata.

Publisher yang hanya membuat definisi generic mungkin kehilangan click.

Solution bukan mempersulit jawaban.

Jangan hide.

Better:

provide deeper value.

Research.

Example.

Data.

Framework.

Local context.

Tool.

Zero-click memaksa publisher punya reason untuk click beyond basic fact.

Ini sebenarnya healthy.

Commodity information semakin murah.

Original information naik value.

Buyer-intent query berbeda

“Agency GEO untuk enterprise?”

AI bisa memberi shortlist.

User mungkin click untuk verify.

Case.

Pricing.

Service.

Methodology.

Security.

Zero-click di discovery bisa berubah menjadi high-intent click di verification.

AEO harus support.

Short answer outside.

Deep evidence inside.

This is two-layer publishing.

We should not intentionally withhold basic answer to force traffic.

It damages trust.

Give answer.

Earn click with depth.

Google generative Search explicitly includes supporting links

AI Overviews dan AI Mode bukan UI tanpa links.

Google menjelaskan generative responses dapat menampilkan supporting links dan bahkan set link yang lebih luas pada context tertentu.

Pada 2026, Google juga menyediakan dedicated generative AI performance reports di Search Console untuk measure impressions in AI features.

Ini berarti site owner punya visibility metric yang tidak selalu identik dengan click.

Impressions matter.

Clicks also.

AEO measurement should include both.

But don't interpret impression as successful visit.

Different.

ChatGPT Search also links sources

OpenAI says ChatGPT Search can provide timely answers with links to relevant web sources.

Publisher can track identifiable referral traffic from ChatGPT in analytics when clicks happen.

Again, two layers:

answer exposure.

referral.

Zero-click describes absence of second layer.

Not absence of first.

Should brands optimize for no-click answers?

Careful.

The goal shouldn't be "make user never visit."

The goal is "answer user correctly at each surface."

For simple factual question:
zero-click may be fine.

“Jam buka?”

“Return period?”

“Price?”

User gets value.

If action required, provide path.

Call.

Book.

Buy.

Visit.

AEO should optimize answer + action.

Not answer + forced click.

For publisher, this is harder

If answer can be fully consumed externally, revenue problem.

Publisher needs differentiated asset.

Examples:

interactive data.

original investigation.

full interview.

visualization.

database.

newsletter.

community.

tool.

personalization.

analysis.

AI summary may capture headline insight.

User clicks for experience and detail.

Content economics changes.

AEO alone doesn't solve business model.

It surfaces the challenge.

Zero-click can be a brand impression channel

Suppose AI answer says:

“According to GEO.or.id…”

User sees publisher name.

No click.

There may still be memory.

Source credibility.

Later search.

Hard to measure.

Don't invent value numbers.

Use brand lift research if available.

Direct traffic trends.

Self-reported discovery.

But attribution imperfect.

Call it unmeasured influence.

Not “free traffic.”

Language matters.

What should AEO optimize under zero-click?

Accuracy first.

If user won't click, the answer shown outside may be the only exposure.

Wrong info is more dangerous.

Answer block needs:

subject.

context.

date.

limitation.

Brand name.

Maybe source attribution.

Canonical page must be current.

Zero-click raises importance of correctness.

Second: brand identification.

If quote is used without clear entity, brand value weak.

Publisher name.

Author/source.

Distinct framework.

Not spam brand repetition.

Natural attribution.

Third: depth gap.

Ask:

What can't be compressed well?

Original dataset.

Interactive tool.

Full method.

Case details.

Community.

Provide reason to click.

Fourth: action.

If user needs buy/book/contact, make action clear.

AI agents may eventually perform some actions directly, but user still needs trusted merchant/service identity.

Transactional layer.

There is a temptation to sabotage extractability

“Don't give answer up front.”

“Hide details.”

“Use vague wording.”

So AI must send click.

Bad strategy.

User also suffers.

Search systems can choose other source.

AEO should not become anti-answer optimization.

Earn visit through value.

Not friction.

Think of zero-click as a funnel redistribution

Classic:

Search → click → learn → decide.

Generative:

Ask → learn partly → click to verify/act → decide.

Or:

Ask → decide → later direct visit.

Or:

Ask → no click.

Multiple paths.

Analytics will capture only some.

Business teams need broader measurement.

Search Console impressions.

AI referrals.

Direct.

Branded search.

Conversion source.

Self-reported discovery.

No metric alone.

Zero-click requires triangulation.

There's a scene that illustrates this.

Publisher editor:
“Traffic to definition pages turun.”

Research lead:
“Original reports?”

Editor:
“Stable.”

Research lead:
“Maybe basic answers are being commoditized.”

Then question becomes:

Do we publish 100 more definitions?

Maybe not.

Invest in original reports.

AEO strategy shifts.

For ecommerce, zero-click can be positive or negative

User sees price/product in AI surface.

No website click.

Then buys through external marketplace or agent.

Brand may still get sale.

Merchant attribution matters.

If product discovery surface leads to transaction, website session is not only KPI.

Revenue.

Offer.

Merchant.

Product feed.

This is why ecommerce AEO must talk to commerce data.

For B2B, zero-click may improve qualification

Buyer asks basic question.

Gets answer.

If not fit, doesn't click.

Good.

Fewer low-quality visits.

If fit, clicks case/security.

Traffic lower.

Lead quality maybe higher.

Don't celebrate without data.

Measure.

Session volume alone can mislead.

AEO needs business context.

Zero-click can also damage smaller publishers disproportionately

Large platform summarizes.

Small source loses visit.

This is a legitimate concern in web ecosystem economics.

AEO practitioners should not dismiss it with:

“Brand visibility is enough.”

For many publishers, it isn't.

Revenue pays reporting.

If source publishers cannot sustain work, answer ecosystem quality also suffers.

There is strategic tension.

We shouldn't pretend solved.

Publisher may need:

licensing.

subscription.

direct audience.

product diversification.

preferred source relationship.

community.

AEO is one piece.

What queries should you protect with depth?

High-value originals.

Tool pages.

Research.

Data.

Case.

Commercial decision.

Not by blocking extraction necessarily.

By adding experiences that summary cannot fully replace.

For simple FAQ, accept that zero-click may happen.

Don't overinvest in commodity content.

This is resource allocation.

How to classify?

Query type A:
simple fact.

Expected click value low.

Optimize accurate answer.

Type B:
complex explanation.

Provide summary + depth.

Type C:
decision.

Provide comparison + evidence + action.

Type D:
original asset.

Make full source uniquely valuable.

Type E:
transaction.

Ensure data/action.

Now zero-click isn't one problem.

Five different situations.

AEO success therefore shouldn't be click-only

Possible metrics:

answer visibility.

citation.

generative impression.

referral.

qualified session.

conversion.

brand query.

self-reported discovery.

For publisher, click/subscription weight high.

For brand, recommendation/conversion.

For ecommerce, product interaction/revenue.

No universal score.

One more thing: click itself can become higher-intent

If AI already summarized basics, user who clicks may be more serious.

Maybe.

Need data.

Don't assume.

Check engagement.

Conversion.

Page depth.

If AI referral sessions convert higher, interesting.

If lower, investigate.

No universal benchmark.

Zero-click debate often wants a simple side:

“AI stealing traffic.”

or

“AI creates opportunity.”

Reality is both.

Depends on business model, query, and surface.

AEO's job isn't to pick ideology.

It's to design information so the user gets a correct answer, while the business still has a reason to be visited, trusted, selected, or transacted with.

If content's entire value fits in one sentence, AI may indeed make click less necessary.

That sentence should still be correct.

Ada satu strategic shift yang perlu diterima: traffic volume mungkin tidak lagi menjadi satu-satunya proxy untuk demand.

Kalau basic answer selesai di AI surface, website bisa menerima lebih sedikit visit tetapi lebih qualified.

Itu mungkin terjadi.

Bukan asumsi.

Harus diuji.

Bandingkan cohort referral AI dengan organic biasa:

engagement.

conversion.

newsletter signup.

demo request.

revenue.

Jika 1.000 organic visits menghasilkan 5 leads dan 200 AI referrals menghasilkan 10 leads, jumlah visit saja memberi kesan salah.

Sebaliknya, kalau AI exposure naik tetapi conversion dan brand demand flat, jangan mengarang success story.

Zero-click memaksa measurement menjadi lebih dekat ke outcome.

Ada juga tension soal source sustainability.

Jika publisher terus menyediakan original reporting tetapi answer surface menyerap value tanpa cukup referral atau commercial return, content economics bisa memburuk.

Ini bukan problem yang bisa diselesaikan hanya dengan AEO tactics.

Publisher harus membangun direct relationship dan business model yang tidak sepenuhnya bergantung pada click dari satu platform.

Newsletter.

Membership.

Data product.

Event.

Research service.

Community.

The exact mix differs.

AEO sebaiknya membantu distribution, bukan menjadi seluruh revenue strategy.

Then ask the harder business question:

What unique value exists after the sentence?

That's where the next generation of publishing will be decided.