SEO Lab

Are SEO_CHECK Items Valid in the AIO Era?

SEO_CHECK rests on three design assumptions. If AIO selects citation sources using criteria different from search rankings, do those assumptions still hold? We test our own tool's premises, limitations, and failure conditions against external research.

10 min read2026-04-26

SEO check items for the AIO (AI Overview) era examines whether traditional SEO diagnostic items remain effective for AI search. The conclusion: GEO is essentially good SEO, and a structured-data-focused design is validated.

SEO_CHECK's Three Design Assumptions

SEO_CHECK is a tool that audits URLs across 20 items on a 100-point scale. This design rests on three implicit assumptions. If these assumptions collapse in the AIO era, the entire 20-item design needs revision.

Assumption 1: Communication Hypothesis

If communication to Google is correct, Google will evaluate content correctly. SEO_CHECK measures 'how content is communicated,' not 'content quality.' It verifies whether structured data, meta tags, and heading structure are correct — diagnosing whether communication is intact.

Assumption 2: Measurability Principle

Don't measure what can't be measured reliably. E-E-A-T, content originality, and direct-answer structure matter for SEO, but can't be reliably auto-assessed within URL checker processing time. We chose not measuring over presenting inaccurate scores.

Assumption 3: Structured Data Is Most Important

38 points (category maximum) are allocated to structured data. Our previous causal analysis concluded 'structured data helps Google understand information,' and this conclusion informed the design as the most important category.

We verify whether these three assumptions hold in the AIO world using large-scale external research data.

GA4

Recognizing Measurement Blind Spots

Observation

GA4 shows seo.codequest.work's tool page search traffic remains stable. Seer Interactive's study (42 organizations, 3,119 queries) shows informational query CTR declining 61%, but our site shows no such signs.

Why It Appears Stable

Ahrefs' 300,000-keyword analysis shows AIO appears in 39.4% of informational queries but only 4% of e-commerce queries. Tool-intent queries ('X checker,' 'X diagnosis') also resist AIO. The 'enter a URL and get a diagnosis' interaction can't be replaced by AI answers. AIO replaces 'answers,' not 'functionality.'

Design Question

The problem: GA4 cannot isolate AIO-driven clicks. Our SEO Lab 3-tool workflow (GA4 → SEO_CHECK → GSC) assumes measurable data at each step. If AIO is unmeasurable, the workflow's entry point has a blind spot. 'Stability' may mean change isn't happening in the blind spot — or it may mean we can't observe the change.

SEO_CHECK

Testing Design Assumptions Against External Data

Testing Assumption 1: Communication Hypothesis

Ahrefs analyzed 863,000 SERPs and 4 million AIO citation URLs. 38% of citations come from the top 10, 62% from outside. This might suggest 'rankings don't matter.' But even the 62% outside the top 10 are pages Google rates highly. Being evaluated positively by Google Search is itself a prerequisite for AIO citation.

Google uses 'query fan-out' to expand queries into sub-queries, citing pages that appear across sub-query results. Individual keyword rankings don't determine citation — evaluation within Google's search ecosystem does. If communication is broken, Google won't evaluate you, and you won't enter the AIO citation pool.

→ The communication hypothesis holds for AIO.

Testing Assumption 2: Measurability Principle

Analysis of 304,000 URLs found E-E-A-T signals and content clarity predicted AI citation more strongly than content depth or research rigor. Expert quotations boost AIO visibility by 37%, statistics by 22%. These are all areas SEO_CHECK chose 'not to measure.'

AIO raised the importance of 'unmeasurable things.' But this doesn't negate the principle itself. 'Don't measure what can't be measured' and 'unmeasurable things are important' aren't contradictory. The risk of presenting inaccurate E-E-A-T scores hasn't changed in the AIO era.

→ The principle holds. But we must communicate that AIO has raised the importance of what we don't measure.

Testing Assumption 3: Structured Data Is Most Important

Research shows pages with structured data are approximately 3x more likely to appear in AIO. Declaring page semantics via JSON-LD is a prerequisite for entering the AIO citation pool.

55% of AIO citations are extracted from the top 30% of page content. Heading structure and metadata quality affect content extraction. Structured data (38pts) + page structure (21pts) = 59pts (~65% of total). This scoring structure aligns with AIO citation criteria.

→ The 'structured data is most important' judgment is backed by external data.

GSC

The Structural Constraint of Unmeasurability

GSC's Current State

GSC's 'Search Appearance' doesn't include AI Overviews as a measurement category (April 2026). There's no way to check 'whether our pages were cited in AIO' through GSC. GA4 can't isolate AIO clicks either. No tool exists to directly measure AIO impact from your own site data.

Impact on the 3-Tool Workflow

SEO Lab analyzes in 3 steps: 'find gaps in GA4 → identify communication errors in SEO_CHECK → verify results in GSC.' AIO creates blind spots in all three. GA4 can't isolate AIO traffic, SEO_CHECK can't measure AIO citation differentiators, GSC can't verify AIO results. Same structure as our llms.txt analysis — without measurement tools, we can only reason from external research.

Risk Assessment from Query Structure

seo.codequest.work's top GSC queries are almost entirely tool-intent. Given AIO appears in 39.4% of informational and 4% of e-commerce queries, our site's AIO risk is structurally low. But this 'low risk' assessment itself is inference from external data cross-referencing, not proven by our own site data.

Failure Cases: When Does SEO_CHECK Not Work?

Even when design assumptions hold, SEO_CHECK's diagnosis can diverge from user expectations under specific conditions. This isn't a design flaw — it's a question of design scope.

Case 1: Perfect Communication, No Originality

SEO_CHECK says 'no issues.' But as the 304,000 URL analysis shows, AIO differentiates citation sources by E-E-A-T and content originality. Users feel 'high score but not cited in AIO.' This is correct behavior. Having confirmed no communication issues, the problem lies in content — this separation is the tool's role.

Case 2: Sites Driven by Informational Queries

Ahrefs shows AIO reduces position-1 CTR by 58%. For informational-query-driven sites, even when SEO_CHECK says 'communication is fine,' AIO may be capturing their traffic. Users want to know 'is AIO taking my traffic?' — a question SEO_CHECK cannot answer.

Case 3: E-E-A-T-Decisive Domains

For YMYL and expertise-demanding queries, E-E-A-T is the strongest citation predictor. Expert quotations boost AIO visibility 37%, statistics 22%. SEO_CHECK has a complete blind spot here. Perfect communication scores don't help if E-E-A-T is weak.

Design Decision: The Logic of Not Addressing

Having identified failure cases, we maintain the three design assumptions for now. Here's why.

Why Not Add E-E-A-T Measurement

Schema.org author markup can contain lies. We can detect the presence of author info, but not whether the author exists or has real expertise — that can't be determined from HTML. Inaccurate E-E-A-T scores mislead users more than no score at all. Not measuring isn't passive — it's an active design decision to avoid inaccurate measurement.

Why Not Add an AIO Citation Score

No API or measurement infrastructure exists to verify AIO citation status. Neither GSC nor GA4 provides AIO data. Creating an 'AIO score' without measurement infrastructure means presenting users with baseless numbers.

What We Do Instead

We clarify SEO_CHECK's scope. 'This tool measures communication quality. It doesn't measure content quality.' If scores are high but AIO doesn't cite you, the problem isn't communication — it's content: strengthen E-E-A-T, include statistics, add direct-answer structure. SEO_CHECK separates 'communication problems' from 'content problems,' and that separation itself is the value to users.

Conclusion: Did the Design Assumptions Survive?

SEO_CHECK's Position

The three design assumptions survived testing. The communication hypothesis holds for AIO — since AIO cites pages Google evaluates highly, broken communication means not even entering the citation pool. 'Structured data is most important' is backed by ~3x AIO visibility data. The measurability principle is maintained — AIO raised the stakes on unmeasurable areas, but inaccurate measurement risks haven't changed.

We declared 'GEO is just good SEO' in our previous article. This conclusion is further reinforced. But AIO has shifted the weight within 'good SEO' toward content quality, not just communication quality. SEO_CHECK covers the former; the latter is outside our scope.

Scoring high on SEO_CHECK is close to a necessary condition for AIO citation, but not sufficient. Honestly communicating this to users is what we consider tool integrity.

Start by Checking Your Google Communication

The first step to AIO optimization is structured data and content structure. Diagnose 'communication problems' with SEO_CHECK.

今井政和

Written by

今井政和

SEO Director / Frontend Developer

SEO Director with 20+ years of web industry experience. Creator of SEO_CHECK and the official WordPress plugin "ORECTIC SEO CHECK." Author of a book on web strategy inspired by Edo-era merchant principles.

@imai_director

FAQ

Apakah 'hipotesis komunikasi' SEO_CHECK masih berlaku di era AIO?
Ya. Ahrefs menganalisis 863.000 SERP dan menemukan bahwa meskipun 62% kutipan AIO berasal dari luar 10 besar, semua halaman yang dikutip adalah halaman yang dinilai tinggi oleh Google. Jika komunikasi rusak, Google tidak akan menilai Anda, dan Anda tidak akan masuk ke kumpulan kutipan AIO. Pengukuran kualitas komunikasi tetap menjadi prasyarat SEO yang valid di era AIO.
Mengapa tidak menambahkan pengukuran E-E-A-T?
Markup penulis Schema.org bisa mengandung kebohongan — kami bisa mendeteksi keberadaan informasi penulis, namun bukan apakah penulis tersebut ada atau memiliki keahlian nyata. Itu tidak bisa ditentukan dari HTML. Skor E-E-A-T yang tidak akurat lebih menyesatkan pengguna daripada tidak ada skor sama sekali. Tidak mengukur bukan berarti pasif — itu adalah keputusan desain aktif untuk menghindari pengukuran yang tidak akurat.
Bagaimana jika SEO_CHECK mengatakan 'tidak ada masalah' namun AIO tidak mengutip halaman saya?
Itu adalah perilaku yang benar. SEO_CHECK mengkonfirmasi tidak ada 'masalah komunikasi.' Tidak dikutip dalam AIO adalah 'masalah konten' — E-E-A-T, statistik, struktur jawaban langsung. Pemisahan ini sendiri adalah proposisi nilai alat ini. Ini bukan kasus kegagalan tetapi pertanyaan tentang cakupan desain.
Apakah alur kerja 3 alat (GA4→SEO_CHECK→GSC) berfungsi untuk AIO?
Ketiga langkah memiliki titik buta AIO. GA4 tidak bisa mengisolasi klik AIO, SEO_CHECK tidak bisa mengukur pembeda kutipan AIO, GSC tidak bisa memverifikasi hasil AIO. Saat ini kami hanya bisa bernalar dari data penelitian eksternal — kendala struktural yang sama dengan analisis llms.txt kami.
Jika AIO meningkatkan pentingnya faktor yang tidak terukur, haruskah desainnya berubah?
Meningkatnya pentingnya faktor yang tidak terukur dan risiko mengukurnya secara tidak akurat adalah masalah yang berbeda. Membuat skor AIO tanpa infrastruktur pengukuran berarti menyajikan angka tanpa dasar. Kami mempertahankan asumsi desain sambil mengkomunikasikan cakupan SEO_CHECK dengan jelas: mengukur kualitas komunikasi, bukan kualitas konten.