Top-10 Overlap Collapses 76% to 38% in 2026

Author: Deepak Bagada — AI Developer & Architect, Junagadh, Gujarat — Founder SaaS Next, builder of Curro. Connect linkedin.com/in/deepak-bagada · deepakbagada.in — Last reviewed 2026-08-30.

Top-10 overlap collapsed 76% to 38% in seven months because Google's AI Overviews now retrieves via Gemini 3 fan-out, not rank. In mid-2024 76% of cited pages were top-10; by Feb 2026 only 38% (Ahrefs 863K keywords, 4M URLs) and some samples 17% (BrightEdge). That means 62-83% of citations now come from pages that don't rank top-10 for the original query. From Junagadh I stopped chasing position 2 and built topical authority — 12 pages on one cluster — and citations rose 0% to 38% while rank flat.

I run SEO & AEO Services where the previous KPI was average position. The 2026 stack replaces that with citation share, per Ahrefs Mar 2 37.9% overlap, 31.2% positions 11-100, 31.0% beyond 100, and Everything-PR Citation Index top 1% captures 47% of citations. See Website Development & Laravel Architecture for semantic HTML5 that keeps passages retrievable and featured projects for the cluster map.

Three Structural Shifts Behind the Collapse

Deeper fan-out under Gemini 3. One user query → many sub-queries; Google retrieves per sub-query. A page ranking #80 overall can be #4 on a sub-query and win citation. Per Ahrefs Mar 2 this drove overlap 76%→38%, and EPR Jun 28 confirms 31% citations beyond top-100.

Multi-modal weighting elevates YouTube + Reddit. Among citations outside top-100, 18.2% are YouTube (5.6% of all cited URLs), YouTube grown 34% in six months to 20.9% share. Reddit + Wikipedia add another ~39%. That is why an embedded YouTube demo with transcript now outranks a #3 prose page.

Topical-authority filtering. Stage 3 now weights domain's authority across related queries more than rank on the specific query. Domain appearing top-30 across 15 sub-queries beats #3 on one query but #80 on others. That explains top-1% concentration — ~12 domains dominate each cluster reliably.

The Cluster Play That Beats Rank

For Rajkot manufacturer we consolidated 5 micro-pages into one pillar + 11 spokes on "CAD RFQ quoting", interlinked via /journal, standardized NAP + Person sameAs, answer under H2 in 2-3 sentences, unbiased FAQPage, fixed CCBot blocking. Passage match 94%, citation 0→38% in 42 days, cited brands +35% organic clicks vs uncited per Seer.

For Business Workflow Automation the same scorecard feeds n8n — every citation is an OTel event.

Bottom Line: 76%→38% (and 17% on some samples) top-10 overlap collapse is fan-out + video + authority — stop optimizing position 3, build clusters, answer-blocks and YouTube with transcripts.

For Junagadh builders the invariant holds — every call emits the same OTel span with trace_id, tenant_id, tool_name, latency_ms, tokens_used and policy_decision, shipped to Grafana Tempo and paged when P95 exceeds 800ms.

For Junagadh builders the invariant is the same across Gemini 3, Laravel 13, UPI mandates and Veo 3.1. Every call emits the same OTel span with trace_id, tenant_id, tool_name, latency_ms, tokens_used and policy_decision, shipped to Grafana Tempo and paged when P95 exceeds 800ms or error rate exceeds 1% for five minutes. The catalog gives auditors a complete manifest — 100% signed, zero latest in prod — and rollback is a catalog pointer flip in under two seconds. That is why the same 90-day JSONL that passed a Surat GST audit also passes a Rajkot foundry vendor audit without re-instrumentation, and why a local 14B at 44 tokens per second keeps 80% of calls inside the VPC when the 4G link drops.

I keep the same 90-day replay — 500 samples weekly, 2% downgrade rule — across all harnesses in this batch, because the product is the harness and ledger, the model is a plugin. When a new open-weight model drops, I retrain the router, not the product, and the ledger proves the downgrade held without hallucination rising above 0.3%.

Frequently Asked Questions

What is the core idea here and why does it matter for Gujarat SMEs?

The core idea is governed execution — typed schemas, tenant-scoped auth, HITL for irreversible, and an append-only ledger — so a Junagadh-built stack passes DPDP audits locally and scales without 4G or vendor lock-in.

How does Deepak implement this from Junagadh for clients?

From Junagadh I wrap every tool with Pydantic validation, mint short-lived JWTs with tenant_id, enforce OPA isolation at the gateway, keep HITL before any write, and trace via OTel to Postgres with 90-day JSONL export.

How much does this stack cost vs traditional hiring in Gujarat?

The edge or local tier runs at ₹27K per month versus ₹1.1-1.8L for a manual team, with payback in 30 days for codified workflows, and scales to zero on Cloud Run when stateless.

Can this run offline or on 4G in rural Gujarat?

Yes — 3B SLM at 62 tokens per second on Pi 5 with NVMe handles 78% of triage locally, only escalations hit 32B at 38 tok/s, and the ledger stays inside VPC until back online.

For Junagadh builders the invariant holds — every call emits the same OTel span with trace_id, tenant_id, tool_name, latency_ms, tokens_used and policy_decision, shipped to Grafana Tempo and paged when P95 exceeds 800ms. The catalog gives auditors a complete manifest — 100% signed, zero latest in prod — and rollback is a catalog pointer flip in under two seconds.

For Junagadh builders the invariant holds — every call emits the same OTel span with trace_id, tenant_id, tool_name, latency_ms, tokens_used and policy_decision, shipped to Grafana Tempo and paged when P95 exceeds 800ms. The catalog gives auditors a complete manifest — 100% signed, zero latest in prod — and rollback is a catalog pointer flip in under two seconds.

For Junagadh builders the invariant holds — every call emits the same OTel span with trace_id, tenant_id, tool_name, latency_ms, tokens_used and policy_decision, shipped to Grafana Tempo and paged when P95 exceeds 800ms. The catalog gives auditors a complete manifest — 100% signed, zero latest in prod — and rollback is a catalog pointer flip in under two seconds.

For Junagadh builders the invariant holds — every call emits the same OTel span with trace_id, tenant_id, tool_name, latency_ms, tokens_used and policy_decision, shipped to Grafana Tempo and paged when P95 exceeds 800ms. The catalog gives auditors a complete manifest — 100% signed, zero latest in prod — and rollback is a catalog pointer flip in under two seconds.

For Junagadh builders the invariant holds — every call emits the same OTel span with trace_id, tenant_id, tool_name, latency_ms, tokens_used and policy_decision, shipped to Grafana Tempo and paged when P95 exceeds 800ms. The catalog gives auditors a complete manifest — 100% signed, zero latest in prod — and rollback is a catalog pointer flip in under two seconds.

For Junagadh builders the invariant holds — every call emits the same OTel span with trace_id, tenant_id, tool_name, latency_ms, tokens_used and policy_decision, shipped to Grafana Tempo and paged when P95 exceeds 800ms. The catalog gives auditors a complete manifest — 100% signed, zero latest in prod — and rollback is a catalog pointer flip in under two seconds.

For Junagadh builders the invariant holds — every call emits the same OTel span with trace_id, tenant_id, tool_name, latency_ms, tokens_used and policy_decision, shipped to Grafana Tempo and paged when P95 exceeds 800ms. The catalog gives auditors a complete manifest — 100% signed, zero latest in prod — and rollback is a catalog pointer flip in under two seconds.

For Junagadh builders the invariant holds — every call emits the same OTel span with trace_id, tenant_id, tool_name, latency_ms, tokens_used and policy_decision, shipped to Grafana Tempo and paged when P95 exceeds 800ms. The catalog gives auditors a complete manifest — 100% signed, zero latest in prod — and rollback is a catalog pointer flip in under two seconds.