Tag

Retrieval

15 posts tagged Retrieval.

September 28, 2026

Outdated retrieval can overturn correct answers

Ahmed, Galke, Poech, and Röttger (arXiv 2609.31342, 28 Sep 2026) show that a single outdated but authentic document can flip answers models already get right without retrieval. Publication date is not validity. Reliable RAG needs selective trust over whether evidence still applies.

Mikhail Drozdov

Click-Chasing: Implications for SEO Indexing and Retrieval Quality

Traffic-first KPIs can degrade content quality, impacting indexing and retrieval. We outline technical checks and measurement for SEO engineers.

Mikhail Drozdov

Google Search Direction: Agent Manager, Infra Limits, and SEO Implications

Analysis of Pichai's interview reveals Google's agent manager concept, infra constraints, and new SEO visibility thresholds.

Mikhail Drozdov

Latest.

15 posts

AI Mode Shortlists: User Acceptance and SEO Implications for High-Stakes Purchases

LLM-powered shortlists are accepted by users, shifting control from manual research to AI curation. This changes ranking signals and SERP selection dynamics.

Mikhail Drozdov

Agentic Web Standards: MCP, A2A, NLWeb, AGENTS.md—SEO Implications

MCP, A2A, NLWeb, and AGENTS.md set interoperability baselines for AI-driven web agents. Early adoption impacts retrieval, indexing, and visibility.

Mikhail Drozdov

Four-Layer AI Access: Beyond llms.txt for Brand Content Control

A new four-layer architecture supersedes llms.txt, enabling granular, authoritative AI agent access to brand content.

Mikhail Drozdov

AMA PPC Archives: Structured Q&A as a Retrieval Signal in Paid Search

SEJ's monthly PPC Q&A archive is an evolving knowledge base. Its structure provides testable cues for SEO and retrieval, impacting search visibility.

Mikhail Drozdov

ChatGPT Ads Expansion: Acquisition Channel or Brand Tax?

OpenAI expands ChatGPT Ads with self-serve tools. This note analyzes PPC implications, indexation risks, and visibility effects for SEO engineers.

Mikhail Drozdov

Google: Why Core Updates Roll Out in Stages

Mueller explains staged core update rollouts. Practical implications for diagnosis, measurement windows, and separating indexing from ranking effects.

Mikhail Drozdov

Evergreen content in 2026: stop length targets, ship maintainable answers

SEJ frames evergreen in 2026 as moving past 2,000-word-for-its-own-sake and focusing on durable usefulness. Here's a testable plan.

Mikhail Drozdov

What AI-Driven Search Systems Actually Reward: A Technical Analysis

Dissects what content features AI ranking systems reward at scale, with falsifiable mechanisms and actionable tests for indexing and retrieval.

Mikhail Drozdov

Wikipedia bans LLM-written edits: SEO implications and tests

Wikipedia now forbids LLM-written or rewritten content (with two exceptions). Here's the retrieval, trust, and measurement impact for SEO.

Mikhail Drozdov

How AI Systems Select Sources: Implications for SEO Testing

SEJ reports analysis of 21k+ citations linking content length, depth, and focus to being cited. Here's a falsifiable 7‑day test plan.

Mikhail Drozdov

GEO Strategies for AI Search: Tactical Visibility in 2026 SERPs

Breakdown of practical GEO tactics to increase brand recommendations by AI search engines, with falsifiable checks and retrieval-focused metrics.

Mikhail Drozdov

Surviving AI Search in 2026: Technical SEO Strategies for Resilience

Actionable, testable strategies for SEO teams to adapt to AI-driven search in 2026. Focus: entity alignment, signal diversity, and retrieval optimization.

Mikhail Drozdov