# Concentration Beats Coverage Under Retrieval ## Claim When a system finds answers by retrieving and synthesizing fragments rather than ranking whole documents, scattering near-duplicate content divides authority instead of compounding it, so concentrating signal into fewer, clearer, canonical sources wins over broad shallow coverage. <!-- atomicity: 1 idea ✓ --> ## Evidence **Source**: [[3. Resources/Readwise/Articles/Why Publishing More Content Is Making Your SEO Worse|Why Publishing More Content Is Making Your SEO Worse]] (Search Engine Journal) > [!quote] > "When similar ideas are fragmented across many URLs, no single page accumulates dominant semantic weight. You are no longer strengthening your authority. You are dividing it." The mechanism is embedding-level: AI search chunks documents into passages, embeds them as vectors, and synthesizes from the best fragments — so near-duplicate pages compete against *each other* for the same semantic space ("vector competition") and none wins. The article names the goal "authority density." The same physics governs a Zettelkasten: scatter a claim across many weak near-duplicate notes and no note accumulates dominant weight; consolidate into one strong, well-linked canonical note and authority compounds. ## Scope & conditions - Holds wherever discovery is retrieval/embedding-based (AI search, RAG) and analogously for note systems that surface answers by linking + search. - Weakens where genuinely distinct user intents need separate pages/notes, or where long-tail coverage still captures un-synthesized demand a single canonical source would miss. ## Links - [[3. Resources/Permanent Notes/Retrieval Quality Is Set by the Unit of Knowledge Not the Retrieval Algorithm]] — extends: canonical atomic units concentrate retrieval authority because unit design sets the ceiling before search and reranking begin - [[Abstracting Drudgery Re-Opens Creative Work to Non-Specialists]] — contrasts: abstraction broadens who can create, while retrieval systems still reward concentrated authority over broad coverage. - [[Personal Knowledge Management MOC]] — applies-in — this is the SEO-domain statement of the consolidation move the whole vault performs: one strong canonical note beats many near-duplicates - [[Atomic Notes and Linking]] — extends — atomicity + meaningful links are how you concentrate signal rather than scatter it - [[Building a Second Brain and the Zettelkasten Method]] — extends — the method that operationalizes concentration-over-coverage for notes - [[AI & the Future of Knowledge Work MOC]] — applies-in — retrieval reshapes how authority is earned, for content and for knowledge work - [[3. Resources/Permanent Notes/Event Recap Process]] — applies-in: maintaining one canonical recap concentrates event authority rather than splitting it across competing summaries. - [[3. Resources/Permanent Notes/Serving Clean Markdown Solves Retrieval but Not Discovery]] — applied in: serving markdown by path doesn't tell an agent which page answers its question (Sepic). - [[3. Resources/Permanent Notes/A Docs Page Must State Its Purpose in the Title and Opening Because Agents Lose Navigation]] — has evidence from: context that lives only in the sidebar is lost when an agent receives a chunk (Washington). ## Implications - For courtneyr.dev / DevRel content: prune and consolidate near-duplicate pages into cornerstone assets; measure AI-citation share, not page count. - Open question: what's the operational metric for "authority density" / "vector competition" with real tools, and when does consolidation backfire by losing distinct intents?