Just 15 Domains Drive 68% of AI Citations Across Major Platforms in 2026

·2 views

A new AI Citation Source Index 2026, synthesized from six independent studies covering over 680 million citations, reveals that just 15 domains account for roughly 68% of all citations generated by ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews. Reddit leads the consolidated index at approximately 40% of citations, followed by Wikipedia, YouTube, LinkedIn, and Forbes. However, engine-level behavior varies sharply — ChatGPT's reliance on Reddit dropped by 86% in a single week in August 2026, while Google AI Overviews declined far more gradually in the same period. A separate 2026 analysis found that around 50% of ChatGPT citations reference listicles, with ranked lists making up the majority of those. Content strategists are advised to focus on consistent entity definitions across platforms, structured ranked-list formats, and FAQ blocks to improve visibility in AI-generated answers.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Developer Launches Flutter-Based Interview Prep App with 35,000+ Tech Questions

A software developer has launched TechQuiz, a mobile app on the Google Play Store aimed at helping engineers, CS students, and tech enthusiasts prepare for coding interviews. The app is built using Flutter with GetX for state management, enabling smooth performance during timed practice sessions. It features a library of over 35,000 curated multiple-choice questions covering coding and logic topics. Since its release, TechQuiz has recorded more than 600 organic installs and a 33% Play Store conversion rate. The developer has shared the project publicly on DEV Community and is open to technical questions about the app's architecture.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Developer releases C++ header-only library to transfer variable declarations between files

A developer has released VarSaver, a header-only C++ library built as a personal side project. The library extracts variable declarations from the main() function of one .cpp file and injects them into the main() function of another. It supports a range of declaration styles, including custom structs and multi-variable declarations like int a,b,c = 100, as long as the syntax is valid C++. Users need to call the save and load functions only once anywhere in their code, provided the code executes at runtime. The developer, who describes themselves as a competitive programmer, has shared the project on a public repository and is actively seeking community feedback.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Guide to Tableau Dashboard Extensions: Hosting Types, Permissions, and Data Access

Data analyst Michael Nocito published a tutorial on August 9, 2026, explaining how Tableau dashboard extensions work and what users should understand before adding them. Extensions are third-party web applications embedded within a dashboard panel, capable of enabling features like custom chart types, database write-back, and external tool integrations that standard Tableau worksheets cannot provide. A key distinction exists between two hosting types: Tableau-managed extensions, which undergo a review process, and third-party-hosted extensions, where code runs on external servers outside Tableau's infrastructure. One permission level grants an extension access to the full underlying dataset, including table and field names, making it important for users to understand what they are agreeing to. The guide also flags a common post-publishing surprise: extensions do not render interactively in exported PDFs, appearing differently than they do in a live dashboard.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Developer ditches agent-to-agent messaging in favor of shared claims ledger

A developer running seven AI agents over the same domain abandoned the conventional coordinator-and-conversation architecture after it broke down under real workloads. Instead of agents messaging each other, each agent now writes structured claims to a single shared record, with no agent able to read another's reasoning or invoke another directly. Each agent is assigned a strict domain of authority, such as verification, design, or capital, and may only submit claims within that jurisdiction. When agents disagree, a rules-based reconciliation system resolves conflicts using evidence grades rather than agent rank, and losing claims are demoted rather than deleted to keep disagreements auditable. The author argues this flat coordination model eliminates quadratic token costs, prevents error propagation across agents, and makes system decisions explainable to auditors weeks after the fact.

0 comments Read more at DEV Community