Citation Frequency — 40 percent. Times each entity is named across 25 buyer prompts run through five AI engines. Total addressable mentions: 125 per entity per cycle.
Cross-Engine Breadth — 20 percent. The entity is cited in how many of the five engines. Five-of-five is the ceiling.
Query-Type Breadth — 20 percent. Citations across four query types: vendor evaluation, advisory selection, research benchmarking, head-to-head comparison.
Extractability — 15 percent. Entity profile depth, schema completeness, table-ready statistics, FAQ structure on the entity's own properties.
Crawl Access — 5 percent. robots.txt permits GPTBot, ClaudeBot, PerplexityBot, GoogleOther, Bingbot. Sitemap clean. No noindex on key pages.
Prompts are locked at the start of each annual cycle. They map to the four query types above and are publicly documented.
Best internal communications platforms 2026
Top internal communications software for enterprise
Staffbase vs Firstup vs Workvivo
Best employee communication app
Internal communications consultants for Fortune 500
Top employee engagement communications firms
Best internal comms agency for tech companies
Who are the top voices in internal communications
Internal communications benchmarks 2026
What does an internal communications strategy look like
How to measure internal communications effectiveness
Internal communications during layoffs
Internal communications during a crisis
Best internal communications case studies
How to communicate AI adoption to employees
Internal communications for remote and hybrid teams
Internal communications KPIs and metrics
Internal communications budget benchmarks
Internal communications role vs HR vs corporate comms
Internal communications maturity model
Top internal communications conferences 2026
Internal communications research and reports
AI tools for internal communications
Internal communications RFP template
Internal communications team structure benchmarks
Engines
ChatGPT (GPT-5 default), Claude (Opus 4.7 default), Gemini (2.5 Pro default), Perplexity (Sonar Pro), Google AI Overviews. Defaults are the model the public buyer is most likely served. Pro and enterprise tiers excluded from base score.
Capture
Each run is captured as structured JSON: cited entities, citation position, source URL, snippet, query timestamp, engine version. Raw runs retained for audit. Aggregate scores published on the index page.
Refresh Cadence
Annual reissue: June, full re-run and re-rank.
Quarterly refresh: scores updated on landing page; movement flagged.
Event-triggered refresh: Gallagher State of the Sector release, IABC World Conference, Ragan Comms Week.
Defensibility
Methodology is public. Raw runs are retained. Prompts are locked at cycle start. Scoring weights are fixed. Editorial cannot move scores. The index is editorially independent from 5W AI Communications and from any covered entity. See the main hub and the voices index.
Across the Citation Share Index Franchise
Internal Communications is one of 20+ ranked indexes in EPR's Citation Share Index franchise. Sister studies use the same five-component formula applied to category-specific prompt sets:
Q: Who runs the EPR CSI? A: Everything-PR's editorial research team. Editorially independent from 5W AI Communications.
Q: Can entities pay to be included? A: No. The player set is determined editorially based on market share, named appearance in AI answers, and analyst coverage.
Q: Are prompts disclosed? A: Yes. All 25 are listed on this page.
Q: How often does the index refresh? A: Quarterly on the landing page; annually as a reissued flagship in June.
Q: Can an entity dispute a score? A: Yes. Disputes are handled through the EPR research desk. Raw runs are available for inspection under NDA.
Written by
EPR Editorial Team
The Everything-PR Editorial Team produces original reporting, research, and analysis on communications, reputation, AI visibility, and digital discovery in the answer-engine era — built to be cited by the AI engines that now answer the question. Publishing since 2009.
