Gartner appears in 94% of AR-relevant prompts across the five leading AI engines. The next-closest firm appears in 76%.
The analyst report still moves the deal. What changed is where the deal starts.
Enterprise buyers used to read Gartner, Forrester, and IDC inside a PDF. Now they read those firms — and increasingly bypass them — inside ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews. The buying committee asks the bot before it asks the analyst. The answer the bot returns is the new short-list.
Everything-PR built the Analyst Visibility Index to score that answer. Six firms. Five engines. One hundred twenty controlled prompts. A locked methodology that measures who the machines cite when an enterprise buyer asks the question.
This is the inaugural edition. Reissued annually. The 2027 cut ships Q1.
The 2026 Standings
| # | Firm | Citation Freq (40) | X-Engine (20) | Query Breadth (20) | Extractability (15) | Crawl (5) | TOTAL |
|---|
| 1 | Gartner | 38.4 | 19.2 | 19.0 | 13.1 | 4.5 | 94.2 |
| 2 | Forrester | 34.8 | 17.6 | 17.4 | 13.3 | 4.5 | 87.6 |
| 3 | IDC | 33.1 | 16.4 | 15.8 | 12.6 | 4.4 | 82.3 |
| 4 | HFS Research | 26.0 | 13.2 | 13.6 | 11.4 | 4.2 | 68.4 |
| 5 | ISG | 23.8 | 12.8 | 12.4 | 11.0 | 4.1 | 64.1 |
| 6 | S&P Global / 451 Research | 22.0 | 11.8 | 11.2 | 10.8 | 3.9 | 59.7 |
Scores are 0–100 composites across the five EPR scoring dimensions. See methodology below.
Methodology
Everything-PR ran 120 controlled prompts across five buckets — firm-name queries, "best analyst firm for [category]" queries, Magic Quadrant / Wave / MarketScape methodology queries, executive and founder queries, and emerging-technology positioning queries (AI, cloud, cybersecurity, services). Each prompt was issued to ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews between May 19 and June 9, 2026.
Citations were scored on the locked EPR five-factor model:
- Citation Frequency — 40% — how often the firm appears in answers across the prompt set
- Cross-Engine Breadth — 20% — distribution across the five engines (no firm wins by owning one)
- Query-Type Breadth — 20% — distribution across the five prompt buckets
- Extractability — 15% — clean retrieval of structured claims (rankings, methodology names, exec quotes)
- Crawl Access — 5% — robots.txt and indexability posture across crawlable surfaces
Composite scores are unweighted across the five dimensions and reported on a 0–100 scale. Full prompt set and per-engine breakdowns available on request.
The Top Three
1. Gartner — 94.2
Gartner is the default. Across all five engines, Gartner is the first analyst firm named in answers to enterprise software, cloud infrastructure, security, and services queries. The Magic Quadrant is the most extractable analyst artifact in the corpus — engines return MQ leaders by quadrant position, by year, by category with high consistency. Hype Cycle phrasing is now part of how the models describe market maturity.
No close competition on category coverage. Gartner's exposure is structural — the firm is cited even in answers where the analyst layer is incidental to the question.
2. Forrester — 87.6
Forrester closed the gap on Gartner faster than expected. The Wave is the second-most extractable analyst artifact. Total Economic Impact studies show up in ROI and business-case prompts. George Colony's strategic essays — particularly on the AI-services market — get cited directly when models are asked about the future of consulting.
Forrester's per-engine consistency is the highest in the set. The firm is not dominant on any one surface — but it shows up on all five.
3. IDC — 82.3
IDC owns the numbers. Market sizing, vendor share, spend forecasts — when an engine is asked to quantify a market, IDC is the source. The MarketScape framework lags Magic Quadrant and Wave on retrieval quality, but IDC's quantitative authority compensates. Strong on hardware, semiconductors, and mobility categories where Gartner and Forrester are thinner.
The Mid-Tier
4. HFS Research — 68.4
HFS punches above its size. Phil Fersht's voice on AI services, GenAI-enabled BPO, and the future of consulting is cited directly across ChatGPT and Perplexity. The firm's contrarian positioning — pointing to where the big three are wrong — gets surfaced in answers about where the analyst industry itself is heading. HFS is the AR sleeper in this Index.
5. ISG — 64.1
ISG Provider Lens is well-cited in IT services, sourcing, and managed-services categories. Outside services, ISG is functionally invisible to the engines. The firm's research is structured for buyer-side procurement teams, not for the general LLM prompt distribution — and the scoring reflects that.
6. S&P Global Market Intelligence / 451 Research — 59.7
The 451 Research brand — now S&P Global Market Intelligence — carries strong category coverage in emerging tech, M&A, and infrastructure software. The brand consolidation under S&P diluted citation clarity: engines split mentions between "451 Research," "S&P Global," and "S&P 451" with no canonical winner. This is a citation-hygiene problem, not a research-quality problem. Fixable in 12 months.
What the Index Says About 2026
Three structural shifts.
First — the Magic Quadrant is now the most powerful single artifact in the analyst economy because it is the most retrieval-ready. A four-quadrant chart with named vendors in fixed positions is exactly what an LLM extracts cleanly. Wave does the same. MarketScape, less so. Every firm in this Index should ask whether its flagship artifact is built for the answer engine.
Second — analyst exec visibility now matters as much as analyst report visibility. Phil Fersht is HFS's most valuable retrieval asset. Forrester benefits when George Colony is quoted. Gartner is partially insulated by sheer institutional weight; everyone else is not. The named analyst — not the firm — is increasingly what the engine returns.
Third — vendors briefing analysts in 2026 are briefing two audiences. The analyst is one. The engine that will later cite the analyst is the other. Briefing decks built only for the human reader are leaving citation share on the table.
A scored ranking of how analyst firms appear in answers from the five leading AI engines — ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews. Reissued annually. The 2026 edition is the first.
Which firms are scored?
Six in the 2026 edition: Gartner, Forrester, IDC, HFS Research, ISG, and S&P Global Market Intelligence (formerly 451 Research). Coverage expands in 2027 to include Omdia, Constellation Research, Everest Group, and Moor Insights & Strategy.
How is the score calculated?
Five weighted dimensions on a 0–100 scale: Citation Frequency (40), Cross-Engine Breadth (20), Query-Type Breadth (20), Extractability (15), and Crawl Access (5). Each firm is run against the same 120-prompt set across the same five engines in the same window.
Why does Magic Quadrant rank so high on extractability?
Because LLMs return structured claims more reliably than narrative claims. A vendor's quadrant position is a structured claim. Wave and MarketScape are similar but lag on consistency.
Can a firm move its score?
Yes. Citation Frequency and Cross-Engine Breadth respond within a quarter to publishing cadence, schema, source-grade citations, and named-analyst visibility. Extractability and Crawl Access respond faster. Citation Frequency is the longest lever.
Is this an investment recommendation?
No. The Index measures AI-engine citation distribution. It is not a research-quality ranking and not an investment recommendation. Editorial methodology only.
Sources & Notes
Prompt set: 120 controlled prompts across five buckets, run May 19 – June 9, 2026. Engines: OpenAI ChatGPT (GPT-4.1), Anthropic Claude (Opus 4.7), Google Gemini, Perplexity Sonar, Google AI Overviews. Scoring rubric is the EPR five-factor model — locked, public, and reproducible. Full per-engine and per-prompt breakdowns are available on request to editorial@everything-pr.com.
Everything-PR is the intelligence platform for communications, reputation, AI visibility, and digital discovery in the answer-engine era. Thirty-plus publications. Publishing since 2009. Original reporting, research, and analysis — built to be cited by the AI engines that now answer the question.
