Live benchmarks · last 30 days

What actually gets cited by AI search.

Anonymized, aggregated across every release the PressGEO citation monitor tracks. No marketing math — direct counts from citation_checks.

Releases monitored
3
Checks run
733
Citations captured
0

Per-engine citation rate

How often each engine surfaces a PressGEO release when its monitor query is asked. Live-citation engines (Perplexity, Gemini, Copilot) and knowledge-probe engines (ChatGPT, Claude, Grok) are measured differently — see methodology below.

Engine
Checks
Cited
Rate
Perplexity
108
0
0.0%
Gemini
108
0
0.0%
Copilot
108
0
0.0%
Claude
107
0
0.0%
ChatGPT
108
0
0.0%
Grok
107
0
0.0%
Reddit
87
0
0.0%

Citation quality breakdown

Every cited answer is classified by an AI grader on three axes. Based on 0 classified citations.

Depth
Primary subject0.0%
Passing mention0.0%
Form
Direct quote0.0%
Paraphrase0.0%
Link only0.0%
Sentiment
Positive0.0%
Neutral0.0%
Negative0.0%

Methodology

  • Rolling 30-day window across every release with citation monitoring enabled.
  • Live-citation engines: a hit means the release URL appears in the engine's public answer page for the monitored query.
  • Knowledge-probe engines: a hit means the company name or URL appears in the model's direct answer to the query.
  • Quality classification runs once per hit via Lovable AI gateway (Gemini Flash Lite). Older hits stay unclassified.
  • No per-user, per-release, or per-company data is exposed. Query strings are shown only when ≥6 checks have been run against them.