JUL 2026Benchmark
Answer Engine Index: Q3 2026
An illustrative benchmark format using fictional figures and example interpretations. It is not a report of measured engine behavior.
12 MIN · METHODOLOGY V1.0
Home / Research
Research Hub
The evidence base for generative search.
Explore the GEO explainer, a sample benchmark format and a dated review of the original website. Sample exhibits are not collected research findings.
01 · Featured
ANSWER ENGINE INDEX · QUARTERLY
Answer Engine Index: Q3 2026.
A sample quarterly benchmark format, using illustrative values for four engines and 1,200 prompts. No prompt collection or quarterly monitoring run has occurred in this build.
FIG. 01 · Q3 HEADLINE NUMBERSSAMPLE
PROMPTS TRACKED
1,200
AVG CITATION RATE
11.4%▲1.2
ENGINES COVERED
4
TOP-DECILE SHARE
61%▲4
FIG. 01 · SAMPLE — NOT MEASURED. ILLUSTRATIVE VALUES PENDING MONITORING API.
02 · Library
Reference pages and sample reports.
JUL 2026Case study
We run the method on ourselves
A historical review of the original build dated 27 July 2026. It distinguishes file inspection from unavailable engine evidence and reports no composite score.
MEASURED (TOOL) · METHODOLOGY V1.0
The answer brief · planned
The research, before it's conventional wisdom.
The Answer Brief is a planned newsletter about GEO research and measurement.
Newsletter subscriptions are not open yet. Please check back for availability.