Methodology
AEO Goal Methodology
AEO Goal is a measurement product, so how it measures is as important as what it reports. This page explains what AEO Goal measures across AI answer engines, how each measurement is produced and verified, and where the limits are, then links to the detailed references for AI visibility and AI citation tracking.
Measurement principles
Four principles govern every number AEO Goal reports.
Measure AI answers directly
AI visibility is read from real answer-engine responses to tracked buyer prompts, never inferred from keyword rankings or other proxies. Google rankings are measured separately by daily rank tracking and reported as their own metric, so neither surface stands in for the other.
Verify before it counts
Cited source URLs pass a two-stage check before they affect customer-facing scores. Citations verified as false are excluded from share-of-model math.
Exclude synthetic data
Missing-key, placeholder, failed-provider, and fallback responses are filtered before persistence, so they never inflate a measurement.
Report honestly
Numbers ship with confidence ranges, prior-period movement, and stated limitations instead of false precision.
What AEO Goal measures
Share of model
Brand mentions divided by total tracked answers in the window, so the denominator is explicit and the figure reconciles on screen.
AI citations and sources
The source URLs an answer engine cites, with quality and position signals, plus verification of whether the page actually mentions the brand.
Competitor overlap
Which competitors appear in the same answers, so wins and losses are comparative rather than absolute.
Sentiment and position
Whether a mention is favorable and where it lands within the answer, not just that it exists.
Market and locale coverage
Prompts carry market and locale context so results reflect the buyer, not a single default region.
Measurement cadence
A scheduled hourly dispatcher runs each prompt on its configured frequency. Measurement is scheduled, not real time.
How a measurement is produced
Every share-of-model figure is the output of the same repeatable pipeline.
Run tracked buyer prompts across configured AI answer engines on a scheduled cadence.
Detect whether the target brand and competitors appear in each answer.
Capture cited source URLs, answer context, sentiment, and position.
Verify cited URLs with a two-stage check: reachability, then body fetch and brand-mention judging.
Filter synthetic, placeholder, and failed-provider responses out of persisted data.
Aggregate into share of model with confidence ranges and prior-period movement.
Measurement references
The full detail behind each measurement, including formulas, verification, retention, and per-area limitations.
Honest limitations
- Measurement is scheduled per prompt, not real time.
- Live engine coverage depends on configured provider credentials and enabled crawler paths.
- Comparative signals inform decisions; they are not guarantees of ranking or citation.
- Vendor security, retention, and contract terms should always be verified independently.
See the methodology on your own brand
Run a free AI visibility scan, or review how AEO Goal handles your data before you start.