Start with a defined measurement universe.
AI visibility has meaning only when the measured entity, category, questions, markets, languages, AI environments and observation period are explicit. Intelligence keeps those dimensions visible so results can be compared over time without silently changing the denominator.
Entity set
Brands, products, services, competitors, categories and public organizations included in the analysis.
Prompt / intent universe
Questions grouped by informational, comparison, purchase, switching, troubleshooting, reputation and other decision intents.
Markets & languages
Country, region and native-language scopes kept separate when answer behavior or demand differs.
AI environments
The supported answer environments and observation windows included in a given measurement.
Demand intelligence maps what people are likely asking AI.
Because private AI conversations are not a general data source, Intelligence does not pretend to see them. It builds approved prompt/intent panels from category research, controlled sampling, first-party signals and authorized platform telemetry where available. Modeled demand is labeled estimated.
| Demand view | Decision supported |
|---|---|
| Estimated AI demand | Prioritize questions and topics with meaningful category interest. |
| Prompt trend | Identify intents or product questions gaining or losing attention over time. |
| Purchase intent | Separate broad research from comparisons, shortlists and buying decisions. |
| Topic / intent coverage | Find strategically important demand families where the brand has weak answer presence. |
Answer intelligence looks beyond simple mentions.
An entity can be mentioned, recommended, cited, compared negatively, described with outdated attributes, or absent while competitors dominate. Intelligence preserves those distinctions.
Presence
Where and how consistently the entity appears across the scoped answer opportunities.
Recommendation
Where the entity is actively recommended or shortlisted, not merely named.
Citations
Which exposed sources/domains appear as citations or references when the AI environment provides them.
Characterization
Recurring product attributes, advantages, limitations, reputation claims and narrative patterns associated with the entity.
Core Intelligence metrics use public definitions.
Relevant metrics include AI Visibility, AI Share of Voice, Mention Rate, Recommendation Rate, Citation Rate, Citation Coverage, Topic Coverage, Intent Coverage, Narrative Frequency and Estimated AI Demand. Each metric is interpreted inside its documented scope rather than as a universal score.
Visibility does not mean recommendation. A direct citation does not automatically prove causal source influence. Estimated demand is not presented as a count of private conversations.
Competitive intelligence shows who wins the same answer moments.
Compare brands inside one defined universe: share of voice, recommendation behavior, topics, sources, narratives, markets and movement over time. Competitor discovery can support customer-defined sets and approved discovery workflows.
Source intelligence
Where citations or source references are exposed, Intelligence records recurring publishers, first-party domains and communities associated with the measured answers. Source influence may be modeled; hidden sources are not invented when an environment does not expose them.
Track what AI repeatedly says about the brand.
Narrative analysis groups recurring descriptions and claims: price, quality, safety, availability, innovation, service, trust, comparisons and other category-specific attributes. The same narrative can then be compared across languages and markets using native-language prompt panels rather than a single translated English panel.
- Find attributes competitors own more consistently.
- Identify outdated or inaccurate recurring facts.
- See which market/language has the largest information gap.
- Separate answer-level observations from assumptions about proprietary ranking logic.
From category question to measurable action.
- Define entities and markets.
Choose the brand, products, competitors, languages and decision contexts. - Build the prompt/intent universe.
Create a scoped panel of relevant user questions and classify intent families. - Measure demand and answers.
Estimate demand where appropriate and observe presence, recommendation, citations, competitors and narratives. - Identify information gaps.
Find high-value demand where the brand is missing, weakly represented or supported by poor source coverage. - Choose an intervention.
Use first-party content, Distribution, paid media or another legitimate information/marketing program. - Measure again and verify when needed.
Compare the same scope over time and use Verify for independent claims/evidence.
Intelligence connects to the rest of SIGNAL without collapsing evidence classes.
| Analytics | Adds first-party crawler, agent, referral, conversion and revenue evidence from connected properties. |
| Verify | Uses scoped Intelligence observations as part of independent campaign, GEO or vendor audits. |
| Ads | Uses demand and competitive gaps to inform paid media strategy without promising paid changes to organic answers. |
| Distribution | Supplies authoritative information into publisher and machine-readable channels; Intelligence then measures downstream answer/source changes. |
See your category the way AI answers it.
Establish a scoped baseline for demand, visibility, recommendations, citations, competitors and source influence.