Glossary

In plain terms. The single 0 to 100 number that summarizes how AI engines see your destination — a weighted blend of DAVI Website Score (the structural readability of your web presence) and DAVI Perception Score (how AI actually describes you). The same composite name applies at the destination, partner, and competitor level.

Technically. Composite of the DAVI Website Score and DAVI Perception Score, rebased to 0 to 100 and re-scored monthly off the canonical scoring run. Specific weights are part of the proprietary DAVI methodology and are configurable per client.

In plain terms. Whether AI engines can actually read your destination's web presence. Crawler access, structured data, editorial density, third-party readability, and the Zertura Destination Layer.

Technically. Composite of structural sub-components covering crawler access, site architecture, editorial surface, external & earned readability, and the Zertura Destination Layer. Weighted by AI grounding impact and rebased to 0 to 100. Sampled across a representative page set per monthly scoring run.

In plain terms. How AI actually describes your destination across the major LLMs, scored 0 to 100. This is the perception layer · what AI says about you, not whether it can reach your site.

Technically. Composite of equally weighted perception dimensions (Presence, Accuracy, Distinctiveness, Recency, Sentiment) measured across leading AI models on the destination's prompt set. Equal-weighted by default; weighting is configurable per DMO.

In plain terms. How well AI's actual narrative about your destination matches each of your brand's strategic pillars. Measured per pillar with a Strong / Partial / Gap grade and a 0 to 100 score.

Technically. For each pillar, AI responses to the pillar's test prompt set are scored against the canonical pillar definition for Strong (matches verbatim or with high paraphrase), Partial (matches partially or with caveats), or Gap (absent or contradicted). Score is the weighted share of Strong + 0.5 × Partial, rebased to 0 to 100. Sourced from direct LLM querying only.

In plain terms. A Net Promoter-style score for how AI describes your destination. The share of AI-generated language that's promoter-positive minus the share that's detractor-negative.

Technically. % Promoter sentences (polarity ≥ +0.5) minus % Detractor sentences (polarity ≤ -0.2) across the AI response corpus for the destination's prompt set. Range −100 to +100. Reported alongside Sentiment Score on the AI Sentiment page.

In plain terms. How positively AI describes your destination across all the responses we collect. 100 means uniformly glowing; 0 means uniformly negative.

Technically. Mean polarity of AI-generated descriptive sentences mentioning the destination, scored on a −1 to +1 scale and rebased to 0 to 100. Sentences are extracted per response, classified by a fine-tuned sentiment model, and aggregated by stage and topic.

In plain terms. How unique the AI's description of your destination is, versus generic 'beach town' / 'Gulf Coast' boilerplate.

Technically. Measures n-gram overlap between AI responses about this destination and a control corpus of generic destination copy. Lower overlap = higher distinctiveness. Reported on the same 0 to 100 scale as the other dimensions.

In plain terms. How often AI engines surface your destination across the prompts visitors actually ask.

Technically. Mention rate across the destination's tracked prompt set × LLM panel. A mention counts when the destination or one of its designated entities appears in an answer to a relevant prompt.

In plain terms. Whether AI returns correct facts about your destination, location, hours, claims, attributions.

Technically. Per-claim verification against a curated source-of-truth corpus (GBP, Wikipedia, partner sites). Each AI claim is labeled correct, incorrect, or unverifiable; score is the share of verifiable claims marked correct.

In plain terms. How differentiated and on-brand AI's narrative is, versus generic copy.

Technically. See Distinctiveness Score. Reported per-entity on the dimension breakdown.

In plain terms. How fresh the sources AI is citing are. Stale citations reduce visibility over time.

Technically. Median publish-or-update age of citations AI returns for this destination's prompt set. Buckets: <90d (excellent), 90 to 365 days (strong), 1 to 2 years (watch), over 2 years (critical). Score is rebased to 0 to 100.

In plain terms. Tone and positivity of AI-generated descriptions of the destination.

Technically. See Sentiment Score. Rolled up by visitor-journey stage on the Visitor Journey page.

In plain terms. 0 to 39. Actively pulling the composite down. Needs immediate attention.

Technically. 0 to 39. Flagged with a red top border on partner cards and surfaced in Priority Recommendations.

In plain terms. 40 to 59. Below where it should be and likely to slide further without action.

Technically. 40 to 59.

In plain terms. 60 to 79. Healthy performance. Protect the gains and look for selective lift.

Technically. 60 to 79.

In plain terms. 80 to 100. Best in class for this destination size and category.

Technically. 80 to 100.

In plain terms. The raw count of times AI surfaced your destination across the prompt set in this period.

Technically. Sum of qualifying mentions across (prompts × LLMs × runs) for the selected date range. A single response can include multiple mentions; they are de-duplicated per response.

In plain terms. An estimate of how many real users likely saw your destination mentioned in AI answers this period.

Technically. Mention volume × estimated query frequency for each tracked prompt × LLM market share. Query frequency comes from public search-trend data and modeled LLM-specific intent share.

In plain terms. The percentage of AI mentions in your category that go to your destination versus competitors.

Technically. Your mention count divided by the sum of mention counts for the configured competitor set across the same prompts × LLMs × period.

In plain terms. How many different kinds of sources AI is citing about your destination (Wikipedia, GBP, Tripadvisor, press, Reddit, partner sites, etc.).

Technically. Distinct source-type count present in citations across the period. Higher diversity correlates with more resilient Presence and Recency scores.

In plain terms. How AI describes the research and inspiration phase: discovering the destination, comparing options, finding info.

Technically. Stage prompt-set covers information discovery, neighborhood selection, weather, packing, and price comparison.

In plain terms. How AI handles the booking phase: where to book, package vs. direct, channel confidence.

Technically. Stage prompt-set covers direct booking, channel selection, package bundling, and rate-parity language.

In plain terms. How AI describes the welcome experience, wayfinding, and first impressions.

Technically. Stage prompt-set covers welcome experience, wayfinding, transportation from airport, and local greeting.

In plain terms. How AI describes what visitors actually do once they're in town: partners, recommendations, concierge.

Technically. Stage prompt-set covers partner quality, recommendation accuracy, concierge helpfulness, and itinerary fit.

In plain terms. How AI describes the wrap-up: reviews, repeat intent, social sharing.

Technically. Stage prompt-set covers review volume, repeat-visit language, and social-sharing behavior.

In plain terms. Zertura's content service. We deploy writers into your content stack to produce the AI-readable owned content that lifts both Website Score and Perception Score. Partners enroll on a tier; the destination underwrites or co-funds the spend.

Technically. Subscription program with three tiers (T1 / T2 / T3) determining monthly content volume, channel mix (partner site, GBP, Wikipedia, earned editorial), and re-scoring cadence. Enrollment is per partner. Active enrollments contribute owned content to the destination composite; Pending and Completed enrollments are tracked separately and excluded from the headline 'active' count.

In plain terms. A flag on a partner card suggesting the destination consider enrolling that partner in the Content Deployment Program. Triggered automatically when a not-enrolled partner sits in the Critical or Watch band.

Technically. Surfaced on Partners and Overview when scoreBand(partner.davi) ∈ {critical, watch} AND partner.coopStatus === 'Not Enrolled'. Click-through routes to /deployment with the partner pre-selected.

In plain terms. The share of designated partners actively enrolled in Content Deployment Program. Higher coverage means a more accurate destination composite.

Technically. Enrolled partners ÷ total designated partners. Unenrolled partners contribute zero owned content to the destination composite.

In plain terms. A partner's individual DAVI score, calculated the same way as the destination score but scoped to their entity.

Technically. Same five-dimension composite, scored on prompts that mention or could mention this partner. Rolls into the destination composite weighted by partner Primary Designation.

In plain terms. A flag that marks an entity as a primary partner of the destination, eligible for Content Deployment Program programs and surfaced in scorecards.

Technically. Boolean attribute set by the DMO. Primary-designated entities are included in destination roll-ups; non-designated entities are tracked but not rolled up.