Primenza Methodology

Primenza is a measurement and optimization platform for technical SEO, Generative Engine Optimization (GEO), and AI search visibility. This page explains what Primenza measures, how scores are produced, and what those scores do not mean.

What Primenza Measures

Primenza measures website signals relevant to classic search engines and AI answer systems: crawlability, on-page SEO, GEO readiness (readability, structure, citation readiness, entity clarity, trust signals), AI visibility signals where your plan and configured providers allow, and optional Market Intelligence prompts across supported AI providers.

SEO Measurement

SEO audits analyze crawled pages for technical, on-page, and related issues. Scores are derived from deterministic checks against the fetched HTML and site structure — including metadata, headings, links, performance signals where available, and configured audit rules. SEO scores reflect audit findings on the measured pages, not guaranteed Google rankings.

GEO Measurement

GEO scoring evaluates how well content can be understood and cited by AI answer engines. Primenza groups GEO metrics into four pillars: Readability (AI readability, semantic clarity, topic coverage), Citation (citation readiness, answer quality, source credibility), Structure (entity recognition, structured information, information hierarchy, content chunking), and Trust (hallucination resistance, fact consistency, expert signals). Pillar scores are aggregated from page-level metric measurements during crawls.

AI Search Visibility

AI search visibility in Primenza refers to whether a brand or domain appears to be discoverable, mentioned, or cited in AI-mediated answers — based on configured modules, simulations, and (where enabled) Market Intelligence provider runs. Primenza does not control AI engine rankings or guarantee inclusion in any model output.

AI Citations and Mentions

A citation is when an AI answer references a specific URL or source attributable to your domain. A mention is when your brand or product name appears in an answer without a direct link. Primenza records these when they are observed in supported measurement flows. A single response is not treated as definitive; repeated runs, timestamps, and provider context are stored where the module supports them.

Observed vs Estimated vs Simulated vs Inferred

OBSERVED

Directly measured from a fetched page, crawl result, HTTP response, or a stored AI provider response at a recorded timestamp.

ESTIMATED

Calculated from documented formulas or models using observed inputs — for example, aggregated pillar scores or visibility rates across multiple runs.

SIMULATED

Produced by controlled simulation or heuristic analysis when live provider access is unavailable or when the product explicitly runs a simulation module.

INFERRED

Conclusions drawn from available evidence — for example, inferring mention context from parsed answer text rather than accepting provider-reported scores as authoritative.

How Scores Are Calculated

Primenza scores are normalized indicators on a 0–100 scale for comparability inside the product. They summarize audit and measurement outputs for prioritization. A score of 83/100 means Primenza assessed strong readiness on the measured dimensions — it is not a promise of ranking position, traffic, conversions, or AI citation probability.

Data Sources

Sources include Primenza crawls, live HTTP fetches, on-page HTML analysis, configured platform catalog data, plan-gated provider APIs where credentials are configured, and user-supplied domain settings. Market Intelligence measurements store engine/model, prompt, timestamp, and algorithm version where applicable.

Limitations

AI responses are probabilistic and can change without notice. Provider availability, plan limits, crawl depth, and sampling affect results. Homepage-only comparisons differ from full-site audits. Primenza cannot observe every AI query or every search engine index state.

How Often Measurements Change

Scores update when you run or schedule a new audit, crawl, or visibility run. Historical results remain timestamped in your workspace. Public marketing counters may include configured offsets for display purposes and are labeled separately from per-domain audit scores.

Research & evidence

Primenza publishes a small set of reference guides and OBSERVED research snapshots with explicit methodology, sample scope, and limitations. These are not mass-produced content marketing — they document how the product measures SEO, GEO, and AI visibility.

Frequently asked questions