VectorGap Help Center

Use VectorGap in this order when you create a workspace or onboard a new client. #### What to do first 1. Create or select the brand from the sidebar brand selector. 2. Complete Brand Hub Settings: domain, industry, target generation, markets, languages, personas, and entity URLs. 3. Add known facts and proof in Knowledge Base so audits can judge AI answers against the right source of truth. 4. Run LLM Perception first to establish how AI providers currently answer about the brand. 5. Add tracked competitors, then run AI Preference or Share of Voice when you need competitive proof. 6. Convert repeated issues into Remediation Center missions and retest after fixes ship. #### Navigation - **Dashboard** is the account-level portfolio view. - **Intelligence** contains LLM Perception, AI Preference, Web Perception, GEO Optimization, Share of Voice, Query Explorer, and the AI Visibility Summary. - **Remediation Center** turns gaps into missions and tracks fix outcomes. - **Brand Hub** contains Knowledge Base and Settings. - **Community** links to release notes, roadmap, feature requests, and interviews. - **Billing & Usage** and **Profile** manage account operations. ---

Dashboard is the account-level portfolio surface for agencies and operators managing multiple brands. It is not meant to duplicate every brand-level audit page. #### What to do first 1. Scan brand cards for score movement, missing setup, weak audit channels, and competitor context. 2. Open the brand that needs attention. 3. Use the Intelligence child pages to inspect evidence, prompts, answers, and provider-level detail. #### What you will see - Latest available LLM Perception, Web Perception, GEO, AI Preference, and Share of Voice signals. - Trend indicators and compact context per brand. - Setup gaps that should be completed before interpreting scores too aggressively. ---

Intelligence is the brand-level monitoring area. It explains how AI providers and public sources currently represent the brand. #### What to do first 1. Start with Intelligence Overview to see which channel is weak. 2. Open the weakest channel: LLM, AI Preference, Web, GEO, Share of Voice, or Query Explorer. 3. Use AI Visibility Summary for the source-grounded narrative and next actions. 4. Convert actionable gaps into Remediation Center missions. ---

Intelligence Overview is the scannable brand command center. It keeps the channel cards visible first and moves the written narrative into the AI Visibility Summary side panel. #### Read it as - **LLM Perception**: Are AI answers about us accurate, positive, visible, credible, and recommendation-worthy? - **AI Preference**: Do AI systems choose us over competitors for buyer scenarios? - **Web Perception**: Does the public web support the story we want AI to repeat? - **GEO Optimization**: Can AI systems extract, trust, and cite our website and entity footprint? - **Share of Voice**: How much of the AI answer space do we own versus tracked competitors? ---

AI Visibility Summary is the board-ready narrative layer. It connects audit movement, evidence, open gaps, completed work, and next recommended actions. #### What to do first 1. Read the summary and confidence narrative. 2. Check what changed: improvements, declines, unresolved gaps, and unexplained movement. 3. Review channel summaries for the audit family behind each claim. 4. Export only after the evidence matches what you want to tell a client. Use the summary for strategy and reporting. Use the underlying audit pages for raw prompts, answers, provider scores, and technical evidence. ---

LLM Perception measures how ChatGPT, Claude, Gemini, Perplexity, Grok, and Mistral answer questions about the brand. It scores the answers across the current six-metric framework. #### What to do first 1. Run or review the latest LLM Perception audit. 2. Read provider performance before diving into individual prompts. 3. Use the Score source breakdown to see how each provider performs across Accuracy, Sentiment, Coverage, Credibility, Visibility, and Recommendation. 4. Inspect issues and repeated provider weaknesses. 5. Open Query Explorer when you need exact prompt and answer evidence. #### Key concepts - **BPI / AI answer score**: Composite score from accuracy, sentiment, coverage, credibility, visibility, and recommendation quality. - **Provider performance**: The configured six-provider panel: ChatGPT, Claude, Gemini, Perplexity, Grok, and Mistral. - **Issues**: Hallucinations, omissions, outdated facts, contradictions, weak recommendations, or missing proof. - **Competitor context**: Competitor rows are tracked inside the workspace and can be audited without creating separate standalone competitor brands. ---

AI Preference compares the brand against selected competitors in direct AI matchup prompts. It answers a different question from LLM Perception: not “Is the answer accurate?” but “Would AI recommend us over the alternatives?” #### What to do first 1. Select the competitors that matter for the current client, market, or buyer persona. 2. Choose the relevant audit preset so language, country, industry, and persona match the buyer context. 3. Run the matchup audit across the complete configured provider panel. 4. Review direct win rate, category weaknesses, provider-level gaps, and the rationale behind losses. 5. Turn repeated competitor advantages into missions. #### Key concepts - **Competitive preference score**: How often the AI panel favors your brand over the selected alternatives. - **Direct win rate**: Head-to-head wins in direct comparison prompts. - **Category gaps**: Buyer criteria where competitors have stronger evidence, clearer positioning, or more trusted sources. - **Provider breakdown**: Whether a weakness appears across the whole panel or mainly in one provider. ---

Web Perception compares the story the brand wants buyers to see with what public web sources currently support. #### What to do first 1. Check the main score and supporting cards for alignment, sentiment, source coverage, and gaps. 2. Identify which public sources weaken or contradict the intended story. 3. Use competitor web audits when you need public-positioning comparison. 4. Create missions for the clearest source, positioning, or proof gaps. Web Perception is the source-quality layer. It explains why AI systems may repeat, ignore, or distort specific claims. ---

GEO Optimization checks whether the website and entity footprint are easy for AI systems to extract, trust, and cite. #### What to do first 1. Review the latest GEO audit. 2. Start with Technical, Entity Health, Content Structure, Source Presence, and Consistency. 3. Open Issues, Citability, and Extractability for actionable details. 4. Add or repair schema, entity sources, factual consistency, and page structure. 5. Retest after fixes ship. #### Key concepts - **Technical**: Crawlability, metadata, schema, robots/LLM access, and page structure. - **Entity Health**: Whether the brand is represented clearly across owned and external sources. - **Content Structure**: Whether pages expose claims and answers in a parseable way. - **Source Presence**: Whether trusted third-party sources support the brand. - **Extractability**: Page-level ability to be parsed and summarized by AI systems. ---

Share of Voice shows how much of the AI answer space the brand owns compared with tracked competitors. #### What to do first 1. Start with the current distribution: brand share versus tracked competitors. 2. Compare provider/platform breakdowns to see whether the gap is model-specific. 3. Read trend movement over recent audits before claiming improvement. 4. Use AI Visibility Summary when you need the narrative behind share movement. #### How to read it - **Current distribution**: Latest split of AI answer mentions. - **Provider breakdown**: Where the brand is strong or weak across ChatGPT, Claude, Gemini, Perplexity, Grok, and Mistral. - **Competitor context**: Uses tracked competitors from the workspace, not public challenge invitations. - **Trend**: Directional proof that remediation or content work is changing AI visibility. ---

Competitors are tracked workspace records used across LLM Perception, AI Preference, Web Perception, Share of Voice, and reports. #### Current workflow 1. Configure competitors from Brand Hub Settings or the relevant competitor controls inside Intelligence. 2. Keep competitor records lightweight: name, domain, industry, and short description are enough for most audits. 3. Run competitor LLM or Web audits when you need benchmark evidence. 4. Use AI Preference for direct buyer-choice matchups. 5. Use Share of Voice for answer-space ownership and trend comparison. Competitors are no longer treated as a separate public challenge flow. They are private benchmark inputs for the workspace. ---

Query Explorer is the evidence surface for prompts and answers. It lets you inspect exact AI responses by audit, provider, template, status, prompt score, and audit score. #### What to do first 1. Choose the audit family or audit run you want to inspect. 2. Filter by provider when a weakness looks model-specific. 3. Use prompt score ranges to isolate bad or unusually strong answers. 4. Use audit score ranges to compare runs within a performance band. 5. Open rows for prompt text, answer excerpt, provider, score, latency, status, and evidence. Use Query Explorer when a client or stakeholder needs proof, not just a summary score. ---

Remediation Center is the cross-audit action queue. It ranks LLM, AI Preference, Web, GEO, and Share of Voice weaknesses by practical action value. #### What to do first 1. Sort by priority, effort, source, or detected date. 2. Open the action to read evidence, affected metric, and why the fix matters. 3. Create a mission when the gap is worth executing. 4. Attach evidence when external work ships. 5. Retest the relevant audit family after completion. This is where audit findings become work. Do not leave repeated gaps trapped inside individual audit pages. ---

Mission Control tracks structured remediation work. A mission connects a problem, objectives, execution steps, and retest outcome. #### What to do first 1. Create a mission from a recommendation, gap, or manual issue. 2. Define the target outcome and the audit family that should move. 3. Complete objectives as work ships. 4. Use Results & History after retesting to verify impact. Good missions are specific: fix one perception gap, source gap, comparison weakness, or extractability problem at a time. ---

Results & History is the archive of completed remediation work and proof-loop outcomes. #### Use it to - Show what was fixed and when. - Separate actions that moved scores from actions that did not. - Explain score movement to clients. - Find recurring issues that need broader positioning, source, or website changes. ---

Brand Hub is the operational home for a selected brand. It contains the Knowledge Base and Settings. #### What to do first 1. Keep Settings accurate: brand facts, markets, languages, personas, entity URLs, and competitors. 2. Use Audit presets to define repeatable market/language/persona targeting. 3. Add durable proof and source material to Knowledge Base. 4. Re-run audits after meaningful changes. ---

Knowledge Base stores trusted facts and source material that help VectorGap interpret audit results correctly. #### Current tabs and uses - **Neural Graph**: Entity relationships and evidence connections. - **Actions**: Mission-ready knowledge actions and AI Preference gap overlays. - **Perceptions**: History of how AI models represented the brand. - **Strategic Pillars**: Trust, Product, Service, Marketing, and Price signals. - **Docs**: Uploaded files and extracted source material. Knowledge does not magically raise scores by itself. It improves evaluation quality and helps identify where public proof is missing. ---

Settings controls the facts and targeting configuration that audits depend on. #### What to do first 1. Save the core profile: name, domain, industry, description, target generation, markets, language, and personas. 2. Configure Audit presets first in the right column so recurring audits use the right market, language, and persona. 3. Add entity URLs: LinkedIn, Crunchbase, Wikipedia/Wikidata, YouTube, X, Facebook, directories, media profiles, and other authoritative sources. 4. Configure tracked competitors. 5. Adjust workspace and alert settings only after the brand basics are correct. Audit presets are central because they control how AI audits model the buyer context. ---

Community links are product-feedback and communication surfaces. #### Current surfaces - **Release Notes**: What changed recently. - **Product Roadmap**: Planned and shipped product work. - **Feature Requests**: Submit or review product feedback. - **Book an Interview**: Schedule research or customer interviews when invited. These are not audit surfaces and do not affect scores directly. ---

Billing & Usage manages plan access, usage, and subscription operations. #### What to check - Current plan and workspace entitlement. - Audit usage and remaining capacity. - Brand, competitor, market, language, API, and automation limits where applicable. - Upgrade, downgrade, invoice, and payment management. The live Pricing page and Billing page are the source of truth for current packaging. Help text should not be treated as a price table. ---

Profile contains account-level settings and developer controls. #### Typical actions - Update account details. - Manage API keys if API access is enabled. - Manage webhook endpoints if workspace automation is enabled. - Review account security and session settings when available. ---

- **AI Preference**: Competitive matchup audit showing which brand AI providers prefer for buyer scenarios. - **BPI / AI answer score**: LLM Perception composite score across accuracy, sentiment, coverage, credibility, visibility, and recommendation quality. - **Competitor**: A tracked private benchmark record used across audits and reports. - **Extractability**: How easily AI systems can parse and summarize a page. - **GEO**: Generative Engine Optimization: making owned and external sources easier for AI systems to extract, trust, and cite. - **Knowledge Base**: Private trusted facts and documents used to evaluate audit results. - **LLM Perception**: How AI providers answer about the brand. - **Mission**: A structured remediation plan tied to a gap and retest outcome. - **Provider**: One AI system in the configured panel: ChatGPT, Claude, Gemini, Perplexity, Grok, or Mistral. - **Query Explorer**: Prompt-and-answer evidence viewer. - **Remediation Center**: Cross-audit action queue for turning findings into work. - **Share of Voice**: The brand's share of AI answer mentions versus tracked competitors.