Evidence job
Open the answer behind the score
Query Explorer prevents black-box reporting by preserving the prompt, provider, category, source context, and answer evidence.
Inspect the prompt and answer behind every client-facing claim.
Query Explorer is the audit evidence layer for agencies. It lets a strategist open the provider answer, prompt category, score, source context, status, market, language, persona, and competitor evidence before a finding becomes a client-ready recommendation. Query Explorer only reduces reporting risk when the team can see both what AI said and which web evidence can support, contradict, or repair that answer.
- Prompt rows with provider, category, score, status, and answer excerpt.
- Source and citation evidence tied to market, language, persona, and competitor context.
- Failure, hallucination, unsupported-claim, and source-quality signals made inspectable.
- Report appendices, mission targets, and retest lists built from the same evidence rows.
7
standard providers
98
LLM prompts
40
Preference prompts
0-100
Generative Brand Index
Agency use
Find the weakest prompt
Strategists can sort by provider, category, status, and target context to locate the exact answer that needs a page, source, or entity fix.
Report path
Excerpt → appendix → mission
Client-safe answer excerpts become report evidence, remediation missions, and comparable retest targets.
What is Query Explorer?
Query Explorer is the drilldown interface that connects VectorGap scores to the actual AI answers that produced them. Agencies use it when a client asks where a recommendation, risk, hallucination, or competitor loss came from. Instead of defending a vague score, the team can inspect the prompt family, provider, answer excerpt, source status, target market, prompt language, persona, competitor context, and scoring notes that explain the result.
Provider-level answer evidence
Every major provider can describe the same brand differently. Query Explorer keeps answer excerpts and provider labels close to the score so an agency can see whether ChatGPT, Claude, Gemini, Perplexity, Grok, Mistral, or DeepSeek omitted the brand, misclassified it, used weak evidence, or recommended a competitor for a specific reason.
- Answer excerpt
- Provider and category
- Hallucination or source caveat
Prompt and category filters
Prompt families become useful when they are grouped by buyer intent: trust, pricing, legitimacy, products, capabilities, recommendations, alternatives, problems, and reviews. Agencies can filter by audit, provider, category, market, language, persona, status, and competitor context before deciding which public proof to improve.
- Prompt family
- Market and language
- Persona and competitor context
Exportable appendix
Detailed query rows are the raw material for a client-safe appendix. The agency can redact sensitive prompt text, preserve provider and category context, include answer excerpts, identify source gaps, and turn weak rows into mission targets that can be retested after the fix ships.
- Evidence appendix
- Client-safe excerpts
- Retest target list
How do agencies use Query Explorer after a weak audit?
The practical path is to open the weakest prompt or category, read the provider answer, identify whether the issue is missing proof, poor source context, entity confusion, unclear pricing, or competitor preference, then create a mission for the page or source that can change the next answer. Query Explorer gives the team enough evidence to decide whether to rewrite a page, add schema, publish a proof block, strengthen third-party sources, or retest a different competitor set.
Diagnose the failure layer
A low score can come from different causes: no answer, generic category label, hallucinated feature, weak citation, outdated price, missing source, unsupported competitor strength, or a prompt that does not match the client’s actual market. Query Explorer separates those failure types before the agency recommends work.
Create the mission target
Once the weak row is clear, the agency can attach a target page, evidence URL, expected metric, prompt family, provider, and retest instruction. That keeps remediation work tied to the original answer instead of becoming a generic content task.
Retest the same evidence
After the page, schema, source, or proof fix ships, the agency can rerun the same target and compare answer quality, source use, competitor ranking, and hallucination risk. That turns Query Explorer into before/after proof rather than a one-time debugging view.
Evidence rows → report proof → Agency OS
Query Explorer only reduces reporting risk when answer evidence becomes a report appendix, comparable retest, and recurring delivery capacity. The buyer should inspect how evidence rows appear in a client-safe report, review pricing proof-before-payment, then buy the Agency OS capacity to inspect, assign, retest, export, and report across clients.
Inspect evidence report proof
Show how answer excerpts, source notes, competitor context, mission targets, and same-target retests become the client-ready evidence appendix.
Open pageReview proof before payment
Use pricing proof-before-payment to connect Query Explorer evidence to Generative Brand Index, Preference proof, Mission Control actions, and retest plans.
Open pageBuy inspection capacity
Agency OS gives the team recurring capacity to inspect prompt evidence, create missions, retest comparable targets, export reports, and defend client claims.
Open pageWhich evidence does AI need before it can reuse the page?
This table turns the page into a structured extraction target: the buyer question, the evidence an AI system can read, and the action an agency can sell or execute next.
| Buyer question | What AI can extract | Agency action |
|---|---|---|
| Where did this score come from? | Provider, prompt family, answer excerpt, status, source context, and category scoring notes. | Use the row as report evidence and identify the exact public claim that needs repair. |
| Why did a competitor win? | Direct matchup context, rationale, cited or missing sources, and category-specific preference reasons. | Create a comparison-safe proof mission and retest the same competitor set. |
| What should be fixed next? | Weak prompt clusters, unsupported answers, source gaps, hallucinations, and missing entity facts. | Prioritize page structure, schema, proof, source, or Brand Knowledge cleanup. |
Questions agencies ask before turning AI visibility into client work
Why does an agency need Query Explorer instead of only a summary score?
A summary score is useful for executives, but the delivery team needs the underlying answer. Query Explorer shows the prompt category, provider, answer excerpt, source context, score, market, language, persona, and competitor evidence so the agency can explain why the score moved and what should be fixed.
Can Query Explorer be used in client reports?
Yes. Agencies can use client-safe excerpts, provider labels, category summaries, source notes, and retest targets as an appendix or internal QA layer. Sensitive prompt wording can stay hidden while the evidence chain remains clear enough for a client conversation.
How does Query Explorer connect to Mission Control?
The weak row identifies the mission: which page, proof block, source, schema, entity fact, or competitor context should be improved. The retest then checks whether the same provider and prompt family changed after the work shipped.