Sample deliverable

Evidence Audit — sample structure

This page shows the shape of a $249 Evidence Audit (holiday launch pricing, ends Dec 31, 2026; standard $497 from Jan 1, 2027). Real audits are built from dated, raw captures. We do not publish a company's results without its written permission, so the result cells on this page are placeholders, not answers. Citebrim will not invent AI answers.

Structure only. Nothing in the tables below is a real Google AI Overview, ChatGPT, Perplexity, Gemini or Grok response. A finished audit fills every cell for each engine in scope with dated raw captures; an engine not measured is labelled “not measured”, never estimated.
Prepared for
Example B2B SaaS company
Status
Sample structure — placeholder cells
Scope
1 brand, 3 competitors, 10 prompts × 3 captures on each engine in scope (up to 5)
Turnaround
5 business days, or a full refund

1. What you receive

  • Ten locked buyer prompts, three captures each on every engine in scope (up to five: Google AI Overviews, ChatGPT, Perplexity, Gemini and Grok), confirmed in writing before your start date.
  • A comparison against three competitors that actually appeared in the captures.
  • Source links returned with the answers, mapped by page type.
  • A written fix plan.
  • One implementation-ready page rewrite, delivered as a draft for your approval.
  • The raw answer archive, labelled as provider-observed responses.

2. Method

Prompts are written as a buyer would ask them, then locked before the first capture. Each prompt is sent to each engine three times. We record the model or API name, the date and time (America/New_York), whether web search or grounding was on, and the full text. Brand names are matched against a prepared alias list. A citation is recorded only when a source URL is present.

API answers can differ from consumer apps. The finished report labels the engine and surface used for every cell. We do not claim that a page change caused a mention. We show what was true before, what changed, and what was true after.

3. Example prompt bank (illustrative)

A real audit uses prompts tailored to the company’s category. These are examples of the type, not results.

  1. Best {category} software for small teams
  2. Best {category} software for mid-market B2B teams
  3. Best {category} software with transparent pricing
  4. {Incumbent} alternatives for mid-market teams
  5. Which {category} tool is easiest to implement?
  6. Best {category} software for {use case}
  7. Who has the best {category} reporting for small B2B teams?
  8. Best affordable {category} software for a team of 20
  9. What {category} tools do people recommend in 2026?
  10. Compare the leading {category} options for small teams

4. Visibility table

In a finished audit this table is filled from archived, provider-observed responses. Here it shows placeholders.

Prompt Google AI Overview ChatGPT Gemini Perplexity Grok
Best {category} software for small teams Placeholder Placeholder Placeholder Placeholder Placeholder
{Incumbent} alternatives for mid-market teams Placeholder Placeholder Placeholder Placeholder Placeholder

5. Competitor comparison

A finished audit names three competitors that actually appeared in the captures, not a generic list. On this sample page the rows are placeholders.

Brand Named in captures Cited with a URL Notes
The companyPlaceholderPlaceholderPlaceholder
Competitor APlaceholderPlaceholderNamed only from live captures
Competitor BPlaceholderPlaceholderNamed only from live captures
Competitor CPlaceholderPlaceholderNamed only from live captures

6. Fix plan and page rewrite

A delivered audit includes a prioritized fix plan and one implementation-ready rewrite, submitted as a draft. It is published only with your approval. Typical fix types — not findings about a specific company — include a clearer public pricing page, a current comparison page against the competitors that actually appear in AI answers, and a product-facts page a model can quote. We do not invent reviews or undisclosed posts as a “fix.”

7. Disclosures

  • No guarantee of mentions, citations, rankings, traffic or leads.
  • If we miss the five-business-day deadline, you receive a full refund. Factual errors are fixed free.
  • Each engine in scope is named in the report with the model, date and whether web search was on. An engine that was not measured is marked “not measured”, never estimated.
  • If you start Core (Visibility Operations) within 14 days of delivery, the audit fee is credited once: your first monthly invoice is $1,001 at holiday launch pricing ($993 at standard pricing from January 1, 2027). If you order the audit by December 31, 2026, the $1,250 launch rate and the $1,001 first month still apply when Core starts in January, within 14 days of delivery. The credit comes as a single-use code sent with your audit.