Prompt selection (ten commercial buyer questions)
We work from intake and sales language to choose ten pre-purchase questions — not branded vanity queries. You can suggest prompts; we finalize the list before testing.
Transparent process
Citation Action Audit methodology
We test ten commercial buyer questions across major AI answer engines, preserve which competitors and sources appear, compare the winning pages with yours, and deliver the three page changes worth shipping first.
We test across the major answer engines above when accessible from our testing setup. Exact engine coverage, account tier, and geography are confirmed during intake — availability varies by engine, session, and location.
Process
Engine coverage and testing conditions
We test across the major answer engines above when accessible from our testing setup. Exact engine coverage, account tier, and geography are confirmed during intake — availability varies by engine, session, and location. We record date, time, engine, account context when visible, and geography used for each run.
Response and citation capture
For each prompt and engine, we preserve screenshots, brands mentioned, and linked source URLs. Evidence is dated and included in your deliverable.
Competitor and source detection
We note which competitors appear, which URLs are cited, and whether your brand or domain is mentioned or linked.
Your pages vs winning pages
For prompts you lose, we open cited competitor URLs and your closest page. We document structural gaps: answer-first copy, proof, schema, comparison tables, citable numbers.
Human validation
Automated rank scores are not used. A human reviews each observation before it enters your report.
Three-action prioritization
We rank fixes by buyer intent, effort, and how often competitors repeat across prompts. You receive three page changes worth shipping first, each with a copy-ready brief.
Limitations you should expect
AI answers vary by engine, session, account tier, location, and time. We date-stamp and screenshot every observation. A rerun is a new observation — it does not prove that a page change caused an engine to respond differently.
What evidence you receive
A written PDF and editable doc with dated screenshots per prompt and engine, tables of brands and cited URLs, side-by-side notes on your pages vs competitor pages, and three copy-ready page change briefs prioritized for shipping first.
AI answers vary by engine, session, account tier, location, and time. We date-stamp and screenshot every observation. A rerun is a new observation — it does not prove that a page change caused an engine to respond differently.
Get the audit for your domain
$299 one-time · 48 business hours after complete intake
Buy Citation Action Audit — $299
One-time founder-led audit · Delivered within 48 business hours of complete intake