Choose the buyer questions
We agree on the category, buyer stage, geography, and questions before collection. A changed question becomes a new record instead of rewriting history.
How the X-Ray works
Every claim should have a trail.
Foremention keeps four things separate: what we saw, what we think it means, what we did, and what happened later.
Save the answer evidence
We review the same questions across the agreed AI tools. We keep the answer, date, visible model label, brand order, and exact cited pages.
Resolve source entities
URLs are normalized without collapsing distinct pages. We record source type, crawler access, brands present, recurrence, and the answer evidence behind each score.
Build the Source Map
Sources are ranked using observable recurrence and relevance. Entry feasibility is a human judgment and is labelled separately from evidence.
Qualify legitimate routes
We exclude fabricated reviews, undisclosed promotion, link schemes, false identities, pay-to-play coverage presented as editorial, and editorial-pressure tactics.
Track what happens
An opportunity moves through found, checked, pitched, accepted, published, indexed, first cited, repeatedly cited, and lost states.
Reporting model
Three layers. Never one magic score.
Work completed
Sources checked, pitches sent, placements published, pages updated.
Change observed
Index status, first-cited date, answer coverage, first-mention share, citation survival.
Business signal
AI referral traffic, branded search, qualified demos, assisted pipeline.