Observed Measurement Notebook

See how your brand shows up in AI answers — with a timestamped, explainable snapshot.

AI Rank Watch is a checker-first visibility tool for marketers who want proof, caveats, and source coverage — not vague promises. This prototype stays in demo/sample mode until a live source is explicitly approved.

Observed snapshotSelected promptsSource coverageTimestampedIndependent tool
Demo/sample mode · no dollar prices · beta/waitlist only · email delivery intent separated from weekly monitoring opt-in
Checker

Run an observed AI visibility snapshot

DEMO/SAMPLE

This prototype does not call any live provider. If no approved source exists, the UI falls back to demo/sample mode and shows explicit limitations.

Observed snapshot · Selected prompts · Source coverage · Timestamped report · Independent tool
Result preview

Explainable visibility, not magic rank claims

Every score comes with source coverage, caveats, and a methodology link so the user can understand what was observed and what was not.

72/100
Timestamp: 2026-08-10 18:45 UTC
Source coverage: prompt-only demo fixture
Limitations: no live provider, no official affiliation, no guarantee of appearance
Mentions presentCitations partialCompetitor strongerStance neutral
PresenceBrand detected in AI output

Counts whether the brand name appears in selected answer fixtures.

CitationsSome source links observed

Link density and citation behavior vary by prompt and source mode.

CompetitorsOne competitor stronger

Relative presence is visible in the fixture, not a live market claim.

MethodObserved, timestamped, demo-only

The score is explainable, but not a promise of deterministic rank.

States

Checker states that stay honest

Every state is visibly labeled so demo-only, queued, limited, or unavailable results never look like live provider data.

demo-only
Default mode

Preview methodology with a fictional brand and sample result.

queued
Capacity

Request placed in waitlist/manual approval queue.

limited
Partial data

Show what is missing rather than inventing a full result.

no-approved-source
Source gated

No live provider has been approved yet; stay in demo/sample.

completed
Approved run

Used only if a later source mode is explicitly approved.

error
Error handling

Surface the failure with a clear retry path.

How it works

Checker → report → email/waitlist

  1. Choose brand, domain, and prompt pack.
  2. Preview a demo/sample methodology or an approved run later.
  3. Read the timestamped report and source coverage.
  4. Email the report or join beta for weekly monitoring.
Email beta

Separate consent for report delivery and weekly monitoring

Emailing this demo report is separate from recurring monitoring. Weekly monitoring remains optional, waitlist-only, and unchecked by default.

Report delivery consent boundary: clicking “Email this report” requests one-time report delivery only; it does not opt you into weekly or recurring emails. If no outbound email provider is configured, the request may be queued, handled manually, or marked provider-not-configured instead of claiming an email was sent. Report delivery records, waitlist records, logs, and source outputs are retained up to 1 week. For deletion, privacy, or unsubscribe requests, contact shenyanying33@gmail.com. See Privacy and Terms.

Methodology

Transparent score model and source limits

Visible caveats are part of the interface, not only the footer.

Score components include presence, citation behavior, prominence, competitor comparison, and stance. The prototype labels source mode and limitations near the score.

Do not imply deterministic rank, guaranteed improvement, all-engine coverage, or official affiliation. The default state is demo/sample until R6/R10 approval.