CLAIMS & LIMITATIONS

Our evidence standard.

Independent of every AI platform. Every report reviewed by a person. No promised rankings. Every SeenPoint finding is built on recorded evidence. Here is exactly what you get — and how we measure it — so you can act with confidence.

Last updated: 04 October 2026

WHAT YOU GET

INSTEAD OFYOU GET
01Speculation about hidden model reasoningWhat AI search environments actually returned — captured, retained and classified.
02Claims about model-internal ranking factorsObservable outcomes: mentions, recommendations, citations, sources and placement.
03A universal AI visibility scoreCounts you can check, by question and environment, every one traceable.
04Promised rankingsEvidence of where you appear, and work specified to improve it.
05Promised recommendationsThe gaps where competitors are recommended, with specified work to address each one.
06Promised citationsThe sources AI search actually cites, and where you're missing from them.
07Promised revenueCommercial questions prioritised by buyer intent.
08Promised ROIA matched retest showing what changed after the work.
09Before-and-after screenshotsMatched retests under comparable conditions, reported as CHANGE OBSERVED, NO MATERIAL CHANGE or INSUFFICIENT EVIDENCE.
10One answer treated as everyone's answerMarket, location, language and session recorded for every observation.
11Platform tie-insIndependent measurement. SeenPoint is not affiliated with any AI platform.
12Vague "always-on" trackingScheduled Pulse observations and matched retests on your locked questions.
13Unattended automationHuman-led, AI-assisted work, with every report reviewed by a person.

HOW WE LABEL EVERY STATEMENT

OBSERVED
Recorded directly in an observation or source.
CALCULATED
Derived from recorded observations.
SUPPORTED INFERENCE
A conclusion the evidence supports.
HYPOTHESIS
A plausible explanation we test before acting on it.
RECOMMENDATION
The specific action we advise, tied to its evidence.

HOW WE REPORT OUTCOMES

CHANGE OBSERVED
The retest shows a measured change.
NO MATERIAL CHANGE
The retest shows no meaningful movement, so you know where to redirect effort.
INSUFFICIENT EVIDENCE
The difference is too small to call yet, so you know to keep measuring.
EXAMPLES ON THIS WEBSITEExamples on this website are illustrative and use fictional data. Each one is clearly labelled ILLUSTRATIVE EXAMPLE — FICTIONAL DATA.

When the evidence points to an opportunity, you'll see exactly what it rests on.

Get Your AI Search Report →