Why it says "Possible X visitor": how attribution is graded

The honest-analytics page: what "Possible X visitor" actually means, which claims are verified versus inferred, and why we would rather hedge than guess.

When a card says "Possible Nike visitor" instead of "Nike watched your reel", that word is doing honest work. Every company claim in HoldFast carries an evidence grade, and the wording tells you the grade.

The grades, plainly

• Verified: a human confirmed it, your CRM matches the person, or the viewer identified themselves (an email gate, a personalized link, a signed comment). Stated without hedges, because it has earned it.
• Inferred: the evidence is real but indirect, most often the company network the view came from, or an email domain. Rendered as "Possible X visitor", because a corporate network cannot tell an employee from a contractor or a guest.
• Anonymous: not enough evidence to name anyone, so we do not. The view still counts in totals.

Why the hedge exists

Company-from-network lookups are the industry's standard trick, and they are wrong often enough that stating them as fact is a lie of confidence. HoldFast would rather show you a hedge you can calibrate on than a guess you will repeat in a client meeting. When the wording is unhedged, you can act on it; when it says "possible", treat it as a strong lead.

The same discipline runs through alerts: live-presence pings fire only for verified contacts, by design, so your phone never buzzes for a maybe.

Moving claims up the ladder

Verification compounds. Every email gate passage, personalized link open, and signed comment creates verified identity, and every company you confirm in the Explorer becomes permanent truth the system never second-guesses. The hedged remainder is exactly the traffic that deserved a hedge.