LLM perception drift vs AI visibility
AI visibility usually measures whether an entity appears, is cited, is recommended, or is associated with a query in a generative answer.
LLM perception drift measures something else: how that entity is reconstructed and how this reconstruction varies over time or across models.
Why visibility is not enough
A visible brand can be misunderstood. It may be cited in an answer while being associated with an overly generic category. It may be recommended while its differentiators disappear. It may obtain strong share of voice, but on an intent that does not match its real positioning.
Visibility answers the question “are we present?”. Perception drift answers the question “which version of us is being produced?”.
Different measurement
A visibility tracker can count mentions, citations, links, rank, or sentiment. A perception drift audit must compare those outputs with a canon and verify fidelity of role, perimeter, category, evidence, and exclusions.
This is not a competing metric. It is a higher layer. Visibility indicates access to the answer. Perception stability indicates the quality of representation.
Reading rule
Use AI visibility to measure presence. Use LLM perception drift to measure representation variation. Do not confuse an increase in mentions with an increase in fidelity.