Claude is included in the daily run of your question list. We record whether it names you, how it describes you and whether it points to sources. Its answers may differ in length and caution from other assistants, so we compare like with like and avoid reading too much into style differences.
Some questions get a careful, hedged reply that names no one. That is useful information in itself. It can mean the assistant lacks enough trusted material about your category, which the source mapping agent can then help to investigate. Your team decides which of those gaps deserve attention first.
Claude results sit in the same share of voice reports as the others, so no assistant is judged in isolation. If your customers mainly use a different assistant, we weight the report toward it and tell you where we did so. Changes are always reported openly.
Claude results show how one assistant describes your category and whether it names you. They do not show how many people saw the answer, and they do not prove that a change in your content caused a change in the reply. We report what was observed and explain the limits plainly. Used with the other assistants, Claude results help you see whether a gap is specific to one model or a wider pattern worth acting on.