Unlike a search results page, most AI answers do not present a fixed, numbered list, which makes position harder to define precisely. Trackers typically approximate it by recording whether a brand appears in the opening sentence, within the main recommendation, in a supporting list, or only in a closing aside, then scoring those tiers so position can be compared across queries and over time.
Position matters because users reading a synthesized answer tend to give more weight to whatever is mentioned first or framed as the top recommendation, similar to the well-documented primacy effect in ranked search results. A brand that is technically mentioned but placed last, or listed only as a minor alternative, captures far less consideration than one named as the leading option.