Being named is not the same as being named first.

Across 594,486 recorded mentions in this study, the average position at which a company appeared within an engine's answer was 5.15. Broken out by engine, the average position was 5.23 on ChatGPT, 5.34 on Claude, and 4.83 on Gemini.

Why position matters more than raw appearance. A generated answer that lists several companies is read the way any ranked list is read: attention concentrates on the first few entries and drops off after that. A company mentioned consistently but always toward the back of the list has, in practical terms, a different commercial outcome than a company mentioned first, even though both would count identically in a simple yes-or-no appearance rate. This is the same dynamic that made position, not just indexing, the relevant metric in conventional search results for the past two decades, and there is no obvious reason AI-generated answers would behave differently, since the underlying human behavior, reading the first items and skimming or ignoring the rest, has not changed.

Reading the small gap between engines. Gemini's average position, 4.83, is somewhat earlier than Claude's, 5.34, with ChatGPT in between at 5.23. Combined with the earlier finding that Gemini names cohort companies less often overall, one plausible reading is that Gemini is more selective about which companies it names at all, and having been more selective, tends to put the ones it does name closer to the front. An alternative reading is that Gemini's answers are simply shorter on average, with fewer total entries in a typical list, which would mechanically pull the average position earlier without implying anything about selectivity. This study's position data alone cannot distinguish those two explanations; that would require also measuring the typical length of each engine's answer list, which is not part of this month's published fact set.

What a company does with this. Position is a more demanding target than presence, and moving it requires different evidence than simply getting mentioned once does. A company already appearing somewhere in most answers, but consistently in the back half of the list, should treat that as a distinct problem from a company that never appears at all. The lever most likely to move position, based on how these systems are understood to work, is the strength and specificity of third-party corroboration, independent sources describing the same facts about a company consistently, rather than anything the company alone publishes about itself, though this study does not yet have a populated source-citation table to confirm which sources are doing that work in this particular run.