Skip to main content
Position measures where your mentions land inside the answer compared to your competitors. Being mentioned 9th in a “top 10” list is very different from being the first recommendation — and being number 1 often means you’re the AI model’s preferred choice. Position captures that nuance.

​
How it’s calculated

3RD scans each response for numbered lists (e.g. “1. Brand A, 2. Brand B”). Your brand’s line number in that list is your position. If there’s no numbered list, it falls back to the order of first mention — first brand mentioned = position 1, second = position 2, and so on. Lower is better; position 1 is the lead recommendation. Average position is the mean across all responses where your brand was ranked.
Position is only calculated against your tracked competitors — brands you don’t track aren’t counted in the ordering. That’s another reason your competitor set matters.

​
How to read it

Pair position with visibility:
PatternWhat it means
High visibility + good positionYou’re the recommendation. Defend it.
High visibility + poor positionYou’re always in the room but never the favorite — often an easier fix than absence.
Low visibilityPosition doesn’t matter yet — get mentioned first.
Track average position per prompt type and per model — models order lists differently, and losing the top spot in one model may be masked by winning it in another.

​
How to improve it

  • Make your differentiators explicit and quotable for each prompt’s specific intent.
  • Win citations on comparison and “best of” sources — models draw their ordering from them.
  • Put your strongest proof points (pricing, outcomes, integrations) on extractable pages.
Don’t confuse Position with Rank. Position = your place inside one AI answer. Rank = your overall standing across everything you track.