Skip to main content
Average Position is the average rank your brand receives when AI lists products in its answer: (sum of all your positions ÷ number of responses where you appeared). Erlin records the position number for every response that ranks your brand and averages across them. Lower is better. It is the AI-search version of average rank on a results page.

​
Why it matters

Users act on the first recommendation AI gives them. Position 1 is what gets clicked, bookmarked, and remembered. Position 5 is background noise. Moving from position 4 to position 2 does not double your visibility, but it dramatically increases how often users treat your brand as the answer rather than one of several options.

​
What Average Position tells you

Average Position measures placement. A score of 1.5 means AI almost always leads with your brand. A score of 5.0 means you are listed but buried beneath four competitors. Position shifts are often the earliest signal of a change: a move from 2 to 4 over two weeks frequently precedes a drop in Visibility Score on those same prompts.

​
How Erlin calculates it

For every AI response that gives your brand a ranked position, Erlin records the number. The formula:
Responses where your brand was not ranked are excluded. Erlin uses two signals to determine position:
  1. Product list order: when AI outputs a numbered or ordered list of recommendations, your position in that list is recorded directly.
  2. Citation start index (fallback): when no numbered list exists but AI cites sources, Erlin uses the order in which your domain’s citations appear relative to other domains.
Worked example: across 6 responses where your brand appeared, you were ranked 3rd, 1st, 2nd, 4th, 1st, and 3rd. (3 + 1 + 2 + 4 + 1 + 3) ÷ 6 = 2.3

​
How to read it in Erlin

A falling Average Position means AI is ranking you higher. A rising number means you are slipping down the list. Compare platforms to find where you rank worst: a position of 1.5 on Perplexity but 4.0 on Gemini tells you which platform’s source data needs attention. Check the Competitor Leaderboard to see who ranks above you and whether their position is improving.

​
How to improve it

Be the obvious answer for specific queries. Position 1 goes to the brand AI is most confident fits the question’s exact framing. Owning “best CRM for solo founders” is more achievable than competing for generic “best CRM.” Target prompts where your product has a clear claim. Strengthen citations on low-position prompts. Cross-check the prompts where your position is worst against your citation coverage for those same prompts. Position and citation tend to move together: if AI cites your content for a topic, it tends to rank you higher on that topic. Watch competitor positions for early signals. In the Competitor Leaderboard, sort by Average Position. If a competitor’s position is improving while yours is stable, they may be gaining prominence that will eventually affect your other metrics. Act before it compounds.

Visibility Score

How often AI mentions your brand across tracked prompts.

Citation Rate

Your share of the sources AI cites in its answers.

Share of Voice

When AI lists products, how big a slice your brand takes.

Sentiment

The tone of AI mentions about your brand.