Metrics Glossary

Every A2Z Reach metric explained in plain English, with a real example for each.

Visibility Score

How often your brand shows up in AI answers — like your market share of AI conversations.

The percentage of tracked prompts where your brand was mentioned at least once. If you track 100 prompts and appear in 45, your Visibility Score is 45.

Example: Allbirds has a Visibility Score of 45 — meaning it appears in 45% of running-shoe-related AI conversations. The other 55% don't mention it at all.

Share of Voice (SoV)

Your slice of the conversation pie — your mentions vs everyone else's.

Of all the brand mentions across your tracked prompts, what percentage are yours? If a category has 1,000 total brand mentions and 107 are yours, your SoV is 10.7%.

Example: Allbirds has 10.7% Share of Voice. Nike, Hoka, and On split most of the rest. Higher SoV = you dominate the AI conversation in your category.

Position

Where you appear in the answer — first pick, supporting mention, or final verdict.

AI responses have zones: intro (the opening recommendation), body (supporting detail), and verdict (the closing pick). Being mentioned in the intro is strongest — it's the AI's top-of-mind answer.

Example: Allbirds appears in position #1 in the intro for 'sustainable running shoes' — the strongest possible placement. For 'marathon training shoes' it only appears in the body, behind Nike.

Exclusion / Exclusion Rate

How often AI leaves you out entirely while recommending competitors.

An exclusion is any prompt where your brand was absent but competitors appeared. Exclusion Rate is the percentage of prompts where this happens. This is your biggest opportunity signal.

Example: Allbirds was excluded from 10 prompts last week. In 'fastest growing athletic brands', On, Hoka, and New Balance appeared — Allbirds didn't. The fix: earn coverage on the sources that prompt cited.

Citation / Source

The websites AI quotes when it builds an answer — the references behind the recommendation.

When AI recommends a brand, it cites sources (review sites, roundups, articles) that back the claim. These citations reveal which publications influence AI in your category. Getting featured on highly-cited sources is the #1 way to improve visibility.

Example: For running shoes, ChatGPT frequently cites runnersworld.com, rtings.com, and tomsguide.com. If you're not on these sites, you're hard to recommend.

Inline vs Reference Citation

Inline = quoted mid-sentence. Reference = listed at the bottom.

Inline citations appear directly next to a claim ('Allbirds uses eucalyptus fiber [tomsguide.com]'). Reference citations are listed separately as sources. Inline citations carry more weight — they directly support a specific statement.

Example: A response might have 8 inline citations (woven into the text) and 8 reference citations (listed at the end) — 16 total sources.

Comparison Appearance

Did you show up in a head-to-head comparison the AI made?

AI often compares brands directly ('Allbirds vs Nike'). Comparison appearance tracks whether you're included in these comparisons — and whether you win or lose them.

Example: Allbirds appeared in a comparison for 'Allbirds vs Nike for running' — included and described favorably on sustainability, less so on performance.

Gap Type

The specific reason you were excluded — so you know exactly what to fix.

When excluded, we diagnose why: editorial_coverage (missing from cited articles), comparison_presence (not in comparison tables), review_authority (no expert reviews), content_recency (your content is stale), category_mismatch (AI doesn't link you to this category).

Example: Allbirds' top gap is 'editorial_coverage' — the publications AI cites for athletic brands don't feature Allbirds. Fix: PR outreach to those specific outlets.