Understanding your first scan

How to read the Overview after your first scan, and which conclusions to wait on.

The scan fills in stage by stage

Your first scan starts as soon as you finish onboarding. The dashboard does not wait for the whole run: modules appear as their data lands, so you can start reading results within minutes.

Start on the Overview

The Overview page is the summary of everything Recited knows:

  • The stat strip: your AI Visibility score with its two parts (topic coverage and mention strength), your share of voice with your top competitor's share, and how many answers cite your domain.
  • Leaderboard: every brand found in your answers, ranked. You are always in the list, even at zero, because knowing you are at zero is the point.
  • By platform: how your visibility splits across the platforms you track.
  • Where answers come from: the domains and pages AI cited most, plus the mix of publisher types and page formats.
  • Since last scan: the biggest movers. Empty after one scan, useful from the second scan on.
  • Latest answers: the newest captured chats.

Read some actual answers

Before drawing conclusions from the numbers, open five or ten chats in the answer log. Read how the AI describes your category, who it names first, and which sources it cites. The metrics will make much more sense afterward.

What is normal on day one

  • A low score is normal. Most brands discover they are invisible for most prompts. That is the gap the rest of the product helps you close.
  • Zero visibility on some prompts is normal. Check whether any competitor is visible there either. If nobody is, the prompt may be too obscure. If competitors are and you are not, that is a real gap, and it will show up in Missing citations and opportunities.
  • Platforms disagree. The same prompt can name you on one platform and ignore you on another. That is signal, not noise.

What to wait on

Deltas and trends need history. Every table shows change since the previous scan, but a single day's movement is often just the natural variation in AI answers. Wait for a week of daily scans before treating a trend as real. See scan-over-scan deltas.

Your baseline

Your first scan is your baseline. When you later publish content from opportunities, this is what improvement gets measured against.