Identity Monitor: detecting tone-of-voice drift

Last updated March 19, 2026

Answer

The Identity Monitor analyses your recent audio and video content and compares your communication style with your stored identity profile. If your style shifts, your profile is automatically sharpened so that future content continues to match who you are now.

How it works

The monitor compares your recent transcriptions with your tone-of-voice profile across 4 dimensions:

  • Tone Consistency: Does your word choice and tone match your profile?
  • Values Alignment: Do your messages align with your core values?
  • Audience Fit: Are you still addressing the same target audience?
  • Brand Clarity: Is your positioning consistent and recognisable?

Each dimension receives a score from 0 to 100. When the average score falls below 95, your profile is sharpened.

Manual analysis

  1. Go to Identity > Identity Monitor
  2. Click Refresh (top right)
  3. The analysis takes 3 to 5 minutes (runs in the background in multiple steps)
  4. You will see progress: "Analysing chunk 2/10...", "Comparing with profile..."
  5. Once complete, your scores are automatically updated

Automatic analysis

The monitor also runs automatically when you have generated enough new content:

  • After at least 10 content generations since the last check
  • And at least 7 days since the previous analysis

You do not need to do anything. After an automatic analysis, your scores are updated and your profile is sharpened.

Profile sharpened

If your score is below 95, you will see a message: "We have sharpened your profile based on your latest content." Your profile is always updated to match your current style. The profile follows you, not the other way around.

Tips

  • Upload new audio or video regularly to keep the monitor up to date
  • The analysis only works with transcriptions, so make sure your content contains audio or video
  • A shift in score is not necessarily bad: it may mean your style is consciously evolving
Was this article helpful?
identitymonitordrifttone-of-voiceprofileanalysis

Related articles