

AI Metrics
19 snips Jul 17, 2025
This discussion dives into the transformative world of AI metrics and their growing importance in the software industry. Key topics include Daily Active Users versus Monthly Active Users and the complexity of tokenization. The hosts reminisce about early internet metrics like hits and page views, illustrating how measurement has evolved. They also introduce Occam's Razor, highlighting the value of simplicity in metric selection. Ultimately, the episode balances humor and insightful analysis, making sense of the dynamic landscape of AI.
AI Snips
Chapters
Transcript
Episode notes
Understanding AI Metrics Challenge
- The confusion about AI metrics stems from not fully understanding AI itself yet.
- This is why defining what to measure is still a challenge.
DAU/MAU and Stickiness Anecdote
- Dave Kellogg recalls DAU/MAU measuring usage frequency or engagement, not stickiness.
- He exemplifies with Google overtaking Yahoo due to zero switching costs, despite high daily use of Yahoo before.
ChatGPT DAU/MAU Performance
- ChatGPT's DAU/MAU ratio at 66% indicates exceptional daily user frequency.
- Different app types have varying good DAU/MAU benchmarks, e.g., travel apps score lower due to usage nature.