Analytics
Drop-off Analysis
Drop-off analysis shows you where in the video viewers stop watching. This is one of the most actionable pieces of data Keep'em provides.
How It Works
The system divides your video into segments (typically 10% intervals) and counts how many viewers were still watching at each point. The result is a retention curve — a line that starts at 100% and decreases as viewers leave.
Reading the Data
A gradual, steady decline is normal. People naturally drop off over time. What you're looking for are sharp drops — points where you lose significantly more viewers than the surrounding segments.
A sharp drop at minute 6 of a 30-minute video means something at minute 6 is pushing people away. It might be:
- A topic that isn't relevant to most of your audience
- A section that's too long or repetitive
- A confusing explanation that needs reworking
- Poor audio/video quality in that section
Taking Action
When you identify a drop-off point:
- Watch that section of your own video critically
- Check the Top Questions to see if viewers were asking about that topic
- Either improve the content in a new recording, or add better documentation to the knowledge base so the AI can support viewers through that section