The Mid-Video Sag: What a Steady Retention Slope Means

Nobody leaves in the middle of a great middle. A mid-video sag, where the curve slides gently across minutes three through eight before recovering at the payoff, means your viewer has gone on autopilot or your pace has. The trough between the hook and the payoff is the longest psychological distance in the video, and most sag is built in the script, not the edit.

The harsh version: a sagging middle is usually a writing sin that editing cannot rescue. Compression can smooth a bump; it cannot manufacture stakes.

Find the first minute of sag

Look for the minute where a slow slide begins. The moment the line angles down and stays down until near the end, you have found the start of the sag. Open the script at that point and read two sections behind it. The boring almost always announces itself a minute early.

Twice I have traced a sag to two adjacent sections doing the same job. Twice more it was one segment repeating its own idea three times with three different examples. Both times the graph looked identical, which is why you read the curve but then read the words.

Why middles sag

• An unbroken run of explanation. Segments of pure 'this is how it works' with no application wear out by minute four.
• Stakes that stall. The middle restates the problem instead of raising it, and the viewer feels the engine idle.
• Research that went in because it was interesting to you. The audience can always smell the difference.
• Examples that could have waited until the payoff, stealing energy from the ending and giving the middle nothing.

The middle of a video is where creators store the information they think they need to share and the audience stops needing to listen. Diagnosis is separating the sharing from the listening.

The recovery tells you the fix

If the curve recovers sharply at the payoff, your structure is holding and the middle just sprawled. Cut the sprawl. If the curve never recovers and keeps sliding past the payoff, you have a deeper problem: the second half failed to promise anything, and the fix lives in the script's section plan, not in a trim.

Either way, measure the sag in points per minute. Take retention at the start of the sagging section and at its end, divide by the minutes between them, and you get an honest slope number. Run the same number after the edit and you will know whether the fix worked before the algorithm tells you.

Key takeaways

• Pin the minute the slide starts and read the script two sections behind it. That is where the sag was born.
• Raise one stake at the midpoint, not five. A single raised stake carries a middle further than a pile of facts.
• Move half of each section's examples to the payoff so the middle stops doing the ending's job.
• Measure slope in points per minute before and after the edit. The number gives you feedback faster than the graph.