Retention is decided in the script, not the edit

If you hire an editor and just tell them “do retention editing,” you're probably hurting your videos. I would say retention editing is the most overvalued thing on YouTube right now.

Most people think retention editing means more zooms, more sound effects, B-roll on every sentence, a cut every two seconds. But retention has two levels.

Level one: narrative

Why do I want the next piece of information? Questions, open loops, progression, consequences. This is what keeps someone watching, and it's decided in the script, before the editor touches anything.

Level two: visual

How clearly you deliver that information. B-roll, graphics, cuts, music. This is what most people call retention editing.

Most editors go straight to level two. But if level one is broken, level two can't fix it. It just hides the problem for a few seconds.

When retention drops, ask a different question

So when your retention drops, the question is not “do we need more edits?” The question is: where did the video stop delivering what the title promised? Find that point first. Then edit.

I'm not saying fast editing is bad. Some niches need it. I'm saying fast editing doesn't create retention by itself. And as more people watch on TV, a cut every second matters even less. I wrote about that in YouTube is becoming television.

So what does a good script look like?

I used to check a script with three questions. Is the language simple? Does it give the audience exactly what they want? Would I watch this?

Those are still good questions. But there is so much AI content now, and so much generic advice, that passing those three is not enough anymore. So now the real question is:

Would the viewer pay at least $100 for this level of education?

If the answer is no, I redo it. Maybe the idea isn't strong enough. Maybe I'm saying what everybody else says. Maybe there's no real takeaway.

When you write with that filter, the content ends up more valuable than most of what's out there. Not because you're better. Because most people never ask.

So I hope this one was worth $100. If not, sorry.