Transcription
When I was hiring writers for my old scriptwriting agency, I gave them this retention graph as a test. Because while you can definitely become too obsessed with retention, not understanding how to diagnose your graphs at all will leave you making some of the most basic mistakes repeatedly, which is just a massive wasted opportunity to grow a loyal audience faster.
So, in this video, I'm going to show you why most YouTubers misdiagnose their graphs, where to look if you're struggling to understand a big dip, and how you can fix this problem at the scriptwriting stage before you even sit down to film.
Now, to diagnose the problem, we need to clear up a common misconception that's stopping you from fixing your next script. Now, take a look at this graph again. This is a modeled version of a real graph from a multi-million subscriber channel that I worked with. And my task when I was in this team was to diagnose why there was this drop-off.
Now, my question to you, if you were looking at this in your own YouTube studio for the first time, where would you think that the problem was? I mean, it's going to be this bit, right? It's where the video has completely flattened out and then all of a sudden dips down quite substantially before flattening out again.
Now, the way most YouTubers would try to diagnose this would be to watch this exact segment of the video maybe 5,000 times and just keep staring at it and keep rewatching it until your eyes start to fall out of your head, trying to figure out what you were doing wrong in that moment. But in the case of this video, the problem the team and I were having diagnosing it was this bit, the bit with the huge dip, seemed like one of the best bits of the entire video. The storytelling was really good. There was a satisfying payoff at the end, and there was no obvious reason why this part in particular was losing so many viewers.
Now, sure, we have to be humble when reviewing our own content and accept that retention dips are often caused by the video being less engaging in the moment with the dip. The reality is, YouTubers frequently misdiagnose their graphs because they don't realize a big dip like this can be the symptom of something you did elsewhere. But where exactly?
If the problem isn't at the dip, where else are we supposed to look? Well, good news is there are usually three places, and we're going to start here with the most obvious. Because your hook actually has a much bigger responsibility than simply getting the audience interested enough to watch beyond 30 seconds. It also has to set an expectation for the video's format.
Now, one of the biggest causes of retention drop-off that I've seen consistently over the last 4 years is when a video does something like this. First, let's imagine a fitness video that starts out really inspirational, then halfway through switches to talking about the detailed science of a particular aspect of fitness. Or imagine a business and finance video that starts out with some mindset advice, but then switches to talking about a detailed stock market breakdown. Or even my latest video where I started with a high-level theoretical explanation of an AI technique, but then switched to showing a lot of quite bland-looking prompt documents and the retention fell off a cliff.
And the reason I mention all this is because we often set the expectation for a video's format during the hook. So, drastically switching the format away from the expectation that you set up in the hook can cause big problems for your retention, which is why I always go back to the hook to help me diagnose a trickier dip that seems to have no other explanation. But in the case of this video, the graph that I'm showing you here, we reviewed the hook and the expectations we set and decided that this drop-off did not really coincide with a format shift, so it had to be something else, which brings me to the next place you need to look if you're trying to diagnose an apparently inexplicable retention drop-off, and it requires the same sort of thinking as the first.
Now, if you're familiar with my approach to scriptwriting, you know I think of every segment of my video as broken up into setup, tension, and payoff. And to diagnose this retention problem, we need to focus on the setup that came immediately before the dip. So, to explain this, let's imagine a video about how to nail a job interview. And I want you to notice how it feels as I transition between two different topics here.
So, that's it. The night before your interview, find one recent news story about the company and work out how it connects to the role you're applying for. Mention it once, naturally, and you'll immediately stand out from 90% of the other candidates in the room. And when they ask you to tell them about yourself, you need to know how to handle this as well because this comes up in most interviews, and most people either go on too long or don't say enough. Boy, okay. That felt That felt boring, right? That felt long and rambling, and there was something just not quite right about that. And the reason for this is that at this moment here, we actually switched from the payoff of the previous segment to the setup of the next one, but it was really hard to tell.
Now, the result of transitioning from one topic to another like this is that your video can start to feel like an endless stream of and then, and then, and then, which will cause the viewer to lose where they are and ultimately lose patience.
Now, we're coming back to the graph in just a moment, but I want you to first look at what happens when I make a much clearer distinction between one segment ending and the next one beginning. So, that's it. The night before your interview, find one recent news story about the company and work out how it connects to the role you're applying for. Mention it once, naturally, and you'll immediately stand out from 90% of the other candidates in the room. But here's the problem. Most candidates waste all of that preparation the moment they open their mouth because there's one harmless-sounding question that comes up in almost every interview that quietly ruins most applicants' chances. When you make a conscious effort to set up a new segment very distinctly, so that it feels like a separate conversation from the previous one, it allows you to do so in a much more compelling way. And without a compelling setup, even the most interesting topic in the world is going to become confusing because your viewers will forget exactly why they're supposed to be watching.
But once again, we checked for this and we were pretty sure that wasn't the problem either, which brings me on to the third and final place to look, the place we ultimately concluded to have been the problem in the case of this video. Really, the biggest thing that I found about analyzing retention graphs is that your brain kind of needs to be in two places at once. Because not only should you be thinking critically about, okay, what am I saying or doing here in this moment? How does it come across? Could a viewer feasibly get lost in this part? But you also have to relate all of these thoughts back to the things you said and did earlier in the video.
Let me explain what I mean by that. So, during this segment, as the graph was dipping, we couldn't really see any obvious problems that would constitute such a major drop-off. But by zooming out and thinking, okay, what is the overall ask I'm making of my audience in this segment? I realized something that I now think about when diagnosing any retention graph. Because ultimately, every segment in your video is asking the audience for their trust, right? Trust that you're going to be respectful with their time and they're not going to regret giving you that time. And looking back at the segment before the drop-off, I reasoned honestly with myself. I looked at the promise we'd made in that segment and concluded that we hadn't paid it off in the right way.
Now, what that may look like in your niche will be different to mine, but imagine a video about how to light a cheap YouTube studio setup at home where we pay off a segment by saying something like this. So, it really depends on the natural light in your room, but a ring light or a key light are both popular options that a lot of people find works well for them. Versus if we'd said, "Seriously, don't bother with a ring light. It just creates this flat, unflattering glow that makes you look super amateur. What you really want is a single key light positioned just off to one side and 45° above eye level."
This will take a bit of self-reflection and a few reps, but if you take a few weeks away from any video and come back to look at it without ego, I guarantee you will be able to spot the times that your video under-delivered in certain key moments. And that brings us back to our graph here because if the payoff from the previous segment was not satisfying, goes without saying the viewer is making a choice as they move into the next segment of, "Do I want to give this creator my time for another 2 minutes? Because I just did that and I felt like they didn't deliver." The truth is, you'll never learn to diagnose a retention graph just by looking at the dips. You need to look at the entire graph in its full context.
Now, diagnosing a retention problem after the fact is one thing, but the creators who consistently get this right aren't spending hours reverse-engineering their next videos from their previous graphs. They've built a system that makes compelling setups and payoffs easy to write without even thinking about it. So, watch this video next where I'll show you the exact scriptwriting system I use with my clients and students and how it keeps these problems out of your script in the first place.