The problem
Your best earner stops getting views. You open it and nothing is wrong. The hook is the hook it always was. Payoff is where it was. Click and Convert have not moved. The video did not get worse, and the distribution went away anyway.
Lesson 23 asks whether a slowdown is the script or the market. There is a third answer that looks like neither, and most people never consider it: the video ran out of people it can profitably reach.
Take a comment reply that has earned beautifully all month. Its lifetime figure is excellent, and that figure is an average across everything it ever did. Split it by when the views arrived and it looks like this instead.
| As the views arrived | What it earned per thousand |
|---|---|
| The first stretch | Outstanding |
| The next stretch | Still strong |
| By the time it had four times the views | Ordinary |
| Beyond that | Barely worth the impression |
The lifetime average still reads as a winner, because the early stretch is carrying it. But those impressions are already spent. Nobody is deciding whether to buy them again. The only question in front of the system is what the next block of views is likely to be worth, and on this video that answer has been falling for a while.
So distribution is not a reward for revenue already banked. It is a prediction about the value of the next impression. Once you see that, two videos that look contradictory stop being contradictory: a comment reply can earn far more per view than a broad video and still be handed a fraction of the traffic, because the pool of people it works on is small and it has nearly finished it.
The example
Which means a video has two separate capabilities, and the curriculum has only been teaching one of them. The first is how well it turns attention into money, which is Lesson 7. The second is how much traffic it can absorb before that stops being true. Cross them and you get four kinds of video, each needing a different thing from you.
| Small audience it works on | Large audience it works on | |
|---|---|---|
| Earns well per view | Demand capture. Comment replies, objection videos, offer and closer videos. Genuinely valuable, and it will finish its pool fast. Expect the ceiling, don't read it as failure. | The real winner. Keeps finding new people at the same efficiency. Rare, and the thing to protect and copy carefully. |
| Earns little per view | Weak. It neither reaches many people nor pays for the ones it reaches. | Acquisition creative with a leak. The hook is doing its job. The money part is not, so fix the product transition, the proof or the offer, and leave the opening alone. |
That bottom-right cell is the one most often misread. A broad video with weak earnings gets a new hook, which is the one part that was already working. Lesson 16 is the habit that prevents it: the fix lives one step before the leak, not at the top of the video.
The pool is set by what the video references
Here is where it gets useful, and it is worth being precise about the mechanism, because the obvious explanation is the wrong one. Paid distribution is not serving a creator’s followers. The system is looking for buyers, so a creator’s existing following is not what decides where the video lands. What the video is about is.
Two videos, same supplement, same greenscreen, near enough the same script. One opens on a figure from the health-policy argument that is in the news that week. The other opens on a chart-topping musician. In a content review they are the same video: same format, same product, same structure. They are not the same asset, because that opening reference places each one in a different corner of the platform, and the people reachable there are different people. Both may well buy the supplement. They are not being found the same way.
So the unit worth thinking about is never “greenscreen.” It is closer to the reference, times the angle, times how much the viewer already knows, times the product, with the creator carrying delivery and credibility on top. Change what the video invokes and you have changed which room it is shown in, whatever the format says.
Which is why the denser converter is not automatically the better asset.
| Earns per thousand views | People it can reach | What it adds up to | |
|---|---|---|---|
| The health-policy reference | Much higher | A shallow pool, finished quickly | A strong, early, limited return |
| The music reference | Noticeably lower | A far deeper pool | More total revenue, over longer |
Neither result makes the other unnecessary. They are working different parts of the market, and the account wants both. The same product can have several winners at once, each one opening a different room.
The third question: are these new buyers?
There is a term the first two miss. Suppose two creators earn about the same per view. If they are reaching more or less the same people, adding the second one does not double anything, it just splits a pool that was already being served. If their audiences barely overlap, the second one opens a genuinely new branch of demand for the product.
So the full question a strategist is asking about a piece of creative has three parts: how well does it convert the attention it gets, how many people can it reach before that stops being true, and does it reach anyone the account was not already reaching. Creative variety is worth money for that third reason, not because fresh content is nice to have. TikTok says as much in its own guidance: a varied pool of creative gives the system more places to go looking.
This is the trap. The first video earns the most per view, so twenty creators are told to remake it, reference and all. Now there are twenty assets competing over one pool, and the second video, the apparently weaker one that was opening a different room, has been retired. The account looks more efficient for a week and has quietly narrowed itself to a single corner of the platform. The aim is not to have everyone produce the highest-earning angle. It is to keep several references working at once, then find out how far each one goes.
What travels to another creator, and what doesn't
This is also a sharper answer to why copying a winner keeps failing. Lesson 27 says a video has layers and most people copy only the words. Split those layers again, by whether they survive the move to a different creator.
| Travels | Stays behind |
|---|---|
| The hook's structure, the mechanism, the proof, the offer, the call to action, the pacing | The cultural reference the video opens on, the corner of the platform that reference reaches, how much that audience already knows, and the creator's own credibility on the subject |
Copy the left column onto a new video and you are carrying the part that works. Carry the right column across unexamined and you have preserved the visible structure of the winner while assuming its reference will land the same way somewhere else, which is the invisible half that made it win.
Sometimes the bar moved, not the video
There is a second way a video loses distribution without deteriorating. The brand raises the return it is asking for. Every impression now has to clear a higher bar, and impressions that were acceptable last week no longer are. Nothing about the creative changed. The threshold underneath it did.
Read that together with the pools and it matters more. The audience did not go anywhere. The brand changed the price it is willing to pay to reach it, and a pool that cleared the old bar no longer clears the new one. Conclude from that "greenscreens stopped working" and you will retire creative that would pay again the moment the target loosens. Lesson 44 covers the budget and the target as the two gates in front of every impression.
This is the same point as the pools above, from the other side. In older ad accounts somebody built a prospecting audience and a retargeting audience by hand, and decided which creative went into each. Here you do not choose the audience, and you are not inheriting one from a creator’s followers either. You choose the message, and the message decides who can plausibly answer it. “Three reasons this matters if you are over forty” can only land on one kind of viewer. “The coupon disappears tonight” can only land on another. So you are not writing videos and handing them to a targeting system. The videos are the targeting, and the system works out where each one can earn.
What the account actually needs
So the right question is never which kind of video wins here. Keep only the dense converters and you get very good at harvesting a pool that keeps shrinking, which reads as strong efficiency and falling volume. Keep only the broad ones and you generate a great deal of attention you never turn into orders. The mix is the thing you manage, and the job is to find several profitable rooms rather than one best video. Lessons 10 and 11 cover how differently the two kinds should be judged.
Do this in Verdict
- 1When a winner slows, open Fatigue first. It separates a pattern wearing out from a bad week, and Lesson 23 walks the read. A video that has simply finished its audience looks different from one the market moved past.
- 2Read the video’s recent window against its lifetime, not just its lifetime. The Leak Finder reads a window, so a winner whose recent stretch has gone soft shows up as soft rather than hiding behind its own average.
- 3Before rewriting a broad video that earns little, check which half is failing. Strong attention with weak money is a body-and-offer job, not a hook job. Open it in Video Breakdown and change the part that is leaking.
- 4When a proven video loses distribution and its numbers are unchanged, look at the room before the creative. The Pre-Slot Screener and its Window Score read spend and payout direction from the outside, which is the closest you get to seeing the bar move.