Summary

From YouTube's point of view, a YouTube video is just a set of pixels and audio that causes you to look at a screen and watch advertisements. If each video causes you to watch an advertisement and watch another video, then YouTube can chain together a long session of video->ad->video->ad->... and get multiple ad sales out of one sitting. Any video that interrupts this chain (like "how to quit YouTube" that actually makes people quit) ends the chain and YouTube misses out on the revenue from every subsequent video+ad that they could have shown. Because of this, YouTube works extremely hard to never show you videos that will cause you to break the chain and leave your screen. Therefore, you will probably never find on YouTube a video that causes you to use less YouTube.

Style/tone/content notes: This is an oversimplification and doesn’t describe all of the factors influencing a platform like YouTube. This is also a pretty typical take on engagement-based platforms and the drawbacks of optimizing for duration of user interactions. It also has a fairly negative bias against the platform, and doesn’t do a great job capturing benefits or nuances of the platform. I’m trying to convey an idea that I hope has a lot of value for some people struggling on the platform, but I’m also not happy with the tone/clarity/bias in the writing. Please let me know how I can improve.

What is the actual end goal of YouTube?

Instead of looking at what happens when a user types www.youtube.com into their web browser and working forward from there, let’s instead work backwards from the end goal of YouTube as a platform and see what steps might lead to this outcome. Thus our first question is: what is the end goal of YouTube?

To make progress towards understanding YouTube's goal, we can look at examples of things YouTube wants to happen:

• a company pays YouTube to advertise to viewers to buy a certain soda. When the viewer walks into a pizza shop later that week, they buy the soda, and the soda company makes more money on the sale than they spent having YouTube suggest it
• a political campaign pays YouTube to advertise to viewers to vote for politician #7. When a viewer is in a voting booth later that month, they recognize politician #7 name, but they don’t know any of the other names. They vote #7
• an airline pays YouTube to advertise to viewers to buy flights with Generic Airlines. When the viewer is looking for flights to visit family for the holidays, almost all the flights are the same price, the same duration, and the same mediocre service. The viewer feels slightly more positive about Generic Airlines because they had seen an advertisement with a smiling flight attendant on a tropical beach. The viewer buys the flight from Generic Airlines instead of the other competing airlines

What do all these examples suggest?

In all these examples, YouTube wants companies to pay money in exchange for showing a user an advertisement. As a secondary goal, the advertisement should create more value for the company than it costs, so that the company keeps buying advertisements in the future. Notice that YouTube could apply more than one ad to a user. YouTube could show the same user all 3 of the example scenarios, and get paid for all 3 ads! So YouTube wants to get paid to show advertisements, and YouTube wants to do that as many times as possible.

What is the point of the YouTube videos themselves?

Notice that our proposed goal of YouTube is to have companies pay YouTube to show users advertisements, and to do this as many times as possible. Notably in this goal statement, YouTube doesn’t actually care about the content of a YouTube video except that it causes a company to pay to place an advertisement and causes a user to view the ad and the subsequent videos.

Don’t think about a knitting tutorial on YouTube as a knitting video. Think of it as a 20 second knitting needle company advertisement, surrounded by some pixels and audio that caused a viewer to hold up their smartphone and look at the screen. Don’t think of a luxury car comparison video as a exploration of different cars, think of it as a car advertisement surrounded by pixels and audio that caused a potential car buyer to look at their computer screen. Don’t think of an election coverage video as an analysis of the political landscape, think of it as a 10 second political campaign ad surrounded by pixels and audio that caused a potential voter to sit in front of their smart TV and look at it.

To YouTube, a video is just an arrangement of pixels and sounds that causes a user to pay attention to a screen. Then, while the user is paying attention to the screen, YouTube can pause the pixels and charge a company to play an advertisement. Then, ideally the user begins the next video, repeating the cycle as many times as possible.

A critical aspect of a video is that it doesn’t break an advertisement viewing chain

Another critical feature of a YouTube video is to cause viewers to stay on the platform and continue to look at ads. Since a viewer can be shown more than one video/ad in a sitting, we can imagine a user’s journey on YouTube as follows:

Ad 0 -> Pixels 1 -> Ad 1 -> Pixels 2 -> Ad 2 -> Pixels 3 -> Ad 3 -> Pixels 4 -> Ad 4 -> Pixels 5 -> Ad 5 -> Pixels 6 -> Ad 6 -> … -> Pixels 37 -> Ad 37 -> user collapses in exhaustion at 3am, falling asleep with their phone in their hand

In this scenario, the chain of pixels showed to the user caused 38 advertisers to pay for 38 ads. Supposing each ad sold for 1c, this means the value of that chain was 38c, which is good for YouTube.

Now, let’s suppose that video Pixels 3 was titled “How to get your life together and go outside for once” and suppose the video actually worked. Everyone who watches Pixels 3 leaves their device and actually goes outside. The chain as viewed from YouTube is:

Ad 0 -> Pixels 1 -> Ad 1 -> Pixels 2 -> Ad 2 -> Pixels 3 -> user goes outside, notably not looking at ads anymore.

In this scenario, the chain of pixels showed to the user caused 3 advertisers to pay for 3 ads. This is awful for YouTube in comparison to the first chain. The including of Pixels 3 in this chain caused the total value to go from 38c down to 3c. Pixels 3 literally destroyed 35c of the possible 38c chain value to YouTube.

This is the core of why any searchable or reccomended YouTube video is unlikely to cause any behavior other than watching ads and watching the next video. Any video that interrupts the chain of showing you ads is bad for business. Any video that causes you to do anything other than sit in front of a screen is bad for business. If a video does anything to modify the user at all away from viewing more ads, the video is bad for business.

This is why you will almost never see in search results or get recommended a YouTube video that makes you do anything other than sit in front of a screen. YouTube would be shooting itself in the foot to show you such pixels and audio. Showing you videos that make you leave the screen interferes with the primary goal of showing you another advertisement.

What about productivity YouTubers, organization YouTubers, self-help YouTubers, etc?

We might feel like productivity videos, orginizational videos, get-your-life-together videos, or other positively-themed videos might cause the viewer to consume less YouTube videos by giving the viewers knowledge, skills, or motivation to do something else. The flawed assumption in that idea is the assumption that the video's title and the video's effect on the user are the same thing. In reality, nothing forces a video titled "how to get your life together" to actually cause a viewer to get their life together. If a user types “how to actually get organized” into search, YouTube could either:

Show Pixels 7649458, “How to get organized by cleaning your desk” which is a video that actually causes people to stand up and organize their desk, leading to the chain:

Ad 0 -> Pixels 7649458 -> user stops watching and starts organizing 

Or instead, YouTube could show Pixels 294860, “How to get organized and stop wasting your life” which is a video that shows a nice looking cleaning montage, shows before and after shots, describes deep-sounding but meaningless sayings like “a clean desk is a clean mind.” The video never has a natural break for the user to leave, and the video ends with, “but a clean desk is only the first step to an organized life! Watch Pixels 3946684 which will explain how to get your life together by cleaning your closet!” The user saw what looked like a story of someone getting organized, and that organized person said the next video was important, so the user watches the next video. The chain for this might look like:

Ad 0 -> Pixels 294860 -> Ad 1 -> Pixels 3946684 -> Ad 2 -> Pixels 636457465 -> Ad 3 -> Pixels 109494 -> … -> Ad 49 -> user has not moved from their seat in 4 hours and collapses in exhaustion having consumed 50 ads.

Every single video in this chain can claim to get your life together. Every video can promise to make you use YouTube less. Every video can give truthful, deep, insightful information about getting your life together. The key insight is the true purpose of each video will be to cause you to watch the ad and watch the next video. No video is likely to cause any action other than watching the next video, because any other action breaks the chain and destroys value. These incentives still hold for videos titled "how to watch YouTube less" or "how to get your life together".

Making this theory pay rent

This is an interesting story, but what would we expect to observe if this model of YouTube is true? Let's make this idea pay rent in both expected and forbidden experiences like in Making Beliefs Pay Rent (in Anticipated Experiences):

Things this model predicts:

• Suggested videos and search results will favor monetized videos (show ads), videos without ads will rarely be suggested or show up in search
• Suggested videos and search results might not be related to what you just watched or what you searched, but rather are chosen by if they will cause you to watch ads and watch another video (irrelevant but engaging videos in watch next and search results)
• Platform designed to engourage next video watching (autoplay, suggested next videos, reccomended, calls to action to watch next videos ("if you liked this video, I think you will love my next video about..."))
• Videos promoted by the platform will rarely contain calls to action involving leaving the platform or doing anything other than watching the next video+ad
• Search results for questions that could be answered in one short video will be long analysis videos with ads, and will end without a satisfying answer, prompting the user to contine watching additional explanations or unrelated content
• Suggested videos and search results will rarely leave viewers with a sense of fullness, completion, satisfaction, or enough-ness. Videos will tend to end with a question or non-answer, leaving the viewer with a longing for the next video to provide relief (it never will, it will end in a way that encourages another video)

This model somewhat forbids/predicts-won't-happen:

• Popular, promoted, searchable videos that are demonitized/ad-free
• Popular, promoted, searchable videos that have a working call to action causing viewers  to leave the platform (example of something unlikely under the selling ads chain theory: "if you liked this video, promise me you will power off your computer and phone immediately and reflect on your relationship with technology for the next 48hrs")
• Platform changes making watching the next video harder or less automatic, like removal of next video autoplay, or removal of end-of-video suggestions
• Popular, promoted, searchable videos that leave viewers satisfied and complete, ready to put down their device and move on with life
• Search result videos that clearly and concisley answer the question without leading into further exploration/questions/viewing

Conclusion

A YouTube video will almost certainly not cause you to get your life together, get organized, use your devices less, go outside, spend more time with friends, or any other non-sitting-in-front-of-a-screen activity. A YouTube video is just pixels and audio that causes you to watch ads. If a certain arrangement of pixels caused you to do any action other than continue watching pixels and ads, it would break a chain of viewing and destroy value for YouTube. This is true even of videos titled “How to get your life together” or “How to use YouTube less.” If you pay attention, these videos almost never have immediately actionable call-to-actions, and they almost always chain nicely into a subsequent video. Videos with such titles are pixels the same as any other, with the same incentives as any other, chosen carefully to make you watch ads and keep the chain of videos+ads long. 

Pixels that change your life may/do exist, it is simply not likely that YouTube will show those pixels to you. Any video that actually causes users to do any action other than continue watching will be measured to be a chain breaker and value destoryer, and YouTube will avoid suggesting it or putting it in search results. YouTube may have millions of videos that would cause viewers to use less YouTube, but YouTube is incentivized never to show any of them.

If you want to reduce the amount of time you spend in front of a screen, and you search for how on YouTube, know that everything presented to you has been experimentally verified to not cause the exact results you seek.

If you want to do anything other than watch a long chain of pixels and ads, do not type www.youtube.com into your browser.

Details, nuances, “it’s complicated”

1. YouTube is not a person or a single-goal entity. In this writing YouTube was somewhat personified and assigned goals, and that is not literally how the platform works. Any platform that big is a disjointed, many-team-many-person organization with different goals all throughout. Maybe a certain manager wants a certain metric to improve in order for that specific manager’s promotion cycle. Maybe a passionate content moderation engineer really does care about the content of some of the pixels. Please use this model as a quick, visceral tool to understand what is going to happen if you visit the site, but please understand in reality the platform is not a single entity and doesn’t have goals this simply.

2. The algorithm might not be as simple as “maximize showing ads and chain length.” For example, advertisers track how much an advertising campaign causes sales to change, and adjust their willingness to pay accordingly. This leads to features like YouTube's "are you still watching?" check, which makes sure that showing pixels+ads has a chance of influencing a user's buying behavior (Advertisers don't wan't to pay for ads played on an unwatched laptop screen). Or maybe certain ads might be so valuable that the platform might want you to leave for the ad, since the value of that exceeds the value of the rest of the chain you would have watched. Or maybe a platform is competing with another platform in a way that pressures them in different directions. Or maybe a platform has legal issues or payment processor issues or corporate image issues that make them care about Pixels 7 in a context other than pure showing-ads-long-chain. Or maybe a new platform is focusing on user growth and isn’t maximizing ads yet. Or maybe a platform has more than one monetization method and sometimes tries to funnel users between these methods. Or maybe the CEO has a moving conversation with someone at a party and pushes the platform in a slightly different direction for non-ads reasons. This pixels, ads, chains model is oversimplified and doesn’t explain everything. Please don’t over-apply this simplification.

3. The same concept applies to many ad-supported digital platforms. You probably won't watch a tiktok that makes you quit tiktok either.

4. I realize both the tone and content in this are emotional and biased, and that is making some (all?) of the points less factual and more exaggerated. I realize the things I say are certainly biased by feeling used and deceived by ad-based platforms. I don’t like putting stuff like that out into the world, but at the same time I wish I had read something like this a few years ago, and I hope that this writeup can help some readers spend potentially hundreds of hours per year in ways more aligned with their goals. If anyone has writing advice, suggested revisions, tone suggestions, or anything else, please let me know.

5. I don't fully know how adblockers play into this theory. My current initial guess is that a platform values network effects, market dominance, and potential future monitization enough that they will continue serving users with adblockers while trying to convert them to ad watchers. Or maybe at some point in the future, platforms will start turning away users that never watch ads? (trials of messages like: Adblockers are against the YouTube Terms of Service, disable your adblocker to continue watching)?