Before this video loaded, you probably watched an ad, and or one will appear right about now.

How did this exact ad get on this video? And what you really want to know, how much money do these things make?

Okay, there are three players in this game. Creators who create videos, advertisers who make the ads, and YouTube who plays Matchmaker Matchmaker Making the Match.

Now, unlike old dying media where a dude in a room could place ads on predictable,

known content at a leisurely schedule, YouTube doesn't work like that.

Can't work like that, because on YouTube, there is no schedule.

At any moment, a wild viral video can appear. On a normal day, 65 years worth of videos are uploaded to YouTube

and people watch a billion hours of video. It's an endless waterfall of random content

that just keeps growing. Hiring enough humans to categorize all those videos and match ads

to them would be impossible. This is a job for bots. The moment each new video is uploaded, YouTube's bots get to work,

looking at the title, the keywords, the captions, the comments, the controversy, all to make their best guess as to what category it belongs in.

Meanwhile, advertisers tell their own bots what category of videos they want their ads to run against and what kind to avoid.

And here is where YouTube earns their money.

Between when you click on a video and when it plays, a YouTube bot holds an auction,

announcing the categories its guest the video is in, opening the floor to all interested

advertiser bots to place their bids.

The winning advertiser bot is the one with the ad that will most likely make YouTube and the creator the most money at that instant. But that's not necessarily the highest bidder. Different ads pay in different ways. For example, skippable video ads, the advertiser doesn't pay if the viewer skips as soon as they can. The advertiser only pays if their ad is clicked or watched.

A super high bid means nothing if the ad is super boring. No clicks equals no dollars for YouTube or the creator.

So there's a bias against Boring. The auctioneer bot will prefer a lower bid for an ad with a higher click rate than a higher bid for an ad with a lower click rate.

But boring is in the eye of the beholder. So in addition to YouTube's bots trying to guess what the video is about,

there are bots that try to guess what the viewer is about, looking at their watch history and device and activity to try and guess their age,

chromosomes, income, location, and interests.

...