The Secret Sauce Baked into Channels
Others before us have enabled consumers to pay producers for their content. Largely, these attempts have failed. Consumers have been trained to expect to pay nothing for internet content.
We’re doing something no one has ever done before. We’re going to enable advertisers to make micropayments to consumers for their attention. Until now, advertisers have paid ad networks (like Google and Facebook) to find places to display their content.
Our view is that if advertisers pay consumers, this will remove the reservation consumers have in paying publishers for their content because they don’t have to get out their wallets to fund those payments.
Here’s the secret sauce:
Imagine that advertisers are bidding for consumer attention. This is bringing in revenue to those consumers. Meanwhile, consumers are buying content at prices set by the publishers. What’s going to happen?
Publishers might charge too much for their content. Consumers won’t be making enough from advertisers and so will run out of money and most will eventually leave in frustration.
Alternatively, advertisers might pay a lot for consumer attention and we’ll start seeing some consumers who try to game the system and just make money by watching ads. That will poison the well for advertisers.
But suppose we could put these things in perfect balance? As advertisers spend more money, the prices on content goes up just the right amount so that consumers neither go broke, nor have the ability to game the system. That’s what this differential equation is all about.
Imagine that the network has a big knob on it that you can turn that will move all of the content prices up or down in proportion. If publishers are charging a little too much overall, then we turn the knob a little to the left and everyone’s prices drop a little. If advertisers are spending a lot, then we can twist the knob a little to the right and publisher prices go up, meaning that consumers will be transferring more of their earnings to publishers whose content they want.
In the equation above, B is the “base price”. This is the unit price of each piece of content in the network. Publishers are free to set their own multiplier on this base price — from 1 to 10. Those who think their content is really valuable set a higher multiplier than others who think that a lower price will result in overall higher revenues for them.
On the right side of the equation, start with the last term: the sum of all promotional cards presented to users divided by the sum of all content cards presented, over the same recent time period. This is a measure across the whole network of the fraction of consumer attention that is spent on advertising. For example, if we’ve displayed a million cards to users over the past day, of which 110,000 were advertisements, then this ratio is 11%. Now subtract this ratio from the target ratio, T, that we set for the network. This is a constant that we determine based on our experience of how much advertising users are willing to tolerate. Let’s say that that number is 10%. Then we multiply that number (1%) by a damping constant, C. Let’s say C is 0.001. So now we can see that dB/dt is 0.00001. That means that B should be increasing by a very small amount. Similarly, if we were seeing a small amount of advertising appearing, then the ratio would be less than the target, and the result would be negative, pushing the base price down slightly.
This is what is called a “control system”. The base price will float up or down slowly in response to the amount of advertising appearing in the network. When there is too much advertising, the prices of content go up. And when there is too little, the prices go down. The damping constant makes sure that prices don’t change too quickly.
The reasoning behind this is simple. Advertisers are competing in an auction for a limited supply of consumer attention. High advertising demand will drive higher prices and consumers will start earning more money. To keep things in balance, they can afford to pay more for the content they are consuming.
Finally, to make the whole thing work, we also need to determine advertising rules not only at the network level but for each individual consumer. When should we introduce an advertisement into a user’s feed? The answer is simple: if and only if they need it. If a user’s balance is above a threshold, then we don’t need to show them any ads. But as a user starts purchasing content, their balance will fall, and that’s the time when we will start showing them ads as a way to offset their purchases and restore their account balance. Whenever there is an available ad slot, the network runs an auction to sell that advertising space to the highest bidder — taking into account the preferences and restrictions imposed by that user.
Consumers are free to eliminate ads by purchasing coins to fund their balance so that it never falls below that target threshold. But we expect most users will simply tolerate a reasonable level of advertising in their feeds in order to pay for the content they want. Users who consume more content will be presented with more ads to pay for it. Those who consume less will see fewer ads.
Simple. If it works, this will create an attractive place for advertisers to find consumers, consumers to find good content, and for publishers to get paid well for their work.
