Is YouTube a market? And does it matter?

Manoel Horta Ribeiro (@manoelribeiro),


Importantly –atleast for the context of this blog post– it also makes me wonder about an even broader idea. Namely:

Is YouTube a market where content creators compete for users?

And here some clarification is due. I am no economist, but it seems clear that we can find, in the YouTube eco-system, buyers, sellers and a scarce resource:

  • Sellers are YouTubers, many of whom make a living out of profits that stem from their video productions.
  • Buyers are users, who watch content on YouTube, indirectly (and sometimes directly) sponsoring content creators through ads.
  • The scarce resource is user’s time, after all, there is only so many hours in a day.

Interestingly, YouTube is one of the few Social Networks where the distinction between Buyers and Sellers is so drastic (take for example Twitter where this is not so clear). Moreover, it is perhaps the Social Networks where profits are more directly linked to content creation, as they have a partnership program with YouTubers where Google shares ad revenues.

Given that there is a limited (although ever-growing) amount of users on the platform, each of them with limited time, competition follows. Yet, I get this strange feeling about “defining” something as a market. Two questions follow: (a) is that enough? and; (b) so what?

With the fear of being pedantic, let us go through them. The first question concerns the falsifiability of the hypothesis. It could be that, although the described entities do exist, the interaction between buyers and sellers is detached from the scarce resource. This seems counter-intuitive to me, but begs the question: which kinds of empirical tests can we make to validate that YouTube is a market? One thing that occurs to me is that we could try to replicate things expected from markets, such as: (i) can supply (content) adapt to the demand (views)? and (ii) do the agents in the market behave to maximize profit? I believe these questions could be answered in a data-driven manner with YouTube data (which I may have been collecting for the past 3 months, who knows :p).

The second question is perhaps deeper, and is the one that bothers me the most. Sometimes, this whole market thing seems to be a center-piece of how to make online social networks better, but during moments of doubt it seems remarkably useless. Let me play devil’s advocate here: on one hand, the YouTube content creation environment, although remarkable, is a mess: (i) content creators frequently become frustrated with how, out of the blue, the rules of the game changes for them (for example, when a whole category of video stops being profitable); (ii) a lot of crazy things happen on YouTube because of exploits of the algorithm (I’m looking at you ElsaGate); (iii) lastly, in general, I wonder to which extent the kind of process pointed out by Munger and Phillips can be attributed to a poorly designed or regulated “market”. All in all, studying YouTube as a market could allow the problems of the platform to be solved less arbitrarily (that is relying mostly on an army of moderators and on reactively banning/demonetizing content).

On the other hand, it is not clear to me how modelling YouTube as a market would help in practice. Even if YouTube is a market, it is a different kind of market, ruled by algorithms, AB tests and living under another ad market. In that sense, it could be that the levers to “regulate” this market are not yet developed. Or even worse, it could be that these market “levers” are very hard to define in a platform where the rules of the game change so often.

Written on January 5, 2020