Content creation & media is an interesting business. It costs a lot upfront to produce but the marginal cost of showing it again is effectively zero. This essentially means the more times I can find ways to show the content in the future, the more $s I can make and thus, the more I can afford to spend upfront.
The no of times I could show any piece of content to the audience was determined by my available distribution. Typically, you would sell a movie to as many distributors as possible. Historically, the value chain in descending order looked something like this -
First-run/Premium theaters
Hotels & airlines
Pay-per-view
DVDs
Cable
Broadcast TV
Now imagine the same movie is launched directly and only on the OTT platform, breaking through this value chain. This shrinks the entire chain into just one step thereby, reducing the ability to make money off of a particular piece of content.
Running a streaming service is hard. Going direct-to-consumer might sound easy but is very hard to pull off - one needs to build & maintain the technical infrastructure but also at the same time worry about customer acquisition & churn along w/ customer support, things that didn’t matter when you were just selling content.
Most media industries excluding music don’t have a record label business but around the flavor of the season. Put another way, an inventory of historical classics works well for the music industry while the content streaming industry is a lot more around novelty (consider a rough Pareto being applicable here).
Enter Gen AI. Non-deterministic compute could shake up a lot of things here. If we’re able to enable creators to be able to produce new, novel content faster, we could be looking at a whole new set of economics as we bring down the cost of production.
A fun question to think about is what happens in that world. Some hypotheses -
We shift to consumer-owned, hyper-personalized content as opposed to consumers consuming content in some shape & form. End-users could get desired endings to their favorite shows or get their own choice of fictional characters to do specific things.
We could likely get to a world where quality content production at scale becomes normal & we’re able to leverage powerful recommender systems to actually get to extreme hyper-personalization where each individual user sees a very different feed tailored to their particular interests.
Live media like sports remain as the only “original content” & the value of it will increase significantly.
Advertising will likely be hyper-personalized. Every unique user will get a set of unique ads & commercials.
What do you think? What did I miss?
Thanks for reading The Bizit Newsletter! Subscribe for free to receive new posts and support my work.
Share this post
[#23] The Media industry & the Generative AI opportunity
Share this post
Content creation & media is an interesting business. It costs a lot upfront to produce but the marginal cost of showing it again is effectively zero. This essentially means the more times I can find ways to show the content in the future, the more $s I can make and thus, the more I can afford to spend upfront.
The no of times I could show any piece of content to the audience was determined by my available distribution. Typically, you would sell a movie to as many distributors as possible. Historically, the value chain in descending order looked something like this -
First-run/Premium theaters
Hotels & airlines
Pay-per-view
DVDs
Cable
Broadcast TV
Now imagine the same movie is launched directly and only on the OTT platform, breaking through this value chain. This shrinks the entire chain into just one step thereby, reducing the ability to make money off of a particular piece of content.
Running a streaming service is hard. Going direct-to-consumer might sound easy but is very hard to pull off - one needs to build & maintain the technical infrastructure but also at the same time worry about customer acquisition & churn along w/ customer support, things that didn’t matter when you were just selling content.
Most media industries excluding music don’t have a record label business but around the flavor of the season. Put another way, an inventory of historical classics works well for the music industry while the content streaming industry is a lot more around novelty (consider a rough Pareto being applicable here).
Enter Gen AI. Non-deterministic compute could shake up a lot of things here. If we’re able to enable creators to be able to produce new, novel content faster, we could be looking at a whole new set of economics as we bring down the cost of production.
A fun question to think about is what happens in that world. Some hypotheses -
We shift to consumer-owned, hyper-personalized content as opposed to consumers consuming content in some shape & form. End-users could get desired endings to their favorite shows or get their own choice of fictional characters to do specific things.
We could likely get to a world where quality content production at scale becomes normal & we’re able to leverage powerful recommender systems to actually get to extreme hyper-personalization where each individual user sees a very different feed tailored to their particular interests.
Live media like sports remain as the only “original content” & the value of it will increase significantly.
Advertising will likely be hyper-personalized. Every unique user will get a set of unique ads & commercials.
What do you think? What did I miss?
Thanks for reading The Bizit Newsletter! Subscribe for free to receive new posts and support my work.