3 ms·
It should be obvious that technological change allows content optimization in a way that was never possible in old media. With a group of different headlines, y
by Tossrock 12y ago
It should be obvious that technological change allows content optimization in a way that was never possible in old media. With a group of different headlines, you can apply A/B testing / multi-armed bandit in real time, and watch the results. With a TV show, you had to make the show first, and then you saw the results from Nielsen after it had already aired. The web allows analytical depth on content that was never possible with newspaper / TV / movies, and the proof is in the pudding - Buzzfeed's rapidly growing revenues, the proliferation of content optimized viral sites, etc.
- jgalt212 12y agoYes, it definitely seems like it would be easier responsive to consumer tastes and preferences in the online world. That being said, you're still trying to model human behavior which as I indicated earlier I believe is largely unmodelable (sp). Of course, any fool can A/B test to show that readers prefer interesting articles over boring ones. That being said, it's very hard to ex ante produce interesting content. Or even enough content such that a high enough per cent is interesting to the consumer. For example, how many articles have we seen on people leaving crazy good/bad tips and the server taking a phone pic of such tip as evidence. I think I read the first few, but they still seem to come out every single day. Are people still clicking on them? HuffPo says they are. Why else would they keep running them? However, I really think those sort of stale genres are now only being clicked on by bots.