Challenges in Search on Streaming Services: Netflix Case Study

Netflix is one of my favorite examples of a search service that would be infinitely improved with a really low-tech change that will probably never get implemented because it doesn't give the data scientists something to talk about over beer: Filtering

So often, what I'm looking for is something like, "G rated content, not by Disney, less than 30 minutes long", or "Documentaries that I haven't watched yet, with an audio track in French, that are available for offline viewing". With a search space the size of Netflix's library. There's no need for a search string there. Nor is there a need for anything fancy like recommendation or expanding to items that are similar to direct hits on the search string, because, these days, a basic filter like that will only yield a page or two of results, anyway. And would be really easy to for them to implement, because they already have their library tagged with pretty much anything I could think to filter on, anyway, so they can use it as features for the fancy-schmancy recommendation engine.

More generally, I stopped having a need for a juiced up search engine and recommendation system in Netflix the exact moment I stopped owning a DVD player.

>and support searches for unavailable (out-of-catalog) entities

This is a subtle killer feature of Netflix's search, and something not often done by others. By including these unavailable entities, it allows Netflix's similarity and recommendation engines to still understand what the user was looking for, and provide things that are relevant and similar, even if they aren't exactly what the user is looking for.

That being said, this is a lot easier for something that has a relatively finite, known set of items (movies) in contrast to things with lots of items that can be differentiated in smaller, subtler ways (e-commerce, web documents).