I don't think it has anything to do with their engineering. Essentially the global annotation product failed, so they doubled down on rap and turned into a media company. As an early fan of the product, I'm still not sure how to feel about all this, though admittedly you could see it coming even from the outside over the past 6-12 months.
It's always been a little confusing to me how annotation could be a defensible product. A large and active user base is, of course, inherently defensible, but the user base of people who will make animated GIFs to explain Jay-Z lyrics has very little to do with the user base you want to explain the AHCA.
I like the idea of annotation and I even sort of like Genius's implementation of annotation. There's a pretty decent annotation of Gatsby on the site. Virtually every line of The Waste Land has a good annotation there too. But, not to put too fine a point on it, this stuff all seems like, from a platform perspective, something Dustin Curtis could have built solo.
If you could turn the whole web into a vibrant, engaged, participatory community and do it all on your platform, then suddenly you're rivaling Facebook. That's a $100B business right there.
It's just that annotations, by themselves, fail to do that. Their VCs must've been betting either on some new killer feature that would come out, or that users would react to the existing annotation product differently than they did. For a VC, though, "could rival Facebook" is often more important than "has a small chance of doing so".
I don't think the userbases are really that fragmented. The popularity of meme annotations slowly faded away to more quality comments and official annotations from rappers, producers, etc. Personally, I annotated rap songs, tech news, as well as "boring" documents.
But the annotations community outside of rap songs didn't take off in the same way.
I think its because outside of rap there isnt many songs with the lyrical density and complexity required for anotations to work. Even most rap songs are very up front about what they are talking about.
Their only other major play is annotating things like scientific papers but I cant see the market for that being big enough to make money.
Scientific papers is actually the very first market I had in mind when Genius launched as at the time I was doing research and ramping up on image processing / computer vision literature. There'd be so much value in having annotations on research papers to speed up that process.
