In the interest of not repeating the same mistakes if another book- or literature-related site is proposed: why did the first incarnation of Literature Stack Exchange (2011–2012) fail? What should we learn from this? How can we avoid repeating the same mistakes?
Since I found this site, I've never been able to square with its site definition. That's a lot of what kept me away.
On the one hand, the most popular topics - recommendations, story identification, and reading order - seemed to me inconsequential and uninteresting. Their utility is obvious, but generally speaking, these are the opposite of expert questions. These are (usually) trivialities and everybody-can-answer questions. On a more active site with wider scope (e.g., SciFi), these questions might be a mere nuisance in my own opinion, but when they seem like the primary focus of the site, they drown it out. It becomes a site that discourages visitors who were hoping for more than that.
On the other hand, the SE format seems an extremely poor fit for getting into deeper literary discussions. Analysis of a work can be structured very naturally as a dialogue; I honestly don't see what can be gained by shoehorning it into a question/answer, problem/solution format. I never have. And while I wasn't very active on the site, I did peek in from time to time, even gave a shot or two myself, and never saw anybody manage to get a serious analytical discussion off the ground at a level that interested me — and I think the SE format is the primary reason for that.
Regarding both these categories of questions, my feeling is that these questions work just as well or better on a regular forum than they do on a strictly-formatted SE platform. This being the case, SE hampers rather than helps.
I can see two ways out of this in the future - though I'm very dubious about both of them. One is, if you've got a large enough userbase interested in your particular content, then you're good to go. Even if I don't like Lit.SE's scope, if enough people do, then maybe you've got a site. (This is my feeling about another SE or two out there, which I feel have a trivial scope, but there are enough people out there who are genuinely interested in it.) But there's really no good way to guarantee that — that userbase might well simply not exist.
The second is, if you figure out a type of question that works really well on SE, you could base your site around that. Again, that type might not exist, or might not have much of a following. But if you did find one, it'd be a new direction. One of the reasons I'm optimistic about SE betas in general is because it's possible, through experience and effort, to gradually develop new types of questions - questions you couldn't even think to ask beforehand, because you never had the venue or the culture for it.
In conclusion, an observation from my own home field that'll be pertinent here too. A beta is an experiment, not a foregone conclusion. That means that, in all probability, you didn't make any mistakes. You tested the viability of the concept, and you found it. Maybe you've gained some important insight into types of questions that can work, or what type of post gets what type of response. Maybe that insight will help form some other related proposal — but maybe this topic really isn't a great match for Stack Exchange.
On the other hand, if you think it's a great match for SE, capable of producing great content, well, then should be able to describe what and how. And that'll be the next proposal.
I think there was a mixture of things.
I think the very first thing was letting recommendations be on-topic. With 85 out of 434 questions they were about 1/5th of the site, but drew little traffic. People get tired of looking at stuff that doesn't interest them.
I think another issue is that we lost a lot of traffic to sci-fi because it shares such a huge portion of content with literature.
Personally, I couldn't find a question for this site that I couldn't ask on SFF... Most of the people on Stack Exchange, despite intentions to the contrary, are technical. Technical people tend to like SFF type books more than general purpose. So at least a part of it was the overlap with SFF...
So, what I'd say in the end is, this topic is just too broad. Broad topics sometimes work on Stack Exchange, but I'd dare to say this is the most broad of any... Also, it's really hard to come by good questions from books, it takes a lot of time and effort to do so. I am having a very hard time finding a book series only on SFF that is popular... I think what happens is finding good questions in books just takes so long, that the question count goes low. If there isn't enough questions, then people stop visiting the site, and then they just start to forget about it.
SFF thrived first from movies and TV series, and then expanded into books. There was a time where Star Wars and Star Trek were the range, but it took much longer to get the Harry Potter questions, and Lord of the Rings. Further more, both of those have movies associated with them...
Perhaps if there were a few smaller sites, one around which people could get a bit more excited. The scope would be a bit tricky, but perhaps Mysteries, Romances, Classic, etc. These would have to allow for various medium as well. Of course, I don't doubt that some books defy the normal categories, but you can't win them all...