Hiptype (YC S12) Launches A Google Analytics-Style Service For E-Books | Hacker News Reader

Hiptype (YC S12) Launches A Google Analytics-Style Service For E-Books

techcrunch.com

90 pointssohailprasad32 comments

I'm just trying to imagine how a book like Catcher in the Rye, Finnegans Wake, or To Kill a Mockingbird would have been changed if it were subjected to these sort of analytics by its publisher.

I can easily see future books being edited and re-written from edition to edition (not that "edition" will continue to have any meaning) in response to analytics on reader behavior.

30% of readers failing to finish a book, and dropping off near the same place? Let's rewrite that chapter to keep people turning the pages. Lots of people highlighting a controversial passage, then not buying any more books from that author? Tone it down, smooth out the "problem areas." A surprising number of young women reading a particular book? Let's "pink it up" and throw in some romantic subplots that tested well in focus groups. How about embedding some multimedia into the book, and some product placement while we're at it?

Literature will survive this latest attempt to measure the creativity right out of an industry. Good movies still make it on the indie circuit despite the focus-grouped blockbusters, good games still get made by small shops. I just can't help but feel a tinge of sadness as I watch books and literature get consumed by obsessive measurement and cold-blooded analytics.

Jayson, thanks for your feedback. You raise an important issue.

A huge proportion of the eBooks downloaded are educational, trade, and reference. Cookbooks, for example, are incredibly popular and these "cold-blood analytics" can be very practical for determining what content is working and what content is not engaging readers.

It gets more tricky when it comes to literary fiction. But I think it's safe to say that measurement and creativity aren't mutually exclusive. Measuring something in and of itself doesn't compel you to make changes.

But it would be interesting for us to do some research by getting into an edition of a classic text and seeing what types of changes our analytics would suggest making. This would mostly be an exercise in figuring out how to improve our analytics, rather than taking it at face-value that a classic text should be edited based on some data points.

I feel your pain, but look at it this way; there were plenty of crap books before this idea as well, the silver lining is maybe having more enticing crap books will eventually lead people down the path to good literature.

I expect to be a lone voice in a sea of positivity...

The very notion of A/B testing a piece of art disturbs me profoundly. The idea that publishers would say, "Our analytics show readers engage with these types of passages so you need to increase your usage of X passages by 50% and use 50% more words from this list", paints a fairly dim view of the future of artistic expression in a world that is concerned with profitability over substance at an increasing rate.

Publishers already have a far heavier hand than I'd like them to or most people know about. Giving middle manager types data when it comes to artwork can only lead to place where self expression through authorship becomes a dead art-form.

Actually, as a reader, I would prefer book publishers to know what gets read in their books and what doesn't. Likewise, I would like to know that information as a reader.

There've been countless times when I purchased a heavily advertised (well reviewed) book, only to find out that it was worthless piece of ... Anyways, those books weren't written for me; they repeated the same stuff over and over and over again.

One example - Good to Great. The first chapter is good, so is the second (maybe). Everything afterwards is just a repetition of the same concept. If I could see ahead of time that most readers stop reading the book 20% in, I'd make my assumptions and not even bother to try to get through it.

Likewise, there are books where the first chapters are quite tedious to read, but the rest are great. If, while reading chapter 1, I could see that most people struggled but then persisted and loved the book, that might encourage me to get going.

Just my 2 cents.

We find it very fascinating that people either love Hiptype or are profoundly disturbed. What we're doing is not different than what occurs on the websites and mobile apps you already use, so the difference is one of culture.

What I'm most curious about is the extent to which these cultural norms will shift as it becomes more common for books to contain analytics like ours.

Preface: I hadn't considered technical books and all comments are about stories, both fiction and non fiction.

As someone who has grown up on the internet for the last two decades there is something decidedly different between a website, a blog, and a mobile app when compared to a book. It's something ephemeral I can't put my finger on.

There is magic in books. From a child's earliest days books provide a window to other worlds and realities. Some of these worlds show us glimpses of greatness and happiness others show us the darkest depths of human sadness and despair. We are given characters to love and hate along side characters to use as reflections of ourselves. In all of this there is a certain magic to books. They provide judgement free escapes for the outcasts and loners; Later in life they provide the community for those very same outcasts and loners to come together.

I realize I sound naive and possibly unrealistic. Also, I acknowledge that this community in particular tends towards efficiency and has a tenuous relationship with the humanities. However, all that aside, this feels like it cheapens the reading experience.

I hope this doesn't become the norm for authors telling stories. While perhaps a bit more hyperbolic than necessary this all feels akin to A/B testing The Sistine Chapel. Let's not go around letting MBAs and Middle Managers fuck with art.

You might just be A/B testing your samples or your jacket text. I'm guessing a lot of authors would agree with you that they don't want to change specific passages in their stories. But I see a lot of them working on making their sales better by providing better blurbs. It's the equivalent of being able to change your outside back cover text to get people more interested.

For fiction and other works intended to be purely art, sure. But for technical, explanatory books...seems like better analytics could be a huge boon.

I agree. If people skip to a section of a coding book all the time, an author could decide it needs expansion or might base a whole other title on the subject.

Of the companies I've read about so far, I believe this is the one to watch. I'm expecting this to be a big deal (I don't know the founders or anyone else even remotely involved). If they can get the right deals (be where authors sell the books and readers read the books), then writers will quickly become addicted to this service.

Ok, the "Hip{something}" naming madness has to stop.

I hope they allow a mechanism for people to opt out. I can't really see this helping the reader, or even proper publishers. It will just aid the Demand Media equivalents of the publishing world in churning out crap.

There is a privacy notice in every book with Hiptype that allows readers to opt-out. We want to make sure that we make it easy for readers who do not want their anonymous usage statistics tracked to disable Hiptype.

I find it interesting that they call themselves the "Google Analytics for e-books" since, while it is true that often the out-of-the-box analytics offered by big publishers is extraordinarily basic (number of downloads, number of opens), the one I have worked with - Adobe DPS, used by Conde Nast and others - does allow you to directly bolt on Google Analytics to your publications, in addition to their own Omniture offering.

Analytics. Another reason to use alternative channels.

(As for anononymization, as they say in security communities, once you create a means of access, who knows how and by whom it will end up being used?)

I'm glad to see someone from the business actively participating here; however, I remain unassured, in the larger context and perhaps also specifically.

A great idea but there is no information on their site about what data you actually get. Maybe innumerate literature types don't care, but us data heads certainly do. I would want to know what sections people read, what they don't, and critically I want to know what the distribution looks like, not just the mean.

Great service and I am really looking forward to see where it goes. When I am ready to write a book, I am going to use Hiptype, no doubt. Will you let me do all the writing+publishing on Hiptype too? That would be really nice!

Congrats James & Sohail!

Thanks! It's incredibly exciting for us to finally share Hiptype with the world and start getting more authors and publishers using the product.

We've been particularly interested with what the HN crowd would think about things made possible with Hiptype, like A/B testing or beta testing versions of a book.

Could this work in a hybrid app for iOS? Specifically, we publish a hybrid iPad app using an open source framework called Baker (http://bakerframework.com/). I know a lot of people publish books using it as well. Only problem is, Google Analytics doesn't track HTML pageviews inside an app, and most in-app analytics are geared towards native apps. Setting up analytics is painful.

We might be able to help, even if we don't currently support this environment. Feel free to get in touch at james@hiptype.com

Congrats guys! Analytics for books. This is such a great idea...I'd thought it would have already been created. Good luck at Demo Day.

Hmm, I visited the site, but it appears you haven't actually launched? It asks me for my email to register for an invitation.

We're inviting users in weekly batches. We're using a fifo system so signing up earlier guarantees earlier entry.

We appreciate your patience, and you can get in touch directly if you'd like us to get your account activated more quickly.

Grats on the launch guys

How do they determine the income of the readers?

This is a good question. None of this data is available from the book itself. We've invested a lot of effort into our analytics backend, and our ability to derive demographics info from a very small amount of uniquely identifying information on the eReader or multi-purpose device.

There are a few reasons why we're not explaining how we get all of the demographics and insights data shown on our dashboard, but we can vouch 100% for its accuracy, and it's all data that is publicly available.

How does the developer get your code inside their book? Also, you might want to consider an intermediate plan. We're over 1000 readers (after a few months), but I'm guessing we're years away from 500,000 readers.

Adding our plugin to books is easy - you just upload the book file (using filepicker.io from our batch) and it is automatically modified to include the plugin.

Also, thanks for the feedback about pricing! We're looking into some changes that would address your concerns.

I'm guessing this info isn't on the website because you haven't got around to it yet, but knowing how things work is what attracts me to services. I'm wary of companies that claim to do things but don't say how it works, so you should consider putting up a "How-it-works" page. Do you have a list of supported formats? Just another comment about the pricing, 500,000 is a LOT of readers. Very few indie publishers (at least single authors) are near that number for a single book. Having a range for authors between 5,000 - 100,000 might be a better sweet spot.

Credit card tied to the purchasing account. Once you have that, you have (for sale) access to in-depth profiles I would assume include more or less accurate assessments of income. Presumes this is being done in coordination/through the e-tailor.

Reply on news.ycombinator.com