We talk with writers, editors & entrepreneurs about, really, anything. All conversations are 'manuscript-first', meaning they were typed as you see them. Small edits have been made for structure.

Our complete list of conversations, including:

Pixelated, the digital, double-blind, lit-inclined author chat

A Bit Contrived, interviews with real authors about improvised books

Episode IV: Melting genre

Published 3/4/15
In the fourth installment, I speak with Neil Baptista, the Founder and CEO of Riffle. Topics include book discovery, "lean" startups, the place of blogs in publishing, getting the ear of publishers, the future of genre & more.

(Andrew Lipstein) I’m here (digitally) with Neil Baptista, the Founder and CEO of Riffle. We’re a kilomile away—to bridge the digital divide, would you kindly describe your current environs?

(Neil Baptista) ‪It's dusk in New York. We're enjoying some fresh breeze in our office, as our 100 yr old cast iron radiators are pumping out so much heat.

The best form of heat by far. For the uninitiated, would you give us the elevator pitch on Riffle?

‪Our mission is that "Riffle inspires people to read more" It does that by connecting you to people who are passionate about books. What floor are we going to? I've got more pitch in New York.

‪I'll give you a few extra floors—how exactly, as a user, am I going to derive value from the platform?

‪Early on we were very focused on this idea of book discovery. We looked at algorithms and content to drive that.

What we found in the end is that people that connecting with people who are passionate about reading and about books is the most powerful driver. A large portion of our early users and core community are librarians, bloggers, editors at publishers, booksellers and authors. Their excitement and pace are contagious. Their reviews, comments and recommendations create trails to fantastic books.

‪I'm interested in this process of pivoting. I know from talking with you previously that you’re a subscriber to the “lean” principles of business-building. Can you tell me a bit how that’s informed the fruition of Riffle?

‪The fundamental principles of "lean" were laid out in the Lean Startup by Eric Ries, although the principles have existed in manufacturing for much longer.

We basically follow a build, measure, learn cycle with everything that we do. We test our assumptions about what readers want by doing interviews with them, with surveys and by watching how much new features get used.

‪It's an organic process, so we release new things constantly, as opposed to the traditional approach which was to release software every 6 months to a year. ‪It is primarily about testing assumptions. It sounds simple, but it actually requires a lot of rigor.

Was there a memorable release that didn't work in a big way (and in turn, gave you insight into your user base in a big way)?

We had an assumption that bloggers would want a tool to curate a bookstore for their blogs and then share in the proceeds that drives. Something beyond what they could build themselves using a blogging platform. Then, basically every blogger would be merchandising their own storefront.

Sounds similar to what's going on at aer.io.

‪It would give bloggers a strong reason to be on Riffle and to provide reviews and other content. 

But, what we found after interviews and a couple of deep surveys is that smaller blogs just wanted to connect with a community and have some comments. Larger blogs where they were monetizing their traffic already had ad tech and an affiliate strategy. The medium sized blogs were just trying to build traffic by focusing on social, better content and guest contributors. We saved at least a couple of months of development by not going further down that path. It also led us to double down and focus on connecting with other readers.

The primary reason that people are using other social book platforms is that connection. The ability to check in after reading a book and continue that experience in the comments or discussions.

Saying the digital publishing landscape is “shifting” is kind of like trying to walk across a pool and saying that the water is “shifting”. How much pressure is there to keep up with startups, both known and unknown?

‪We definitely survey everything new that comes out. The book business is counter intuitive in many ways. So there is a lot to learn for any new startup entering the space. Especially, as the landscape is shifting. We've been in this space since 2009, so we are kind of veterans in some sense.

All the big publishers are definitely quite shrewd about startups. They are aware, but they also know their business very well. When something tips and becomes successful, they start to take notice.

‪How much of "tipping" do you think is "awareness" (i.e. social media) and how much is concrete data (sales, members, etc.)?

Press certainly doesn't hurt. Neither do buzzwords like big data or real-time marketing. I don't think that publishers want to be left out of any of the waves of tech. Although, members is certainly at the top of everyone's list in terms of assessing a new player. It's a bit difficult sometimes, as 'retention' is really the key metric for a startup until they have a product-market fit. That's where the real pressure is. That is to grow membership. But if you look at most of the successful startups out there, they really focused on activation and retention, before they scaled their user base. 

Getting publishers to take notice before we've scaled has only been possible because of our long relationships with them.

That makes sense. That feels in line with thinking "lean"—make sure your assumptions are sticky with users, before worrying about gathering as many as you can. Now I want to switch gears a bit, and talk more about Riffle's building process—specifically how you use genre, an undeniably crucial part of your structure. It’s a tricky issue in publishing, the thorniness of it compounded by its necessity. How much internal debate was there about how to demarcate your genres?

Ah, you would love to be a fly on the wall for those conversations. We've debated ONYX files, BISAC codes, tagging strategies and taxonomies. We have granular lists that break things up into 249 categories and our top line list of genres that is 25 deep. We've decided that it's something that will evolve. So, we've started with identifiable genres that readers can relate to based on how books have been merchandised in the past. As, the data grows the categories and genres will melt away.

‪Melt away, in that we won't even be using words to describe genre, or melt away in that we'll be searching for "literary thrillers involving chefs, unwanted pregnancies and Los Angeles"?

Yes. We'll allow you to drill down and categorize by genres if that's how you think about books. But allowing for search is also on the roadmap. I'll keep your example in mind as a test case. Ideally, though, you'll be relating to people on Riffle and what they recommend or review, so in that sense categories as a construct aren't really needed.

‪It seems like the future might hold a place where we're "tagging" books in the same we are tweets. As data (meta and otherwise) become more common, what risk is there in shuffling past work into the "classics"? (Kind of like how when you turn on a "classics" radio station you'll find back-to-back songs that would have attracted entirely different listeners decades ago.)

‪The complexities of applying data to books are immense. It's one of the reasons that most recommendation engines fail. There are too many nuances to try and account for. It turns out that people are still the best at sussing out relationships between books and making suggestions. We have a simple visual that shows you the books that you have in common with another user. That gives you a quick sense of what your reading overlap is. Are there a couple of best sellers mixed with some obscure fiction? That's a jumping off point for us that we believe is just much stronger than going with "classics" as a tag.

‪Agreed.

‪I'll be the first to admit that the math and promise around algorithms is alluring. We've just learned from readers that even if an algorithm predicts something that they might like to read, mostly they don't have any connection to that recommendation. It's a bot.

‪Excellent answer.

We only have a few minutes left, so I want to move even farther into the deeply abstract. Riffle’s mission isn’t just to capture readers, but, hopefully, to create more readers (or at least “avid” ones). What’s your objective, cold prediction: is the world gaining avid readers faster than it’s losing them? Evidence?

We are definitely in the midst of a media and screen overload. Pew Research seems to say that people are reading more books. In general, more of the world's population is becoming literate. So, in absolute terms there will be more readers. Although, I don't think that there will ever be enough avid readers. What we do know is that many people who are avid readers like to express themselves about what they read. Like Instagram forms community around images, Riffle intends to support community around books, including quotes, discussions and experiences from readers as the central media elements. If you are a reader, your bookshelf if part of your identity and we're hoping to support that expression and make that infectious to others. We'll measure everything we do against our mission to inspire people to read more.

That's a mission I think we all (or at least everyone who's reading this) would agree with. Thanks for playing Neil!