The Painful Stagnation and Inevitable Rebirth of Social Reading

Understanding the Future Ecosystem of Connected Readers

Think, for a second, about how we read and communicate online. Since the dawn of Web 2.0, we’ve commented liberally on any content that crosses our screen. Whether it’s a news alert, a Facebook post, or political turmoil on Twitter, we expect to be heard — we expect to consume the wisdom, or folly, of others, and we expect to comment on it.

Now, think about books. It’s odd to leave our commentary everywhere except inside books, since they are our deepest, most valued repositories of culture and education. Inside the very format that carries our ideas across generations, we don’t argue, agree, commiserate, or reminisce with like-minded readers. Compared to other forms of content, looking for conversations in published books puts you at the center of a vast wasteland. You are alone.

Compared to other forms of content, looking for conversations in published books puts you at the center of a vast wasteland. You are alone.

A small group of people worked to change that for a decade; they cultivated a vision of connecting people inside the written word. Between 2007 and 2017, more than 25 products for social reading launched, with the goal of offering conversations in the margins of long-form text.

Struggling to get conversations into the margins of books sounds almost quaint at this point. What’s the hold up? In a world where Spotify and HULU serve up my entertainment, it shouldn’t seem crazy for two-dozen sci-fi fans to share their commentary inside A Canticle for Leibowitz.

It’s also true that we can’t talk about social features without addressing concerns of misinformation and privacy, from fake news to Gamergate. These are important, but not a reason to snuff out community. Our experiences are only as good as the systems we create to house them; the conversational systems we design for books can certainly account for the vagaries of human nature — they just need a plan to manage abuse.

Defining Social Reading

To begin, social reading is the discussion of long-form text from inside the text itself. The primary objective is to further discussion as the text is being read. The interface often includes:

  • Comments posted asynchronously: attached to paragraphs, highlighted-passages, or sentences
  • Responses to comments
  • Real-time chat interactions, sometimes limited to chapters
  • Rich media comments (images, videos, audio)
  • Curation (discussion around specific subjects)
  • A roster of readers

The interface, commenting features, and devices are often the focus of critiques. As a user experience designer, I see where people are coming from: interfaces and design matter. That said, the key truth about a good social reading experience is that it depends on two things more than anything else:

  1. Available content
  2. The group, and its dynamics while discussing that content

Available content isn’t a surprising requirement— without something people want to read, social reading experiences are dead in the water. But the group is key, and often overlooked. The content that people want to discuss needs to hold their interest, but the people discussing it must be nudged at the right point, they must be talking to the right people, at the right time, and they must be highly motivated to do so. All design and content decisions must stem from the group.

The content that people want to discuss needs to hold their interest, but the people discussing it must be nudged at the right point, they must be talking to the right people, at the right time, and they must be highly motivated to do so.

Obviously, the group is a complex structure. It’s not just about connections, but also network affects. It’s about the Dunbar number, a rule of thumb that anticipates how much a user shares with whom. The Dunbar number suggests there’s a core handful of people (5–10), that you might share thoughts with on a daily basis, but another, larger group (15–25), that you only speak to a few times a week. There’s probably an even larger network of acquaintances (50+) that you work with and speak to a few times a year. Like everyone else, you share different items with different people based on your relationship. Frequency of sharing, formality, and all the other group dynamics influence the cadence and the tone of the discussion.

Dunbar’s number related to Social Reading activity. Larger version.

The interface must be flexible enough to address individual expectations for group sharing. For some people, marginalia is a serious business, with drafts written, references considered, and spelling checked. For others, social reading is a reaction driven by the direction wind blows, a full-on impulse grounded in emotion. The peer group sets the ground rules and the tone, the interface must rise to meet it. In the end, social reading is about the group more than the features. It’s an ecosystem.

There is a very simple but profound answer to [why people read]: people read books to make connections. By connecting very small patterns together into larger ones, we connect concepts back to the real world around us, to real people and places.
A book and its patterns, and the place we sit reading it, and the person we fall in love with, can become forever tied together. It is at this level that reading interests and addicts us. In that sense, it is entirely social…
Much of the value in books lies in discovering that they are shared experiences, and seeing the spectrum of individual connections to them. It’s in that spectrum that the Web offers so much for the medium of the book. As a network of people and communities, it is also a network of voracious readers.
— Above the Silos: Social Reading in the Age of Mechanical Barriers, 2013

What Came Before

In 2006, what we consider today to be the social web was growing exponentially. Facebook opened to non-student users that year. Registration soared, a flood of humanity registered, eventually 1 out of every 6 people on the planet created an account. Also that year, Twitter jumped ahead of other microblogging competitors, and became a serious contender.

Facebook is predicated on real-life social connections, so by using your real name, and connecting with people you already know, you establish an online version of that relationship. On Facebook, for the most part, groups are clusters of people already in our social circles, with varying Dunbar levels. Twitter, on the other hand, can be anonymous. Users on Twitter don’t necessarily have a reciprocal relationship — you can follow a Twitter user without being followed back. On Twitter, the subject matter, or the author’s reputation on that subject matter, is the axis of connection. Granted, there are exceptions as social networks continue to venture into new territory and compete on the same footing, but to keep things simple we’ll stick with these primary social structures.

About a year after the rise of Facebook and Twitter, social reading startups began to launch. They used either the follow (Twitter) or friend (Facebook) relationship metaphor, or a combination of the two. In that sense, we can say some products were driven by the interest graph, others, the social graph. These group decisions, in turn, influenced the overall user experience of each product.

Social Reading Ventures: Lines of Influence

In our connected world, no industry can be an island. Not even Publishing. The following startups saw the white space, the gap between what consumers would use if they could, and what the industry actually offered, and built products to fill it. The infographic below highlights social reading product launches to date, showing connections of influence along the way.

In our connected world, no industry can be an island. Not even Publishing.
Graph 1: Social Reading Systems, Color Coded by Platform Type, 2007–2017. Larger version.

Although these companies had similar features, there were a few key differentiators. As you can see in Graph 1, most platforms launched on the web, but apps had more sophisticated digital rights management, which made it more attractive to publishers. Often apps had both a web component for discoverability and an app component for reading.

Table 1: Social Reading Platforms 2007–2017. Larger version.

Of the platforms above, groups had various levels of accessibility and focus. Some groups were hidden (Subtext) or required confirmation to join (Social Book), or required users to be initially accepted by the group owner, but then allowed users to invite anyone they wished (BookGlutton). Other systems allowed groups to be completely open (BookShout, The Copia), where joining happened automatically with the click of a button. For Readups and ReadSocial, the group was created with a hashtag, so it was somewhat hidden but still public. Changing the hashtag on a book changed the group and its conversations, but you’d need to be told what that hashtag was.

Medium was the most successful startup to date, in terms of social reading, although it’s billed as an authoring platform. It launched in 2013, and validated social reading, bringing it to a larger audience, demonstrating real use cases, and launching an interface so efficient it just seemed natural, as if we’d always used it.

Graph 2: Hyperbolic Tree of Social Reading Startups. Larger Version

The Definition of Success

Ultimately these endeavors didn’t disappear due to lack of interest by readers. Most failed because they couldn’t solve for one of the two core needs: content or group dynamics. Many of these products started from a simple, noble idea: create conversations around content, so they focused on interface, and content acquisition. The building of many features, like the ability to upload work or write content within a reading system, even the creation of full digital bookstores, were often a response to the content acquisition problem.

In short, the proverbial content acquisition fires were flamed by the issues of ownership, piracy, and the complexity of book distribution, specifically:

  1. Debate over content ownership — who has the right to decide which margin notes appear
  2. Fear of piracy — concern that readers can’t open books in an open environment (like a browser) because exposing text could allow it to be stolen
  3. The indirect path of book distribution — the company that sells you a book isn’t the one who makes it, or decides what features and content should be in it

It’s easier to digest the complexities of the book market if you understand one ultimate truth: publishers don’t sell consumers books, book stores sell people books. This means book stores in the digital sense: Amazon, Apple, or Google.

Consider this: just like a warehouse of paperbacks, publishers require that their digital content be preserved with the proper formatting, protected from theft, and be distributed as needed. They don’t cultivate a close relationship with the actual reader. Reader relationships are owned by the bookseller: Amazon, Apple, Google, Barnes and Noble, etc. These digital booksellers preserve the design as completely as possible, protect it with DRM (digital rights management), and distribute it to the end user.

This brings to light a major point of confusion for social reading startups: it was never clear who exactly to partner with. Social reading product owners felt like they needed permission from the publisher to alter the content, to attach user comments. After all, if user comments would be there when someone bought a book, even if they were hidden as a default, isn’t that considered altered content? Moreover, that startup would also need to partner with the digital bookseller (Apple, Amazon, etc.), to implement the interactive commenting features and group structure inside the reading system. This gave booksellers pause; adding in commenting functionality from a third party platform could create problems — latency, security or privacy concerns, design conflicts. Risk. Besides, most booksellers often thought they’d build these features. (But have yet to successfully do so, since it’s not a core part of their businesses).

Many startups tried to solve for these challenges by allowing the user to create their own content (as discussed previously, to support content acquisition). Others became the bookseller themselves, building a full social reading system, and effectively replacing the bookseller/tech company. Unfortunately, most of these efforts languished in obscurity. Readers don’t want want to have to find and include all their content manually, especially if they have to do illegal things, like strip off digital rights management (DRM) to import it. Users expect to be able to acquire/read what they see sold everywhere else, from the airport bookstore to the subway ad to the banner ad. Also, readers generally don’t want to download a bevy of different reading apps and have content in each one (a fractious landscape that also further complicates group creation). Readers want to use reading systems they already have installed, and tap into groups that already exist.

The orbits of readers and publishers rarely cross, and digital reading was no exception. Today, even if a reading system secures permission from publishers to have conversations inside books, it needs the bookseller to partner up. Or the social reading system needs to go it alone, and build not just publisher connections, but the reading system software and all its requisite features. It’s an intimidating task. But the future of social reading is inevitable, and we’re about to witness convergence on a much more massive scale.

Social Reading is Not a Feature — It’s an Ecosystem

In the end, social reading is a combination of features, content, and groups. It is also a long-term endeavor. Social Reading is a full-on ecosystem, where disparate elements are tied together between users and content, a system comprised of interconnected relationships that someone needs to manage and run, long-term.

It’s the same challenge that casts a shadow over the Internet of Things (IoT), the future of robotics, wearables, and essentially most cutting-edge technologies. You can’t just slap connectivity on something and call it done. When a company releases a connected product, (a connected thermostat, a connected water pitcher, a connected car), the company evolves from a product company to a product + services company. The company needs to handle tech questions when things don’t work. It needs to listen to user requests. Respect user privacy. Plan for security. Innovate and improve the system of communication. Encourage adoption. Figure out how to evolve the service into more valuable services. When listed, the tasks seem daunting, but many companies deal with these hurdles every day. Social reading endeavors will need to, as well.

Eventually, the point of conversion, from reading systems to social reading systems, will be driven by booksellers, and they will own it. Booksellers (the tech companies like Apple, Amazon and Google) will be the ones to establish the future success of online reading systems in general, drawing on all the work done to this point. This is because booksellers have the most power: direct access to users. They have access to content. Moreover, they have deep pockets and real investments in features, machine learning and artificial intelligence, which will become key to making social reading work.

Social Reading is Connected Reading

Considering the history of social reading, the industry challenges, and the complexity of the user experience, what will spur booksellers to incorporate these changes?

Four things need to happen for social reading to graduate to the next level:

  1. Publishing houses need to accept the idea that holding back the flood of discussion is ineffective, and that as long as comments can be hidden by the user, moderated, and contested, this is not a publisher’s battle. Give the lawyers the day off.
  2. Each bookseller needs to build in their own form of discussion, predicated on a set of mutually agreed-upon standards. If someone bought a book, she’s already locked into a specific reading system, but if her conversations need to flow across systems, interfaces should have the same basic subset of capabilities.
  3. These capabilities need to tap into a handful of major social networks as a baseline for accounts. Connecting each comment to a current social network account (Facebook, Twitter, Google, Apple) allows for cross-system consistency. This single sign-on means people don’t need to sign up for new accounts, and conversations can be tracked to their source, for moderation.
  4. Moderation needs advanced technological support. It will be dependent on artificial intelligence (AI) that can help keep trolling to a minimum, with machine learning (ML) capabilities to allow it to improve over time. Google’s Jigsaw, a product that the New York Times is using to help moderate comments, is a good example. Jigsaw evaluates comments on a sliding scale of toxicity, and then partners can choose what they allow to appear next to content.

Of course, that still leaves the issue of user experience. A key component of adoption will be how easy it is to actually use; user experience will be the wild card. By crafting a user experience that focuses on simplicity, allows the group to drive the conversation, and anticipates user needs, these companies will ultimately give social reading the launch it deserves.

User experience will be the wild card.

Fear and The Future

For such a simple idea, putting comments next to words, these four factors that will drive social reading feel heavily invested in technology. But a challenge this big, connecting this many people across this much content, with the overall outcome being a simple interface tool, requires a lot of tech. This is the power of technology, the promise of algorithms: that we have the ability to create something that is more valuable than the sum of its parts; something to connect each of us across the cold lattice of digital representation to embrace something better, to create something that makes us feel like the world is improving.

This is the power of technology, the promise of algorithms: that we have the ability to create something that is more valuable than the sum of its parts; something to connect each of us across the cold lattice of digital representation to embrace something better.

That said, like all technology that stretches beyond our expectations, there will be challenges. What makes it accessible and seamless also makes it trackable and prone to manipulation. What are the inherent risks of connected reading? The trolls, the misinformed, the nefarious government powers seeking control? Those challenges can only be solved through moderation and curation. Like all good systems, there need to be rules for content. Rules for interaction. And the next wave of systems that support these rules will be assisted by artificial intelligence and machine learning.

The genie isn’t going back into the bottle. We should make the governance and improvement of our systems our priority, and build what social readers want: ways to connect.