A Second Act for Audiobooks Draws Publishers’ Attention

One of the dominant themes to emerge from the 2023 Digital Book World Conference held in NYC last month was a belief in the unrealized value of audiobooks to publishers, authors, and audio listeners. Several talks given at the event revolved around this theme, such as “The Third Pillar: Spotify and the Audiobook Market,” “Synthetic Voice And The Future Of Audiobooks,” “AI-Enabled Quality Control for Audiobooks,” “The Audiobook Phenomenon: Accelerating Your Audio Business,” and “Rethinking Your Platform: Monetizing Audiobooks and Podcasts.”

It was evident from the tenor and the content of these conference talks that audiobooks are moving toward center stage in the publishing world. A big catalyst for this trend was the announcement just over four months ago that Spotify was launching audiobooks on their platform as a new category of audio content. They cited two main reasons behind their entry into the audiobook market: the 20% annual rate of growth in the category and the prospect of engaging a large, untapped audience of listeners in audiobook content for the first time. The launch was described as “just the first iteration” of audiobooks on their platform. Perhaps most notable of all, the company said it intends to “innovate on the format to benefit listeners, authors, and publishers.”

While audiobooks have been around for a while, the user experience has not changed much over time, and, at least until recently, there has not been much innovation in the space. With Spotify’s recent expansion into audiobooks as a new content format and its commitment to innovate on the format, and the arrival of frontier technologies offering new possibilities for publishers and authors, interest and excitement in the field are gathering pace. Let’s explore what kinds of opportunities are brewing for the publishing industry at large, and where things may be headed in the future.

Solving Audiobook Discovery

In a world awash with digital content and entertainment, audiobooks can be hard to discover. A good approach to solving this problem is the use of recommendation engines, which continue to advance in their sophistication and ability to provide highly personalized recommendations to users.

For example, last year, Scribd enhanced their recommendations service with a new personalization model using embedding-based retrieval architecture, which resulted in an increase in the number of users who clicked on a recommended item, an increase in the average number of clicks per user, and an increase in the number of users with a read time of at least 10 minutes (in a three-day window), boosting levels of user engagement.

Blinkist, a company that distills non-fiction books into 15-minute summaries of key ideas, started with a collaborative filtering algorithm for recommendations but has since focused on refining and improving their personalization model with content-based recommendations. These efforts stemmed from a desire to make their recommendations more customer-centric because, according to Melisa Kleen, Head of Business Intelligence and Data Science at the company, “With collaborative filtering, you always highlight the popular books. That doesn’t enable the deeper learning that a group of customers wants. They want to go deeper into certain topics and not stay on the generalist level across many different things.”

Spotify’s large global customer base puts them in a powerful position to provide valuable audiobook recommendations to listeners, given the amount of user data they can pair with audiobook metadata, making audiobook discovery easier for users. Their prior experience using technology to solve music and podcast discovery can also be brought to bear for audiobook discovery. It is a challenge they have worked to solve before by leveraging their users’ music listening data and history as a bridge to develop custom-tailored podcast recommendations. And in terms of audio content similarities, the leap from podcast to audiobook recommendations is arguably smaller than from music to podcast recommendations.

Spotify has a proven track record in filtering, analyzing, and leveraging audio content metadata (e.g., track title, author and producer credits as well as mood, style, and genre tags, and more) for their music and podcast recommendations engines. They also incorporate text analysis using natural language processing (NLP) models. NLP models analyze song lyrics and audio text to establish the prominent themes and the general meaning and scan for potential “clues” mentioned in the text that might be useful in creating relevant recommendations. Their deep, years-long experience building a complex and intricate recommendations system for music and podcasts, with multiple algorithms and models employed across various levels of data, should, in time, translate into a highly effective personalized audiobook discovery experience on their platform.

Social media is another important force for audiobook discovery and has already played a big role in elevating levels of book interest, especially with younger generations. Booktok, a subcommunity of the popular Tiktok social media platform, leads the way in this regard and has become a driver of growth for the publishing industry, earning over 100 billion views to date.

Solving the audiobook discovery challenge could dramatically increase the size of the global audiobook market by helping to convert the much larger pool of today’s music and podcast listeners into audiobook consumers. This arena should therefore continue to attract high levels of investment in innovative solutions to help consumers to find the right kinds of audiobooks, and the interest of many players in the publishing industry.

Measuring and Analyzing Audiobook Consumption

A strong focus on the need for better and broader analytical information on audiobook consumption is powering development in data collection and analytics tools from various audiobook platforms and technology companies. These tools will become ever more valuable to publishers as the audiobook market grows.

Zebralution, a digital audio content distribution company based in Germany, developed an Audiobooks Analytics tool to “help identify our audiences and their behavior. This data supports our publishers, for example, in content creation and marketing to consider the audience demand.” Zebralution also believes their analytics tool raises consumer interest in audiobooks by improving the relevance of recommendations.

Last year Veritonic, a leading audio research and analytics platform, released a first-of-its-kind API for audio scoring. Their API provides publishers and agencies with access to audio measurement scores for several performance indicators, including recall, engagement, and intent for every asset in their library. Data from their audio-first measurement solutions can also help sponsors of audio content to increase the ROI of their audio campaigns, acting as a driver of audiobook sponsorship revenues for publishers.

The demand for audiobook measurement and analytics is not lost on the large tech players either. Performance data for every audiobook is expected to be made available to publishers as Spotify’s audiobook platform is rolled out across all markets, according to Nir Zicherman, in an effort to address the industry-wide need for clearer and better analysis and transparency.

Text-to-Speech Platforms and AI Narration Tools

There is plenty of exciting innovation taking place in the field of text-to-speech platforms and AI narration tools that make it possible to lower the cost of audiobook production for publishers and authors and offer new sources of revenue to production stakeholders.

DeepZen, for example, is a company working on generating synthetic voices that replicate human emotion for AI-powered book narration. They also license voices from professional voice artists in return for royalty payments to those artists when their voices are used to produce audio content. Their platform provides a user-friendly publisher portal, and their technology currently supports four languages. Smaller or more independent publishers and authors may be particularly drawn to digital voice solutions like DeepZen that reduce the cost and time pressures of studio recording sessions. And the fact that DeepZen is licensing professional artist voices in their solutions demonstrates that AI-narration tools can serve as a potential new revenue stream for narrators and voice actors rather than displacing them.

Another interesting company in this space is Pozotron, who helps publishers to produce quality audiobooks in less time and at a lower cost. They provide a suite of data-driven and AI-powered software tools for publishers that help narrators with script preparation, perform automated audio proofing in multiple languages, and provide audio analysis and narration feedback to authors.

The AI technologies underlying these text-to-speech and AI narration solutions are not perfect, especially in terms of the quality of their auto-narrated content. But they will continue to emerge and mature, enhancing the value proposition of audiobooks for publishers and authors in new and exciting ways. To name just a few of them:

• Audiobook production with advanced AI allows for faster and cheaper production of books in multiple languages, bringing titles to new book markets. As Bookwire explained recently, “In terms of accessibility and backlist, text-to-speech offers a rapid and cost-effective way for publishers to tap the full potential of their catalogs and make their titles available for everyone.”
• Small publishers and self-published authors, who may be interested in making dozens of their titles available as audiobooks but cannot afford the traditional cost of hiring professional voice artists to narrate each novel, have more affordable options to do so with AI technologies.
• A virtuous circle effect is taking place as the growth and advancement of text-to-speech platforms and AI-narration tools are creating a wider and more interesting range of opportunities for publishers and authors to participate in audiobook production, expanding the marketplace for audiobooks further.
• The launch of Google Play Books’ auto-narrated audiobooks technology last year was broadly welcomed by publishers who said it presented real opportunities for accessibility.

Looking Ahead

Thanks to the arrival of exciting new technologies, better software solutions, and ongoing efforts to innovate in the audiobook space, publishers have never had more opportunities and possibilities to monetize audiobooks and greatly expand the universe of audiobooks on the market. Given these developments and the fact that audiobooks are now the fastest-growing format in publishing, audiobooks may be poised to enter a period of rapid growth akin to that of the podcasting space over the last few years that will have a lasting impact on the book industry.

With greater innovation in media and entertainment content moving forward, different content mediums are likely to merge, crossover, and evolve in the future. Some audiobooks may follow a similar adaptation process to podcasts that have been turned into a TV show, movie, or book. The rise of an original audio content medium designed to be heard, not read, and consumed more episodically than the standard audiobook format, may also emerge more fully. Audio narrative experiences may also start to encompass more hybrid formats, combining storytelling, interview conversations, and narrative commentary, as when Malcolm Gladwell developed an interview with Paul Simon into an audiobook experience last year.

Audio content platforms, like Spotify, Audible, Scribd, and others, will continue to strengthen their tech capabilities in order to provide publishers with all manner of audiobook services, from hosting, distribution, and promotion to measuring their consumption and potential and to deliver the best possible user experience to their audiences. Today’s largest tech players in the audio content market, Amazon and Apple, may go even further towards building their own seamless audio content eco-systems through tighter integration of their audio delivery platforms, voice command software (Siri / Alexa), and endpoint devices (AirPods / Echo Buds).

From a longer-term perspective, perhaps the most significant anticipated development on the horizon is a substantial rise in the level of audiobook production. As new text-to-speech technologies lower the cost of audiobook production for publishers and authors, many more audiobooks will become available to us in the coming years.

Not only will there be a wider array of audiobooks to choose from in the future, but the discovery process will become easier, and there will be more ways to access and listen to them as more platforms and apps for audiobooks become available, add new features, and improve the user experience. This means we will be listening to audiobooks much more often, at all kinds of hours and from all sorts of places, because these platforms and apps reside on mobile devices that for much of the world are now ubiquitous and integral to our daily routines.

As more of us spend more time with audiobooks, the desire to connect with others through books will only grow. Digital audiobook and social media platforms will help foster those connections through online book communities, book clubs, virtual events, and moderated discussion panels, where the journey of engaging with a great book can be shared, explored further, and deepened with others. There is a promising future ahead for audiobooks and their publishers, authors, and consumers, and that should excite all of us.

Author: Doron Fagelson,
Vice President of Media and Entertainment Practice at DataArt

Originally published on https://www.dataart.com/blog.