4 Ways Book Publishers Can Take Advantage Of Artificial Intelligence

true

For years, the stock exchanges, manufacturing, shipping, and airline industries were the talk of the town when it came to spearheading artificial intelligence (AI) systems integration. Now, that conversation has shifted to the creative industries.

And yet for all the hullabaloo you hear about AI being on track to revolutionize publishing and transform the consumption economy, for inquiring minds there sure seems to be a lack of detail about how, exactly, it will do so.

The trick to AI is approaching it with one very important factor in mind: you.  What problem do you need to solve? Since AI is built specifically to adapt, the possibilities can seem at once endless and difficult to know where to begin. And with an industry as established, venerated, and culturally significant as publishing, there’s an understandable fear of gumming up the works. But, it turns out that integrating AI into the everyday decision-making of stalwart industries like book publishing is less daunting than it may have seemed.

Here are 4 ways real companies can–and are–taking advantage of AI:

1. To Increase discoverability

Discoverability–how easy it is for a consumer to find a specific product or find what the right product–is one of the most common concerns of publishers today. And understandably: if your audience can’t find your book, how will it ever get read and shared and on the New York Times Bestseller and be picked as one of Oprah’s favorite things??

Only in the past few decades–recently, in the centuries’ old publishing industry–have specific factors shown themselves to be key performance indicators to a book’s discoverability:

  • Thorough ONIX Metadata completion
  • Title (both in substance and formatting)
  • BISAC Categorization
  • Long Tail Keywords
  • Reviews
  • Ratings
  • Social Chatter
  • Comparative Titles
  • Historic Customer Behavior Data
  • Digital merchandising (how the cover comes across in an ebook store setting)

So, now you just go and optimize each of these metrics–measure, analyze, implement, test and then re-calibrate when sales data is available– for each of your backlist and front list titles. Easy Peasy. Except…it’s a lot of data….and takes a lot of time to go through and do it justice.  And most publishing companies just don’t have the resources to open a full data science department with staff to crunch those numbers and compare them with industry norms, editorial trends, and company brand.

But artificial intelligence can do it in minutes.

In 2016, the Associated Press implemented natural language processing (NLP) AI to aid its journalists in the newsroom, beginning with earnings reports. In their associated whitepaper, AP Manager of Corporate Strategy Francesco Marconi offered a great context for implementing AI:

“Artificial intelligence understands the environment it operates in and performs certain actions as a result of it. AI seeks to learn what its users want and how they want it…. AI can help journalists do more investigative work by analyzing massive sets of data and pointing to relationships not easily visible to even the most experienced reporter.”

We’ll dive into more of their findings later on, but here’s the major point: AI can ingest, and analyze massive amounts of data using specific parameters and produce exacting results with direct applications.

It can analyze the text of a book and find the major themes, characters, and settings to use as keywords, which has pretty much been the norm for those doing it by hand. But it can also compare the text to reader reviews, emotional response statistics, social chatter, consumer behavior studies, and more to produce keywords and other metadata using language that the perfect reader for this perfect book uses to talk about themselves and the things they like and dislike. All in minutes, so you are freed up to go continue being awesome, armed with not just great keywords but inherent insight into how your reader is going to discover your book and how it’s going to make them feel.

2. To Increase Efficiency and Quality of Output

AI promises to help publishers maximize their output and create a higher level of consistent quality of that production. Specifically, AI can:

  • Save time for data entry with consistent and reliable data
  • Speeds up decision making by supplying relevant information and analysis from the get-go
  • Ease scaling challenges with its ability to ingest backlist content
  • Enhance comps to go beyond subjective opinions/in house awareness

Today’s market moves faster than before, has more options than before, and more pitfalls than before. As a result, you don’t just have to be the smartest and the most creative in the room, but the most efficient and accurate, too. And to accomplish all of that is no easy task.

But by applying artificial intelligence and machine learning programs to certain areas of your business–categorization, data entry, audience targeting, product positioning, statistical analysis, etc.–publishers and other industries stand to move faster and generate more lucrative products.

Hit the ground running

The sheer quantity of titles and details that editors and marketers are responsible is astounding–it’s not surprising that some titles don’t get the same attention as others do. Consider the discussion of metadata in section one: what if instead of beginning the marketing plans with a blank page and notes from the editor, you began with a customized set of metadata keywords and comps?  Suddenly you go from an against-the-odds rag tag group to a brain trust with an extensive analysis of not just this book, but the whole publishing landscape for comparison and battle-planning.

Enhance gut instinct with objectivity

We all work as hard as we can to read the newest releases as well as the classics…but it’s impossible to know every detail of every book and every market. As a result, most comps are either limited to one’s own backlist titles or to what that person or team is aware of. One of the features of artificial intelligence is its ability to ingest and analyze huge amounts of data–like the entire New York Times bestseller list since 1973, for instance–and produce in minutes an accurate comparison based on whatever data point you tell it to measure for. You know the voice of this book is something special, so you can find comps based on that. Or the setting, or the readability index. You’re in the driver seat–AI is just like your GPS.

With artificial intelligence, teams can start off with metadata and insights specific to each title and use their expertise to edit and collaborate.  After all, it’s harder to start with a blank page than it is to edit a full one.  

This all sounds well and good, but would it actually work for a creative field?  It did for the Associated Press: since implementing AI algorithms to produce automated earnings reports in 2014,  the Associated Press estimates that at up to 20 percent of journalists’ time was freed up as a direct result, “allowing those reporters to engage in more complex and qualitative work,” as stated in their full report on the implementation, “The Future of Augmented Journalism: a Guide for newsrooms in the age of smart machines,”

Through automation, AP is providing customers with 12 times the corporate earnings stories as before (to over 3,700), including for a lot of very small companies that never received much attention,” said Lisa Gibbs, AP’s global business editor.

The AP saw their journalists produce not just more earnings reports, but more work and work of a higher quality across the board thanks to their new-found free time. They saw an increase in their writers engaging with user generated content, multimedia report output, and actually “pursue investigative work and  focus on more complex stories.”

The AP actually used Natural Language Processing, just one facet of their applied AI, to condense the entire report for the Executive Summary. There, they praise AI’s ability to fasttrack the necessary but time-consuming demands of journalism:  

This summary was generated using Agolo NLP (natural language processing) technology. The algorithm took the entire text of this report as input, and organized and summarized its content. We will expand on the inner workings of these algorithms later in the report.

  • AI can enable journalists to analyze data; identify patterns, trends and actionable insights from multiple sources; see things that the naked eye can’t see; turn data and spoken words into text; text into audio and video; understand sentiment; analyze scenes for objects, faces, text or colors —and more.
  • Not only is it imperative to save time and money in an era of shifting economics, but at the same time, you need to find ways to keep pace with the growing scale and scope of the news itself.

3. To optimize each book’s marketing and sales plans

So we’ve talked about the tedium that AI can potentially help you avoid, but what about the exciting, edge of your seat, game-changing creativity it seems to promise?

One of the biggest impacts will be to the bottom line: the more time you save on decision-making and data entry throughout the life of a book, the more resources are available to you in those critical stages before a book goes to market. In the same way that AI services can A/B test website landing pages and run audience tests based on demographics, AI can ingest data to predict behavior and put you one step ahead.

More marketing spend per book

Marketers’ resources are split and their time stretched thin, dominated by efforts for frontlist titles that publishers paid huge advances to– which makes sense. The publisher needs to make their investment back. But it also means that money is left on the table for books that could be huge successes with an extra push.

In addition to the extra time AI can free up, the contextual data that AI tools provide marketers and sales could show opportunities for cross promotion, ideal publication timing, etc. Evaluating successful book adaptations could also provide a road map for positioning a book to be optioned by a movie studio more quickly, or for finding the right studio to approach first.

4.  To increase user engagement (and conversions)

In 2015, Pinterest added an AI recommendation element to its pins — a visual AI, not text-driven–and increased their own user-engagement by 50%. Netflix estimates that its own recommendation algorithms have saved the streaming giant upwards of $1 billion in cancelled subscriptions. Bringing it back to actual publishing, German group Ebner Verlag began using AI to recommend products to specific readers and recorded conversion rates almost 20x the global average.

What did both of these billion dollar companies realize that drove them to invest in their AI?

That giving readers quality recommendations that speak to what they really want would increase their user engagement,  a critical metric for any company’s growth and success.

Find Your Niche

With all of the information at your fingertips, publishers have an opportunity to meaningfully capture direct-to -consumer sales in a way that hasn’t been as achievable before. Today’s readers and shoppers have so many options constantly available to them, so having a genuine product to sell them–and knowing how to sell to them, in particular–stands out.  “These revolutionary technologies are boosting the capability of publications to create and deliver more personalized experiences to and for consumers in increasingly fragmented markets,” notes MediaRadar. “AI can help deliver relevant content to niche markets and individual users with ultimate precision.”

With AI’s demonstrated ability to generate keywords to increase a book’s discoverability, publishers have a view into the kinds of things their readers are interested in engaging in, and can use that to target ads. By running comps on some of those facets, they can expand or narrow that audience even further, or bundle and offer recommendations that will nurture that relationship with the consumer.

Customization to drive growth

“AI can understand and predict the best and most relevant articles for each reader,” explained Boomtrain CEO Nick Edwards to Publishing Executive in a 2016 interview. The media consultant firm drives engagement through AI-powered content marketing/targeting and report  that their customers often see 100% increases in email engagement rates and over 50% in social shares of their content.  

“Boomtrain’s AI is designed to learn how to best engage each individual reader with the exact articles — sent through the right channel at the right time — that will be most relevant. It’s like hiring a personal editor for each individual reader to curate the perfect reading experience.

“The right AI can augment and extend the capabilities of a human editor to deliver more meaningful experiences for readers.”

By performing sentiment analysis and personality comparisons on a book, a publisher can immediately know what kind of language should be used to adequately convey the book’s emotions and can tailor short-form, high-engagement messages on twitter, snapchat, and other social media platforms.

Chatbots: for more than just support

Some publishers are already testing the wheels out on this particular application in the form of chatbots, which have seriously grown up since the days of AIM SmarterChild.

Built to imitate voice, tone and trained on typical conversation patterns and predicted outcomes, publishers use chatbots to drive user engagement and deliver an intimate experience to their readership.

AuthorBot Examples
Authorbot, from Fastbot.io, powers Pan Macmillan’s chatbot on Facebook Messenger

Book publishers HarperCollins and Pan Macmillan are the most notable to start testing the chatbot waters thus far. Both houses implement bots via Facebook messenger, with Pan Macmillan outsourcing its development to Bam Mobile’s Authorbot, which “offers a way to quickly and easily create an author chatbot that readers can interact with..[and] find out about plot, characters, backstory and access exclusive content through messaging” without adding to author time, commitment, or cost. Authorbot is also available on Slack, Amazon Echo, Google Now, and Telegram.

In a recent Publishing Executive article on chatbots, Tech magazine VentureBeat keenly identified the four major elements that any AI worth its salt needs to guarantee increased user engagement when building their own personalized news delivery bot on Google Home (specifically):

  1. Serendipity – Based on user preferences and past behaviors, bots can recommend stories that are likely to be of interest to readers before they begin looking for that content.
  2. Unlearning – Just as the bot can predict potential future interests, it can also detect and eliminate past areas of interest and not send certain topics based on lack of interest.
  3. Recency – Chatbots allow readers to stay in the know and keep up with latest stories within their topics of interest.
  4. Learning – Perhaps the most powerful function of chatbots is the ability to learn based on both implicit and explicit input and behaviors of users — the stories they read, the amount of time spent with each story, and the sum of all of their interactions including likes and shares.

After all, most technology we develop is for a singular reason: we want to make life better, easier, more enjoyable, more accurate–more something. It’s just human nature to want to better and expand our experience for the short time we have. It might sound over blown, but because of that very instinct, publishers have some really exciting possibilities for improvement and expansions with artificial intelligence.

Filed Under: user engagement, discoverability, marketing, Publishing, sales, Artificial Intelligence for Business, chatbots, quality control

SUBSCRIBE TO EMAIL UPDATES