Publishers Are Tapping into the Chatbot Craze to Deliver Personalized Reading Experiences
It’s been years since people seeking information had to travel to the nearest library, access the card catalog, and search the shelves for books or magazines to find the answers to their questions. Today virtually all consumers know that if they have a question -- whatever that question might be -- they can simply “Google it.” And now, as more consumers seek information on their smartphones and on social media platforms, they can simply ask one of the growing number of chatbots on Facebook Messenger, WeChat, or Kik. Using chatbots, consumers can ask questions about the latest news stories, receive customer service assistance, or even manage their online subscriptions. They can also receive automatic updates on stories that might be of interest to them, receiving relevant information before they even think to look for it.
Publishers have found that chatbots are a powerful way to engage with their audiences and monitor that engagement in order to gain valuable insights on reader interests. While chatbots may appear to be simply the digital form of a automated call center, they can offer a more customized and engaging experience to consumers. Using digital insights like geographic location, interest, past behavior, and more, chatbots can help publishers provide the right information at the right time to their readers. In this article, Publishing Executive will explore how innovative publishers are rolling out their own chatbots and virtual assistants and how they are leveraging this technology to increase audience engagement and, ultimately, revenue.
A Paradigm Shift From Apps to Chatbots
Chatbots are reshaping the media market as their popularity among consumers, publishers, and platforms soars. Facebook Messenger now has more than 100,000 bots, says Gracie Page, creative technologist with the agency Y&R London. Facebook is the dominant player now, but Amazon and Google are nipping at its heels, and other platforms are planning chatbot rollouts, says Page. “WhatsApp is rumored to be getting bot functionality by the end of the year, and Apple has just announced their arrival on iOS in the fall,” says Page. Likewise, publishers such as VentureBeat, TechCrunch, and Paste are experimenting with this technology.
The popularity of chatbots is spurring a paradigm shift in how readers consume content, says Beerud Sheth, CEO of Gupshup (which means chit-chat in Urdu and Hindi), a chatbot developer. “Every decade or so the user interface changes,” he says -- from desktops in the mid-1980’s, to the Internet in the 1990’s, to smartphones in the mid-2000’s. “Now we are on the verge of yet another major shift -- this time to bots.”
Bots represent a means of driving content to consumers, versus them needing to seek out content and functionality by downloading apps. “Even though there are millions of apps, most users are not downloading and most use no more than a dozen their mobile phones,” he says. But, he notes, there is one app that is “truly ubiquitous” -- the messaging app.
“The whole industry is saying, ‘If that’s what people like to use, then why not squeeze the whole internet inside the messaging app,’” explains Sheth. “What if you could use the messaging app to talk to a news site or banking service, or gaming service, or anything else you wanted to do -- order a pizza, book a taxi, send me the latest story?” In this scenario, the messaging app becomes the new browser, says Sheth.
Delivering Content When & Where Readers Want It
Travis Bernard is director of audience development at TechCrunch, an early adopter of bots. TechCrunch is well positioned to leverage chatbot technology because the publisher has a ready-made audience of tech-savvy, early adopters. Even so, Bernard says that a key to success is keeping it simple. In fact, consumers don’t even need to know that they’re interacting with a bot -- they’re really just concerned with functionality, what the bot does not what it is. That straightforward approach has driven results for TechCrunch, which boasts over 100,000 chatbot subscribers that use the service five to six times a month.
Initially, the TechCrunch chatbot delivered news, says Bernard. “We were really thinking of it as kind of an alternative to newsletters,” he says. “It works similar to a newsletter: you set up a time when you want to get a digest delivered and it will send you an update and you can subscribe to different topics.” Subscribers can select to receive news on robotics, AI, Facebook, and even Mark Zuckerberg.
TechCrunch’s chatbot has evolved over time to offer greater personalization and more discrete topic options. Most recently, TechCrunch has been exploring the ability to generate breaking news alerts. But having a small editorial staff that has limited bandwidth with which to write these alerts remains a challenge, says Bernard. One day, TechCrunch hopes to deliver news alerts based on the reader’s geographic location. “When we’re doing an event in Austin, we could send users a message and spur greater attendance,” he says.
Enhancing Chatbot Functionality With Artificial Intelligence
Other publishers are using AI and machine learning technology within their chatbots to anticipate what content consumers may be interested in. In June, Paste, a destination for pop-culture and music enthusiasts, announced that it had teamed up with Gupshup to provide customized content to users on Facebook Messenger. Readers can chat with the bot to discover the most recent articles and audio content, and the bot learns about user preferences based on how users interact with it. Paste plans to add omni-channel functionality in late 2017 to make the bot available on platforms like Slack, Telegram, Google Home, and others.
Josh Jackson is president and editor-in-chief of Paste. He says that along with providing more relevant information to readers, the chatbot also allows for greater, and more reliable, interactions. “It’s increased our response rate to 100%,” says Johnson. Social media chatter can be excessive and human monitoring is not 100% reliable, so using a chatbot allows Paste to automate engagement and ensure readers receive the information they’re looking for in a timely manner. “Judging from the messages from our readers, I think they’ve been entertained by the experience and have found cool stuff on our site,” says Jackson.
VentureBeat is another example of a media outlet that has chosen to provide next-level functionality to users. It has built a chatbot on Google’s voice-activated device Google Home and delivers personalized news to readers. Matt McDonald, product manager with VentureBeat says, “A lot of our competitors had already come out with newsbots that were basically a glorified RSS stream that you could connect with your social channel or messaging service of your choice. We didn’t want to just jump into that pool. We really wanted to think strategically about personalized news and what personalization in a newsbot entails.” McDonald has identified four key elements of personalization that are critical for ensuring, and maintaining, audience engagement. His team uses these elements to guide their chatbot development:
- 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.
- 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.
- Recency - Chatbots allow readers to stay in the know and keep up with latest stories within their topics of interest.
- 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.
“A combination of all four of those things allows the VentureBeat newsbot to stand apart from a lot of our competitors,” says McDonald.
The more personalized the experience, the greater the likelihood that users will become engaged in that experience -- importantly, notes Sheth, that engagement can be monetized. “If content is always relevant and interesting, then consumers tend to consume more; ad rates could be higher because you’re doing a much better job targeting to their interests.” Moving from broadcasting to “nanocasting,” delivering very specific topics to a discrete audience, “can drive engagement and monetization much higher,” says Sheth. Personalization can also drive other revenue models within chatbots, such as personalized native content, targeted ecommerce offers, and affiliate marketing.
Making It Happen
The good news for publishers, says Page of Y&R London, is “these are not million-dollar projects to produce.” There are plenty of boutique bot-building agencies in the market and publishers could produce these bots in-house, given the extensive documentation online, says Page. The bigger challenge, she adds, “remains identifying the real commercial opportunity that a bot can help to unlock, and then designing the technology so that it adds value to the customer’s life rather than more digital noise.”
As with any new technology, form should drive function and not the other way around. When exploring the viability of chatbots, publishers should first consider what types of interactions they want to automate in order to improve user experience: accessing relevant information quickly, being automatically informed of new content aligned with their interests, customer service activities, etc. Bots can perform all of these activities -- it’s just a matter of identifying and prioritizing the ones that will be most appealing for your audience. For publishers, the obvious, high-priority functionality is likely the ability to deliver desired content in a timely manner to subscribers.
The next step is making it happen. Bernard recommends working with a vendor. TechCrunch uses Chatfuel, while VentureBeat and Paste have worked with Gupshup. Many vendors specialize in a certain platform, for example Facebook Messenger, Kik, or Telegram, so publishers should have an idea which platform or platforms they want to use. In selecting the appropriate platform, publishers need to consider where their people are most likely to be. Not surprisingly, Facebooks Messenger is one of the more popular platforms.
One good way to narrow down potential vendors is for a media company learn which companies have built the bots that have the functionality it’s looking for. This Botlist provides a wide range of examples across multiple categories, including news. Unless the publisher is a direct competitor, chances are the company would be glad to share which vendor its team worked with.
The cost to create a chatbot will vary widely and be dependent on whether publishers take a do-it-yourself approach using in-house talent, use a self-service platform, or hire a development firm. According to VentureBeat, self-service platforms can range from $19-50 per month, offering various functionality; software development costs can range from $5000-$10,000 per bot for setup, plus monthly maintenance fees in the low thousands.
Publishers wishing to handle chatbot creation internally can take advantage of tools like Chatfuel or Pandorabots. Existing tech talent may well have, or be able to learn, the tools and techniques needed to manage and maintain basic request/respond types of bots. Higher-level bots that incorporate AI and machine-based learning will require higher-level technical skills that, of course, will come at a higher price. Due to the rapid changes taking place in this space, for now, as Bernard suggests it may be more practical, and more cost effective, to outsource this development.
A Trend Publishers Cannot Ignore
Based on the rapid growth in consumer interest and platform development, it’s likely that chatbots are a technology trend that publishers won’t be able to ignore. Bots are providing consumers with greater control over not only the information they consume, but also every aspect of their environments -- from playing music to purchasing products and services. As Sheth argues, chatbots may one day be the way consumers discover new information. That opens an opportunity for publishers to solve new problems for their audiences and more deeply engage with them. In the chatbot era, the publisher-reader relationship will become more immediate and more personal than ever before.
Lin Grensing-Pophal is a business journalist who has written on topics ranging from health and wellness, to relationships, careers, profiles and HR-related topics, to marketing communications and social media. She’s written books, articles, white papers, reports, newsletters, e-letters, brochures, web sites and blogs.