Sometime around the year 2000, small business (SMB) owners woke up one day and realized they needed a website. The case for SMBs having their own websites is so clear as to be moot today, but that wasn’t the case in the early days of the World Wide Web. Businesses could be forgiven for failing to see the value of staking a claim on their corner of the web.
Obviously, that has changed in the past two decades. Eighteen years later, with the advent of voice commerce and constant improvement of voice technology and ecosystems, Guestfriend Founder Bo Peabody thinks the same could be about to happen with chatbots.
The conversational ecosystem has evolved to a point where consumers are spending as much time in messaging platforms like Facebook Messenger, Twitter’s direct message folder, and their smartphone’s native messaging app, as they are on the web. That evolution is not about to slow down, Peabody said — and now, virtual voice assistants like Amazon’s Alexa and the Google Assistant are weighing heavily into that equation, too. Therefore, he said, successful brands must meet them on the text and voice platforms they’re already using.
In another 20 years, Peabody expects that voice will be a channel everyone takes for granted, just like the web. Future children will be laughing at their grandparents for reaching for the phone at all. Here’s what he says it will take to get from here to there.
Conversation As The Future
Consumers need to find out information about businesses, said Peabody: address and hours of operation, as well as — for restaurants — menu items, dietary accommodations, wine list, dress code and, of course, the ability to make a reservation. To find this information, today’s consumer would rather send a text than an email, and would rather speak to Alexa or Google, than pick up the phone and speak to an actual, live human.
Enter the chatbot.
Whether text-based or voice-based, a chatbot lives where the customer lives, Peabody explained. If a “chat” button can gain the same level of consumer-facing ubiquity as the “Call,” “Website,” and “Directions” buttons on a Google search, he predicts consumers will quickly adapt to leveraging chat to find out everything they need to know before visiting a business in person.
Chat Buttons In Search Results
Indeed, Peabody said this is something Guestfriend is in conversations with Google to accelerate. The tech company’s aim is to automatically generate a basic chatbot for every business on the internet, populating automated responses with information that is available from the business’s website.
If a business wants to do more with the conversational channel, then it can pay a monthly subscription fee to claim its bot and start making it more robust. But for many time-strapped business owners, the bare minimum of conversational capability will be enough to get them started.
For those businesses, simply choosing to activate their “chat” button on Google search results can start to deliver some of the advantages of conversational commerce, Peabody said. The bot can then take over where a business would normally rely on a human to respond to inquiries. Then, if there’s a question it can’t answer, it can direct customers to contact the business by its preferred method, whether that’s phone, email, or web form.
Peabody says the gating item for chatbots achieving ubiquity will be speech recognition. Once tech companies get that part right, it’ll just be about having the right data available, then using natural language processing and machine learning to answer the question.
But before any of that can happen, he said, the bot must understand the question. Google demonstrated at its recent I/O conference that the tech is there for understanding and responding to simple, well-articulated questions, although compound and complex questions still present a challenge.
Peabody said the role for a partner like Guestfriend is to power the information surfacing on the back end. Google’s strength is initial intent detection — that is, figuring out who or what business with which the user is trying to interact. To complete the experience, he said, secondary intent detection is needed — understanding the language of the industry to respond to specific questions — and must be linked to the initial intent detection so that the option to go deeper is right in front of customers when they need it.
So, a customer could ask Google Home about the dessert menu, whether a restaurant is kid-friendly, or which Pinot Noirs are available. With Google’s speech recognition, the bot knows the customer is asking about wine. With Guestfriend’s back-end support, he said, it can find and surface the list.
Peabody said it made sense to start the chatbot revolution with restaurants because their rich ecosystems make it easy to collect the data necessary to power automated conversations. To create bots beyond the restaurant vertical, more speech recognition development will be needed, as language can be very industry specific. Hotels, spas, salons, and exercise boutiques all have their own vocabulary that a bot must understand so it can respond. Then, even more challenging, it will be time to dig into home services, retail, and automotive — what Peabody refers to as “the Yelp verticals.”
Finally, Peabody said the ability to place orders and make reservations through either text or speech-based chatbots will be a key future development. He said those working in the space recognize that voice tech needs to do that — it’ll just be a matter of getting there, and reaching great enough ubiquity that ordering platforms want to integrate with a chatbot creator like Guestfriend.
Peabody says it may not be self-evident yet that SMBs must enable natural language conversations, both by text and by voice, if they hope to remain competitive. But he believes the chatbot revolution is a matter of when, not if.