We live in the age of mobile applications. There are currently several million apps available. This profusion of choices means it can be difficult for users not only to choose which apps to download, but to manage them all — a phenomenon we call “app fatigue.” This situation creates both a need and an opportunity to engage users on a single platform. Today, that platform is increasingly becoming messaging apps.
We think the next era will belong to “the conversational layer” — both text- and voice-driven — that will use chat, messaging, or natural language interfaces to interact with people, brands, services, and bots. This shift is currently evidenced by the massive adoption of messaging apps such as Facebook Messenger, Echo, and WhatsApp, which together host more than 60 billion messages daily. According to eMarketer, messaging apps will reach 2 billion people within a few years. WhatsApp users average nearly 200 minutes each week using the service, and many teenagers now spend more time on smartphones sending instant messages than perusing social networks.
Messaging platforms can also alter the way businesses can communicate with their customers. Currently, conversational interfaces within well-known messaging platforms such as Facebook Messenger, Slack, Skype, WeChat, Kik, and Telegram allow companies to chat with their users.
While mobile chat platforms are interesting, the arrival of artificial intelligence-powered engines called bots have made them a powerful tool for sense-making and commerce. Bots use machine-learning techniques to understand text and provide better responses to user queries. They are present in the background, and they make sense of the conversations taking place and convert them into actions using apps, such as scheduling a meeting or ordering a pizza. For example, imagine you are chatting with your business partner using Messenger and discussing a visit to a client site in Boston. Using machine-learning algorithms, a bot can recognize that you are talking about travel and initiate a transaction with your favorite travel app, such as Expedia, or offer a link for a ride through Uber. The messaging platform effectively becomes a distribution channel for software and services without leaving the conversation.
1-800-Flowers recently launched such an experience on Facebook Messenger and has since expanded to Amazon Alexa and the IBM Watson platforms. Customers can order flowers directly from their experience in these conversational layers. 1-800-Flowers is very focused on customer support and maintaining a relationship with their customers, so the company jumped at the opportunity to be one of the first in the space. Of the tens of thousands of people who have ordered flowers through the chatbot integration, more than 70% are new customers — and these new customers skew toward younger demographics than the company’s existing customers.
Companies should position themselves for the conversational layer to be more widespread five to 10 years from now. Individual users will most likely want to interact with trusted brands to fulfill their needs through natural language interactions. This interaction will occur at the exact time the user demands a product or service, and in the exact terms she thinks of that product or service, in the language and communication methods she typically uses (intent, words, shortcuts, emojis, etc.). Companies need to strengthen these natural language capabilities in their products, apps, and bots to allow users to communicate with them with ease.
Individuals will begin to welcome and even expect this type of service from brands, but companies must remember that this trusted personal space is precious. Poorly designed interactions can irreparably damage the customer relationship. For example, when Microsoft’s Tay posted racist remarks on Twitter, it had to be shut down temporarily. The bot industry as a whole has yet to come up with the “killer bot” that tips the scale for wider adoption, but bots continue to grow in sophistication and power.
Pick a platform. When reaching users, a brand needs to understand where current or potential customers spend their chat time. The platform choice is an important early decision. This is similar to how brands and engineering teams initially opted to launch their products on iOS, then Android, and other mobile OS platforms early on. In order to reach user conversations today, brands will need to decide which platforms to target and build on. Different platforms have a diverse set of capabilities (i.e., user identity, cards, and buttons on Messenger, and work team and slash commands on Slack) and demographics target.
Run strategic experiments. It is not clear if customers would use the conversational layer for quick responses or for broader conversations. Brands like Amex Finance are using chatbots to provide notifications to customers about their new products or alerts about forthcoming travel dates. These limited experiments would allow Amex to set the bar for the nature of interaction with clients. The enterprise social networking platform Slack is using chatbots to automate routine managerial check-ins, reducing the need for meetings.
Look for innovative uses in other sectors. Simple examples like ordering airline tickets or pizzas are emerging. However, sophisticated bots that understand the context and make intelligent decisions have not yet been developed. In fact, recent articles on automated email assistants have shown that they rely too much on human intervention.
Companies can also learn from examples outside their industry. For example, a Georgia Tech professor used a bot as a teaching assistant for a programming class. Using machine-learning techniques, the bot was able to handle 97% of the student queries. In online education, where dropout rates are quite high, a high level of engagement with a “tutor” could make a huge difference in retention.
Pilot bots with your customers. Many of the tools that are provided by the messaging and bot platform providers are from the open-source space, and companies can perform low-cost experiments with a reduced set of users to learn more about conversational interactions and use cases that yield the desired results.
The conversational layer of computing may have not yet fully arrived, but it is coming. Companies should begin thinking and experimenting now about how to use these new avenues to support their brand and their business today, so they can be ready for that conversational future as users demand and engage with this type of experience.