Categories: Mobile Apps
Chatbots make it possible for businesses to deliver 24/7 customer support without paying overtime. Driven by AI, they respond to customer service tickets and requests in real-time. They’re able to handle basic (but increasingly complex) tasks, performing a sort of triage that means that only the tougher stuff gets passed up to humans.
It’s good for business, and it’s good for people. Most customer service requests are simple, run-of-the-mill things. Automating them away takes the pressure off your CS team and ensures that customers get a prompt response to their question.
Chatbots are on the rise.
While chatbots can’t yet completely replace human agents, they pretty convincingly pass the Turing test when it comes to basic tasks. Think about the last time you chatted with a customer service rep online. Can you confirm that there was a person on the other side of that screen?
By 2025 the chatbot industry is expected to reach $1.23b. In response, Google recently launched its Chatbase analytics platform, designed to optimize bot responses and improve conversions. The aim is to get your chatbots performing so accurately that they’re a seamless substitute for the real thing.
Your own personal chatbot.
Which leads us to Google’s latest release – the AI-supported “Reply” app. Unlike chatbots designed to speak for a brand, Reply speaks for you. If you’re a Gmail user, you know Gmail’s smart reply options, which draw on message context and your past responses to come up with a series of canned replies.
Reply serves up contextually meaningful responses across messaging apps including Facebook, Slack, WhatsApp, Hangouts and Twitter DMs. You pick the appropriate response, and the app sends it. The app makes light work of responses to simple questions and messages. Want to know where someone is? Just ask, and the app will determine their GPS location.
There’s even an auto-responder feature, where your personal Reply bot digs into your calendar and serves up a response accordingly. This same feature kicks in when you’re driving or biking – or even sleeping. It uses accelerometer data for the former, and presumably clock data for the later. The bot can also read sentiment, which lets it tag certain input as needing an “urgent” response.
Chat is just the beginning.
Reply will also be able to determine commute time and provide ETAs, as well as respond based on what you have scheduled in your calendar. The current Beta build isn’t quite there yet, but these highly anticipated features in the works.
Currently the responses are simple, straightforward and devoid of personal tone or idiosyncrasy. But as the AI gets “smarter”, chances are that we’ll soon start seeing responses indistinguishable from what you’d type…if you could be bothered.