How many times a week do you get an email saying “can we set up a time to chat?” Or a derivative of that message like “How does tomorrow look for a call?” or “Can we talk later this week?” Think about how much time you could save if you had a personal assistant to quickly answer these emails with a response like, “Sure, what day looks best for you?” Or, “Sounds great, what time are you thinking?”
Google’s recent artificial intelligence (AI) add to its Gmail service, Smart Reply, does just that. Smart Reply finds underlying patterns in email communication to “understand” the everyday language of an individual’s email correspondence. It is then able to predict what phrases the user is most likely to type when responding to certain types of messages. It provides several options, or templated responses, for the user to choose from and send—keeping them from having to compose their own response from scratch.
Smart Reply might be the closest many of us will ever come to having a personal assistant, but it’s also confirmation that machine learning is becoming a very real application people can identify with and use daily. In fact, Gartner recently identified Advanced Machine Learning as one of the Top 10 Strategic Technology Trends for 2016.
Machine learning algorithms learn from all types of data points – marketing engagement, customer interaction, employee responses – to help guide actions and decisions based on real-time input. From the moment a prospect engages with a brand, as they use the brand products and services, and throughout the customer lifecycle, machine learning “learns” from this historical data to help businesses be more efficient and precise.
One of the areas we’ve seen tremendous success with when applying machine learning is customer support. When customer support departments apply machine learning to data like questions, complaints and responses, they are better able to answer tickets quickly and efficiently. Support departments are using the same approach as Google to understand context in tickets (similar to how Google is applying machine learning to email) to automatically suggest the best response template for agents use. This not only saves time, but improves consistency and ultimately customer satisfaction.
Have you experimented with Smart Reply yet? We’d love to hear what you think. You can read the Wired article on the Gmail AI app here.