3 Elements That Make Bots Work In Facebook Messenger

The brave new world of bots is marching onward with each passing day to the steady drumbeat of growth in messaging apps worldwide. People love to text and messaging has emerged as the preferred communication mode for many. With the growth in Facebook Messenger and its platform, talking to companies and brands through messaging is slowly starting to gain ground. As with any breakout thing, there are opinions on both sides of the spectrum of whether it is hype or whether it is something tangible that will stick around. Whatever the case may be (I’ll leave my crystal ball post for another time), we at CartSkill have observed this development through other bots and more directly through the bots we have built ourselves on Messenger over the last months. I jotted down some thoughts and interesting findings that we have uncovered.

  1. People love to chat to ”things” that chat back. There is not much hesitance to talk to a bot. Rather it is first and foremost driven by the users interest to get their problem solved (e.g. purchase a product or find information). And there is also just the share curiosity to see what happens when you talk to a bot. We’ve seen people say ”thank you” to our bots countless times. Would you say thank you to an automatic parking garage door that opens for you? Probably not because it is a machine. A bot is also a machine. But you chat with it, it chats back, gives you the information that you need and you are grateful for that. And when you say ”thank you”, the bot says ”you’re welcome”. How many times do people say ”you’re welcome” when you say “thank you”? Not all the time but a bot says it every time. And this makes you feel good.
  1. Yes, artificial intelligence is still limited but in many cases human responses and queries are not all that unpredictable. Unless you are building a very general search engine for everything (and going to challenge Google), a bot is most likely to offer value in a specific use case. And the way that people approach this use case is not that different. Say when a person is searching for a train’s departure time. There are only so many things they can enter – a time, a city, a date, arrival city, train company name. Yes, there are different ways to formulate the questions but it is possible to pick up the intent and offer meaningful results. Sure, there are off topic comments and rants but again, these can be focused within boundaries.
  1. Chat is a focused and a quick medium. Attention is the online currency and a person will dedicate it to bots in a frugal way unless they have time to kill and are looking to Turing test a bot (this has happened to us several times and deserves another post on its own J So the conversation exchange is rapid-fire back-and-forth on a certain topic. Once a bot passes that first trial, the person becomes more open and is more likely to return. So Shakespearean bots need not emerge but rather quick witted sidekicks that can jump in with relevant answers at the right time.

I’ll be speaking more on this topic at the international Refresh conference on product design on September 9 in Tallinn.

Subscribe to our newsletter for the latest in AI