Messaging is the new browser and bots are the websites.
– Mike Roberts, Kik Head of Messaging and Bot Experience
Bots have been all the rage last weeks. Whether it has been the NSFW Microsoft bot (not only racist, but also encouraging pot smoking in front of cops) or the ability to build chatbots on top of FB messengers.
Why sudden interest in bots?
They are not really a new phenomenon. Eliza was already created in 60s (test it here) and Siri has also been around for a while (test it in your phone). The main reason for the chatbots to gain importance especially now is because of the changed digital landscape. For majority of users, messenger is their digital starting point. Users don´t want to use messaging over Internet, they want to access their Internet to from their messenger. Therefore ability to help, serve and sell to users within messenger is paramount. Short text message (or emoji) is the default way of communicating, should it be also the way to communicate with the brands?
“Conversational commerce is about delivering convenience, personalization, and decision support while people are on the go, with only partial attention to spare.”
– Chris Messina
We are still having long way for the bot economy and below are the core things to fix before chatbots will evolve from novelty to actual user behavior:
1. The bots need to understand normal talk
“They aren’t taking natural language; they are taking menu names,”
Bruce Wilcox,the author of Rose, the winner of the most recent Loebner annual chatbot competition.
Many of the recent Facebook bots are still quite clunky in terms of discussion. People are more casual when they are thinking that they are conversing with real person. The challenge is for the robot to be casual but at the same time providing the transactional value. Current examples have not been particularly promising as they are either pushing you products in unnatural way or trying to be funny but not providing any value:
2. The bots need to become more predictive and fast
Going back and forth with your bot to order a pizza is tedious process. Getting weather details in an hour is just ridiculous. They need to become way more intuitive to use to really rival Google for getting your fast answers. The novelty factor will wear off quickly. If bots are not able to give you solutions fast, they will not be used.
3. Bots are not a destination but a way to enhance the existing discussion
E.g. instead of going to separate weather bot, you should get the weather details when you are chatting with your friend and need that info. Mark Zuckerberg raved about bots as replacements for apps, but with the current experience, it is actually just easier to go to that weather app and get your answer. Ideal situation would be that your messenger would recognize opportunities for commercial interaction from your discussions, but how to build that experience so that it is not creepy?
We are living in the early days of conversational commerce. Using messenger for repeated purchases (like pizza delivery) seems like a no-brainer, but will people actually start browsing products within messenger and asking help from the chatbot?
That depends on the user experience. If AI behind the chatbot actually would know your taste and it would be effective and enjoyable to chat with, messenger economy could become true game changer. Opportunity and potential demand is there, but building a good recommendation engine alone is difficult not to mention that you have to add enjoyable interaction with a robot on top of that. And the core question is, will people want to interact with bots?
Time will tell.
One thing is for certain. Bots will not kill the web, but they will permanently alter it.