Also, they made a move into messaging with iMessage. Instead of a bot API like Messenger or Kik, or HTML mini web apps like WeChat, they are releasing an SDK to create mini apps that live on conversations on iOS devices, together with a new App Store. I didn't see that coming, really.
On the Google front, I believe it's just a matter of time for Google to release an SDK for Google Now.
The author focuses on economic, cultural and ecosystem factors behind the bot craze we're seeing (as opposed to the technical drivers, which have been widely discussed already). Pervasive communications, consumerization of software expectations, the pervasiveness of mobile in China are just a few of those discussed in this thorough post.
A couple of quotes from David Marcus, VP in charge of Messenger: “And it was the exact same thing, overhyped in the short term, underhyped over the long run”, “The first websites were really, really bad, the first apps were really, really bad.”
Kik's CEO shares some interesting metrics, and provides a classification of chat sessions in three categories: active, passive and sporadic. They call it the attention metric, and it's based on the time taken between messages during a conversation.
The business models that we are starting to see are: Bots as a service (SaaS style), sponsored bots (an evolution from native advertising), bots that leverage affiliate marketing and retail sales bots. The author also suggests other models: bots for research (probably connected to humans doing manual research on the other end) and lead generation bots. In this list, I miss customer support automation bots, which is a field I'm very invested in.
This post lists 11 recommendations for bot design, together with the rationale for them. To me, the most important rule is having the bot set the expectations clear while introducing itself.
MB16 is launching the first ever national Botathon, that will take place the week before the event. I'll participate as a judge for the NYC batch on July 9th and 10th.