The bot is using it’s feedback database to gradually improve the packages it buys. As we get ready to launch Bobcat In A Box 2.0, I just wanted to highlight this short article on how the bot handles feedback on products from subscribers.

We have run a few experiments with using feedback data from subscribers to pick new products. This week this capability will be rolled out to everyone. The exciting part is that the bot is learning from the packages it sent to find better, more interesting ones. Given the volume of packages ordered, the bot might be able to quickly get a quite refines sense of what the most surprising and delightful small products are on the internet.

Be assured that we are not sharing any of this data. Re-targeting can be creepy and that is not what this is about. We are simply improving the bot over time using feedback data.

Questions? Email me at alexander@bobcatinabox.com