Let us first define the expected chatbot functionality – or a bit ominously, “let us now gather the requirements and create a functional spec”.
We will call our chatbot GiftMinion. 🙂
It is a gift recommendation chatbot which will help the users find gifts after they tap a few buttons. For now, we will assume that the user only interacts with the chatbot by clicking and tapping, and not by actually typing messages.
That might seem a little strange/ironic/funny given we are building a “chat” bot. But under the hood, we are just converting user clicks and taps into words for the chatbox.
In other words, the interaction is still primarily text driven.
The rich controls help simplify the process.
So the user can select one of four choices.
Browse bestsellers by Amazon launchpad category
Here, we use the Amazon LaunchPad most gifted items as an inspiration. The user can choose from 5 categories, and once the user makes a category choice, we will show them the top selling product.
The data is not dynamic. I use this data primarily for instructional purposes. You can be pretty damn sure that this chatbot is not going to turn me into the next tech billionaire. 😉
Browse by category
This is similar to the first group, except the categories are based on traits such as food lover, music lover, traveler etc. For now, this chatbot only recommends gifts for men under this category.
However, once the user selects a category, we show them a list of choices based on editorial recommendations I gleaned from across the web.
Browse by price range
This is fairly straight-forward. There are three budget ranges and user first selects a budget range and we show them a list of products.
Again, these products are editorial recommendations from across the web.
Here, the basic idea is to ask the user a few questions about the recipient and provide personalized gift recommendations.