A case study on conversational UX/UI and building out a chatbot prototype
Chatbots are notoriously annoying but when designed effectively, they can serve as an excellent resource to users.
I decided to build out a chatbot for an earlier UX design project which I thought would provide a good application. The project was a business pivot design and platform for customers to connect with local businesses delivering activity kits paired with online classes.
An augmentation based deep neural network approach to learn human driving behavior
Small U-Net for vehicle detection
How to Quickly Improve your Chatbot’s Retention & Engagement
I discovered through discussions with parents juggling full-time jobs while social distancing at home was that that in the absence of outside help, they need a way to educate and entertain their children other than mind-numbing videos on youtube. Before diving into a chatbot service to build out my prototype, I conducted some research to identify use cases and created a persona to inform the conversation flow.
I initially sketched out how I thought the chatbot could serve as a resource to visitors to the website through providing class suggestions or directing them to local business pages.
This spurred me to think more about other ways that a chatbot might add value to customers visiting the Kits & Classes website without being a nuisance. I thought particularly Nicole’s user scenario and developed four main use cases for the chatbot:
- Suggesting class offerings
- Tracking a package
- Discovering local businesses
- Donating to a local business
I added a fifth use case of providing humor and entertainment to customers since I discovered from my research that many people are craving social interaction during quarantine. Although a robot is not a replacement for a human connection, it can provide some comfort by anticipating the customer’s needs and sharing a few cheesy jokes.
One thing that people find frustrating about chatbots is that they often claim to be a real person behind that rectangular pop-up. Customers generally see right through that so I decided to embrace the “bot” in my chatbot and created a corny but endearing little robot with a pinch of spunk called “KC”.
Once I had a basic conversation flow and chatbot personality down, it was time to find the right program to implement it. There are tons of free chatbot programs on the market that have similar offerings so you just have to choose what works for you. I started out working in Chatfuel since that was the program our instructor used to demonstrate how chatbots work but decided to switch to Botsify after trying it out because it was easier to use and had a more intuitive layout.
I started with the greeting message and then built out the main stories for each of my user cases and scenarios. I incorporated gifs and images to liven up the dialogue and featured products in carousels with quick links to add items directly to your cart from the chatbot messenger.
