Having poor experiences with customer service, we looked online for the experiences of Rogers customers, and we found that it's very common problem. Some pain points include poor scheduling of technicians and the inconvenience of needing someone in the age of majority to be in the home during repair. 33% of people in North America say they’ll consider switching companies after just a single instance of poor service. We explored ways to relive these pains through technological innovation. Perhaps it would be valuable to Rogers as well to be able to save costs and improve customer retention by increasing sentiment of customers. In our flow, we targeted the common problem of slow internet causing by the customer having multiple IoT devices.
What it does
This project uses multiple technologies of an auditory chatbot and AR to assist a customer who is having slow internet in their smart home. The customer, using their Google Home asks Raj, the Rogers Customer Service Chatbot. Raj suggests resetting the router. It wasn't the right solution but from the feedback in tonality and keywords adjusts his responses to alleviate their frustration. Meanwhile Raj reads the connection setup of her home and finds out that there are many IoT devices in their Rogers Smart Home, slowing it down. Raj suggests an upgrade with a new modem for a discounted but profitable amount that will fix the issue. The customer confirms and Raj sends a new modem delivered by Rogers Drone.
When the modem delivers, it has a QR code on it. Once the customer scans with her phone, it opens Rogers AR, an app that uses AR technology. Rogers AR helps the customer where to plug in cables and the best location for her modem in her home for wifi reception. This removes the need to schedule and send a Rogers technician, and for the customer to need to be in the home during that time.
How we built it
We broke the workflow into multiple steps -> Problem, Pain Points, Ideas, Storyboard and Task flow finishing this by noon on Saturday. We divvied out the tasks to each person. Gopi made the dialog for the chatbot using sticky notes in an arrangement of bot speech then user response. Chris built this dialog in Dialogflow. Together they worked out the bugs to make the mockup VUI a seemless experience. Vanin built a true functioning version of the voice VUI in Python, using TextBlob API to conduct sentiment analysis on each user response. Hina built the Hi-Fidelity wireframes that runs concurrently with the chatbot, providing additional context to the user. Lucy built a mockup of the AR experience after the customer scans the QR code on the modem box and uses it to find where to install the cables and where to place the modem for best reception.
Challenges we ran into
We had the Google Voice Assistant working on Chris' voice with the bugs edited out.