Chatbot Design - Vibot
Vibot is a chatbot which enables easier knowledge sharing for community leaders via video streaming, and monetizes communities with conversational video.
In 2017, we launched the first Vibot in cooperation with an online guitar tutor who is also a well-known Youtuber in Taiwan.
Goal and design challenges
Vibot's main goal is to monetizes communities with conversational video. The video transaction is the key metric to evaluate business, and the videos are produced by our customer, Mashushu.
After launching Vibot, we have been tracking users' behavior to measure our goal metric, however, the transaction of the videos hasn't meet our expectations in the past 30 days and the data indicates the following problems :
1. Only 0.6% of users purchase the product
2. 1.13% of users return to vibot within 30 days
3. 50% of users leave the chatbot after they see the video list
50% users drop the chatbot after the product(video) list is displayed
0.6% users purchased the product
Design hypotheses
In order to increase the number of transaction, our customer decided to release new videos (guitar tutorial videos) and lower the prices to attract his fans (target audience). Our team also came up with the design hypotheses below to correspond to the customer's strategy.
​
1. Sending push notification to the users who have already interacted with Vibot (Hypothesis 1 and 2 are correlation):
Hypothesis 1 If users read "Push Notification", then they will see the video. As a result, the "View Rate" of video list will increase.
Hypothesis 2 If the "View Rate" of video list increases, then the "Transaction Rate"of the video will increase.
​
​
2. Refining and simplifing (reducing greeting dialogue) onboarding flow:
Hypothesis 3 If users see the "Filters", they could tap the selection that they are interested in, and then the "Abandon Rate" decreases.
​
Push notification flow
Onboarding flow
Evaluation and summary
After sending the notification, the result shows that the hypotheses are slightly effective.
1. The transaction of new videos increase 0.92% within 14 days.
2. The trailers of new videos are viewed more than previous videos.
​
In summary, we are faced with the prospect of community leaders' influence capacity in social media, and domain knowledge is ease of sharing broadly and quickly through the networking. Therefore, we utilize Vibot to help community leaders to enable their knowledge sharing service which can efficiently respond their client's needs.
Shopping through Facebook and Messenger has been seen as an opportunity for businesses, and we try to verify the "Product Market Fit" at the current stage. Obviously, there are still some questions waiting to be answered, the most challenging one will eventually be "How can we monetize the community with conversational video."
Previous project
