USER TESTING: OnBOARDING + PRODUCT-MARKET FIT
How compelling is the current onboarding?
Koko is a therapy bot targeting late teens that integrates into messaging services such as Facebook and Kik. You can tell it your issues and the bot responds based on real-life training data from human-to-human therapy conversations in the Koko community. We tested the onboarding flow and researched teens' general perceptions of using a therapy bot.
How easy is the onboarding? Where are roadblocks? What is unclear? What copy could be improved? Are there steps we can remove?
How engaging is the onboarding? Does the offering seem compelling to users? When during the onboarding do users seem more engaged, less engaged?
Is the product compelling? What are teens thoughts and perceptions of using a bot for therapy? What reservations do they have? What are their ideas of bots in general? How do they describe bots?
ON BELIEFS AND BEHAVIORS
Anonymity and confidentiality are important
Appreciated the reiteration in onboarding that posting is anonymous
Comfortable sharing feelings with bots, even if presumed it would be weird initially
“Yah I do it with Siri”
One girl initially said sharing feelings would be weird because “machines don’t have emotions”, but then once she did and received a response, she thought it was useful “it’s comforting to have someone to talk to”
Issues to share came readily, were top-of-mind
Value getting an opinion outside of their social circles
“Not going to get what you’d expect. A different view that will make you think.”
Value sharing with bots instead of friends because it doesn’t burden anyone
Weary of unknown apps. Seek to understand what app is about and some social proof for credibility before using
“If Kik is redirecting me to this [Koko], I should know about it [Koko]”
Wanted to look up Koko online to learn about it first. Had issues understanding what it was based on homepage
Uses FB Messenger instead of Kik because “everyone is there and it’s easier to find them”
They possess a flexible notion of what a bot is and what it can do. Had some preconceived notions about what robots are, but generally both users were open-minded to new interactions with bots and tolerant of some irrelevant advice / issues
“Robots shouldn’t have emotions -- or maybe I just think that because of movies. No one ever pictures robots and humans together, even though it might be positive.”
Language used to describe bots is anthropomorphic (this could be indicative of true perceptions or perhaps just easier phrasing)
“It’s comforting to have someone to talk to”
“Regular people” - when referring to non-bots.
- Unclear from homepage what koko is for (currently reads more for third-party tools / investors than end-users)
- Lack of social proof / activity on the homepage
=> Show a feed of latest koko app activity
=> Show that people are on koko (# of posts so far)
Engaged in, and followed, onboarding pretty easily. A few points of confusion
One button said “Nice :)”. User was unclear she had to hit this to move on
=> Make button copy more typical / clear “Got it!”, “Next”
The initial copy of the responding tutorial is “repetitive” (I believe the text was ‘here is an example..’)
=> Remove redundant explanation
When posting, the request for a positive spin can feel disingenuous
=> Add an option to “skip” the positive spin
At one point the user was replying to someone’s post and simultaneously received a reply to her previous post. This confused her.
=> When showing an incoming reply, include what the response is reference to “Someone replied to your post ‘My friends don’t res..”