Glow is a new way to buy workers comp insurance. It's workers compensation insurance that doesn't end when you leave work. For a company doing things a new way in an old industry, the experience needs to be right.
To complete our onboarding process - the user is presented with a chatbot to wrap up the complex process of purchasing workers compensation insurance.
Product launches are important. Glow was looking to create a first-in-class onboarding and first-time user experience.
The challenges included the complexity of the Workers Comp insurance industry, the many legal hurdles and industry archetypes to address and understand. I worked closely with the CEO and senior industry analysts to create a first-time user experience and onboarding flow that is engaging, unique, speaks to the brand and most importantly is effective.
Create first-time user experience and onboarding flow that is engaging, unique, speaks to the brand and most importantly is effective.
As part of initial research for Glow, the CEO had many interviews with restaurant owners, as this was seen as a focus industry for Glow. I was able to use these recordings to understand who the target users were and get a deep insight into their thoughts on workers comp and how it relates to their business.
User Research - Restaurant Owner Persona - We need to understand who purchases workers comp insurance and how they make their decisions.
Week by week - I created a prototype, recruiting business owners as test users. Recorded them using the app and analyzed the results of my participants. I would then present findings to the rest of the team on Friday. I would then make changes to the prototype and start the process over again. When all our questions were answered, we knew we had the correct implementation to start.
Product prototype - user onboarding experience.
The solution to our design challenges was to end our onboarding flow with a chatbot concierge experience that tied up all the loose ends and was flexible enough to account for any path the customer had taken to this point.
The chatbot was not an actual AI, rather a question and answer tree with specific responses to guide the user to their desired outcome.
Product Screenshot - The system analyzes your companies risk profile.
Product Screenshot - The user is interacting with the chatbot to schedule a start date.
As we iterated through our sprint cycles, we saw problems arise through user testing, then be addressed and eliminated - making a cleaner user experience. It was a qualitatively better experience each iteration. We asked a series of questions to score the experience and we were able to also see quantifiably better results in reduced time to complete and overall satisfaction of the test subjects,