Commercial Claims Research
Project Overview:
Conduct generative research to create a strategic path forward for the commercial insurance claims user experience projects roadmap.
Project Aspects:
Full strength multi-month research project including UX researchers, interaction designers, and content specialists
Research methods included, but were not limited to:
User surveys
User interviews
Card sorting
Interactive prototype building
“Jobs to be done” analysis framework
Semantic network analysis
The Ask:
Commercial insurance claims had been completely neglected in previous user experience research efforts
There was the assumption that commercial clients would want the same service or types of service as individual clients
The project was to get a better understanding of user needs
The Users:
Commercial insurance customers (B2C)
Commercial independent insurance agents (B2B)
Key Questions:
How do customers perceive commercial insurance and claims differently than they do individual insurance policies and claims
Do commercial clients want to use self-service tools to file and follow claims like individual clients seem to?
Who owns what part of the claims process?
Where can the company make improvements?
Are there “low-hanging fruit”projects?
Methodology:
Interviews
10 - 12 Commercial Agents
10 - 12 Commercial Clients who had made a claim in the past year
Create a cluster of statements from the interviews
Conduct a “Jobs to be done” analysis of the clusters
Back up “Jobs to be done” qualitative research with Network analysis of semantic interview network
Network Analysis:
Use an online engine, “Ifranodus,” to parse the semantic connections between all the words used the transcripts of the interviews
Then looked at the network analyses of the semantic network
Determined keyword “nodes” and looked for relationships
Network Analysis Findings:
Correlated about 60% of the big qualitative insights from the interviews
Correlated about 50% of the secondary qualitative insights from the interviews
Has some issues, but would definitely use it again
Network Analysis Issues:
Huge correlation between the words used in the actual questions
No simple way to deal with disambiguation
No simple way to enforce similar lexicon for all interviews (Client could be “client,” “insured,” “Bob,” “owner,” etc…
Correlation is not causation
Tons of extraneous nodes (words like “swan”)
Potential Network Analysis Solutions:
Couldn’t solve for all the issues
For one interview, enforced limited vocabulary
Removed all nodes with low importance score (nodes with very few connections)
Removed nodes with too many connections (words like Insurance or process since the whole interview is about them)
Research Findings:
Commercial Agents don’t want to get involved in the process
Commercial clients want to be informed of what is going on for their claim at all times because it may be their livelihood on the line
Some commercial clients want to do everything themselves
Some commercial clients don’t want to have to think about it and want their agents to take care of everything
Customer trust boils down to radical transparency
Commercial Claims Next Steps:
Implementing a “Radical Transparency” strategy
Letting customers know where they are in the intuitively
Giving customers the ability to easily ask, “What’s next?” if it is not readily apparent
Deeper dives into active agent vs active policy holder