One of the key players on Polish insurance market decided to upgrade legacy software dedicated for insurance sellers to levereage effectiveness of their internal document flow as well as experience for employees having regular contact with retail clients. The reorganization of internal software pipeline turned out to be a very sensitive process, since the replaced software was a heart of company’s income stream. New designed producet hat to ensure a smooth transition from legacy software so that they daily work would not be affected.
To empathize with insurance agents I participated in field research trips to insurance agencies in three cities in Poland: Skierniewice, Łódź and Warszawa. During agency visits I had a chance to get to know reality of day to day work of the agency as well as conduct IDIs with agents. At the end of each visit I had time to shadow agents during their negotiations and conversations with clients coming to get an insurance.
This was a project where we had a chance to do a lot of competition benchmarking. Since our app would be one of many that each agent would use day to day, we wanted to understand what agents found working well in these apps. And to apply conclusions coming from Jakob’s Law, we decided to maintain from mentioned apps what was already working and give our own solutions to those elements that did not met expectations of agents. Thanks to that, I knew that we will answer to the requirement of keeping the transition smooth and simple.
Vehicle and real estate qoute estimators were the modules I was working on single-handedly. In course of designing them I had to gather requirements from two sides of the process: Warta’s product owners as well as insurance agents, and I also had to have client’s expectations in mind. This required from me to find a compromise between business goals and agents’ goals. Even though most of the goals of each side were synced, there were some conflicting expectations where I had to find a satisfactory solution for both sides.
Before we began prototyping we created few artifacts that let us keep an eye on big picture of the project, so that we could quickly spot processes that should be treated as crucial to be included in app’s MVP.For most crucial processes, such as quote estimations, I created high fidelity prototypes in Axure, based on the design system that we created for Warta in a team of two UX and one UI designer. For other screens we created static mockups most of the time and added detailed explanation of behaviour in project’s documentation.
Each module was designed via iterative process. After gathering business requirements and creating low fidelity prototypes, we were scheduling usability sessions for high priority modules or consulted minor modules with business experts. After each round we were implementing changes resulting from gathered feedback.