Jeff Kirsch

After launching our improved financial profile, we noticed that while collected client data had improved there were still gaps being discovered when clients spoke with their financial planners - instances where they had answered questions accurately but neglected to mention pieces of their financial picture that, having been left out, there was simply no way to systematically recognize. In addition, clients felt somewhat let down when the completion of a lengthy and detailed profile left them with no immediate payoff. At the same time, we had little more than anecdotal information about where the data problems were most severe.

Instead of taking on a big-swing, multi-sprint effort to build a solution that we knew would ultimately be based on a large number of assumptions, I successfully argued for an MVP approach that focused on providing only the things that would allow the client to answer the question "Is this me?" when looking at their financial picture, pushing them back into our profile experience to edit pieces that felt wrong. In doing so, we gathered and lightly processed their raw inputs to provide a new but recognizable view of their situation, providing a low-impact but satisfying experience. We were able to build the streamlined solution very rapidly (and with lots of hooks to test our assumptions and client behavior) - buying us the ability to iterate based on those findings down the road.