Most Data Science product development processes focuses on the data, algorithms and the definition of the right problem statements. I have witnessed a ton of solutions that fail to get the adoption they deserve because the solution is not integrated into a user workflow or because some assumptions were made that don't make sense in a real world. This is why I advocate using design thinking to not just write a problem statement, but also to get a verified product concept, wireframe and roadmap before writing a single line of code.