How to Do Data Science in your Company to Get The Most Out of it. Part II.

That is why we have started started to cut through projects from end-to-end right from the beginning with an initial version, to see where the pitfalls lie.From the perspective of the complete Novel Data Solution project portfolio we began to collect and incubate a wide range of topics at an early stage, in discussion with all key stakeholders from product management, development, marketing, customer relations and management..This manifests itself in a list of big questions that we revise on a regular basis..This list helps us to develop initial thoughts and hypotheses on a subject first, include the perspectives of experts and the market momentum, until we decide on the next projects to focus on..We have learned to “grow” Novel Data Solutions step by step rather than planning to hard..But we always coordinate carefully with the running projects, to see where we stand and to be able to control the efforts and risks..We thereby use a kind of sowing, breeding, rejecting or harvesting approach..And this is important to share: do not fix long term roadmaps!.Because of the very nature of this kind of projects, do only plan the next step and revise the status quo regularly (at least quarterly) and only afterwards plan the next projects..However, we diversify risks over a clear portfolio of selected moonshots that we run concurrently, as in some cases the outcome can be nothing, but in other cases the outcome can be everything.It is very important to communicate well with all stakeholders throughout the whole process..In the beginning, the main goal is to bring business knowledge and context into the project..Later, we want to make risks transparent, find solutions to potential obstacles, and sharpen goals..Towards the end of the prototype development, possible results must be incorporated or tested as part of the company’s products, and concepts and data models must be put to production.. More details

Leave a Reply