In this chapter, we looked into two use cases of self-service analytics. One is about world indices with a straightforward dataset, and one is about traffic-light violations, for which we had to join two datasets. Our main takeaway was, that you should always start by planning your work. Next, get to know the data, use descriptive statistics, and adjust your plan according to your intermediate results.
A dashboard is never really finished; you can always change things; your audience might change; stakeholder wishes might differ. Deliver a working visualization with basic functionality and continue to develop after you get feedback. Depending on your backlog or other circumstances, the basics might be enough.
The next chapter will be all about improving performance. With more and more data, performance is key, and could make the difference between success and failure...
