Data analysis involves representing data in various forms both for exploration and for identifying anomalies (missing, cleaning, handling, ..) to be fixed before using them. But what techniques and libraries are available?

The presentation of the results is another equally relevant point, which if not expressed in the best possible way, could determine the failure of the experiment. So which representations are more effective for which stakeholder ?

The state of progress (or improvement) of the experimentation is another salient point, without which we would have difficulty in defining whether we still have a long way to go or if we can stop. Which types of graphs are more incisive for which concept ?

During the talk we will try to answer all these questions: starting from a series of examples, we will explore needs and possible solutions, passing through the various types of data visualizations, design thinking and the principles of data-driven storytelling.

By the end of this talk, the next time you step into the chaos of customer data, you’ll feel more empowered, because let’s be clear: I can show you techniques and tools, but only the data owner can confirm that that graph has meaning.