The Process

As with any Sherlock Holmes story, the case starts with a client ...

Problem / Challenge

In order to find something we need a problem definition or a challenge in which area to look. Whether it is a product owner or just somebody else, it all starts with the question from a person: what do we want to find out?


What data is available, how do we get access to it? Usually, somebody has an old copy of the data or there exists a very slow interface to it. Data should be current and access to it fast.


In order to use data from different sources we have to find a common format for it. Often called ETL for extract, transform, load, this step ensures that data is not empty and that data points contain what we expect them to.


In my favorite part, numbers are now presented visually. Whether it may be in classic way as line or pie charts or if one uses a more abstract approach.


Looking for data always costs time and money, therefore every data discovery process should produce results that are valuable for the person who started it.



Great insight from your data comes from a great process of data discovery. Methods from agile programming help in ensuring that the outcome corresponds to the given problem. This site aims at giving hints and tricks on how to find not the murderer, but good answers to your question.

In the mood for a visualization of small data? Try my Danish vocabulary trainer