Data mining and knowledge discovery denote a set of methods and algorithms for exploration and analysis of (often) large volumes of data aiming to infer rules, associations, unknown trends, specific structures so that useful information may be returned in a concise form for supporting decisions. Despite its fast-paced development the data mining is still vaguely defined and lacks an integrated approach. This situation causes difficulties in teaching, learning, research and application. The success of a project in any field of activity of contemporary organizations is often compromised by the general propensity to develop solutions before identifying problems and formulating statements. The article focuses on several important aspects such as the nature and quality of data used in the application of data mining, the most commonly used methods, the choice of the main objectives, problem formulation that should be adequately addressed in the context of data mining common strategies.