To decide on a particular clustering model, it is essential to determine what precisely the model is going to predict and the level of accuracy that is necessary with the output in predicting based on the data. Determining how one is going to group the data to find commonalities among the data requires an understanding of what the data is producing and how that data will be used, an important first step in determining a clustering methodology. It must be determined how many clusters one is wanting to determine the similar properties in the data sets. Taking the time to determine the best clustering technique is essential to producing the right predictive analysis.