When clustering, the algorithm is going to compute a multidimensional Euclidean distance, so the scale of numerical features is very important. For instance, think of a dataset with house prices and number of rooms among other fields. Since the values of house prices are larger than the number of rooms, you will end up having the house prices dominating the distance measurement in that cluster. Therefore, you need to scale those fields.
You can easily do it through the clusters configuration panel, either adding manually the scale you want to apply to the each field or enabling Auto-scale fields option (balance_fields in the API), as shown in the image below. When you have this option enabled, all the numeric fields will be scaled so that their standard deviations are one. This makes each field have roughly equivalent influence.