In CRM Analytics, when dealing with incomplete data, specifically when certain respondents have not filled out fields like income, the Predict Missing Values transformation in a recipe is highly effective. This transformation allows you to predict values for missing fields based on patterns from the existing data. Since the consultant finds this field informative despite having data from only 1% of respondents, applying this transformation can estimate these missing values, which ensures that the dataset remains useful for predictive purposes without discarding important variables.
Reference: CRM Analytics Recipes and Predict Missing Values