We have seen that gallup data analysis requires both dimension reduction and clustering. As it is the case with traditional solutions, like factor analysis, it is easy to loose some minor but significant information when trying to find the clusters directly, using one prototype or factor per cluster. The elegancy of the Self-Organizing Map is that it provides several sub-clusters in topological order, combining the advantages of clustering and profiling methods. As shown by the example, results are easy to achieve and interpret.