Lead segmentation

Stages of guest journey can be defined in terms of prospect, interested, qualified, purchaser. In each stage, unique marketing strategies are applied to attract, engage and delight. Segmentation of guests done for each stage based upon what kind of engagement behavior they have, which kind of marketing strategies they are getting exposed to. Also, factors such as location, demographics are taken into account in order to track the inherent behavior of guests. 

A "meaningful clustering" algorithm can be applied here which will give guest segments as output along with their main characteristics which differ that segments of clusters from other one. 

The cluster effectiveness can be checked by checking how clusters are intact using variance check within a cluster or by calculating Silhouette score.  

Meaningful clustering:

Guests can be segmented using different clustering algorithms like K-means, DBSCAN, Hierarchical clustering algorithms. In k-means algorithm, the output we get in terms of cluster centroids. If the centroid of cluster is assumed as one class then we can get as many classes as clusters. The classification algorithm like decision tree or random forest can be applied to get importance features of each cluster called as a meaningful cluster. If central tendencies of important features are analyzed, we can differentiate behavior of guests in different segments.

The meaningful segments of guests are then utilized to apply appropriate marketing strategies.