One of the Leading online matrimonial services platform. The strategy team wanted a better understanding of this audience in terms of usage behavior, perceptions and attitudes; specifically, to uncover the existing segments/user groups, profile them and understand their specific needs[BP1] . To this end, they partnered with Juxt Smart Mandate, an analytics and consultancy firm
A large-scale pan India research was conducted amongst potential, current and past matrimonial assistance seekers. A 21,000 sample was collected online[BP1] . Although the study had a larger scope, this case study specifically describes the segmentation process that was followed.
The input data for this process was the Importance ratings for various parameters/ features that they would like on a matrimonial website/app. The features ranged from assistance required, sorting/selection features/ verification etc.
The first step was factor analysis. The objective here was to reduce the number of variables and to identify underlying latent constructs. The inter-correlated attributes (parameters/features) were condensed to a number of composites using data reduction techniques (factor analysis).
The above table illustrates derivation and profiling of the 4 factors. These factors can be interpreted as the fundamental needs that people are looking for. The factors are
- Basic background information
- Trust/ confidentiality
- Ease of communication/connectivity
- Additional information
The next task was to classify the people based on these factors/needs. Cluster Analysis (CA) was carried out sequentially. The first step was hierarchical cluster analysis [BP1] where the optimum number of clusters is arrived at. The next step, non– hierarchical clustering (K means) to derive the clusters.
To get a better understanding of the clusters, they were profiled by the demographic and usage behavior information[BP2] .
The sizes of the three segments were as below
1. Urban Survivors – 26%
2. Trust/ Background profile seekers – 23%
3. Ease of communication/ additional information seekers – 51%
As the next step, the clusters were evaluated using discriminant analysis, the functions derived were found to be significant. This classification algorithm is used to classify future respondents with an accuracy of 90%.
Through this segmentation analysis, Bharat Matrimony has been able to understand their audience better and has been able to cater to specific needs by coming up with customized offerings and services.
