Review the material on the How to Interpret tab.
tell me: 1. which output table you used for your decision, e.g., one, two, or three clusters/segment. 2. identify the target customers by age and gender.
Here is the cluster analysis that these questions are based off
Output for FOUR Clusters/Segments \begin{tabular}{|l|} In which cluster/segment is the respondent allocated? \end{tabular} \begin{tabular}{|c|c|c|c|c|c|} \hline No. in Segment 1 & 99 & 75 & 76 & 29 & 35 \\ No. in Segment 2 & & 24 & 3 & 19 & 7 \\ No. in Segment 3 & & & 20 & 18 & 20 \\ \hline No. in Segment 4 & & & & 33 & 34 \\ No. in Segment 5 & & & & & 3 \\ \hline TOTAL & 99 & 99 & 99 & 99 & 99 \\ \hline \end{tabular} 1 You need to select ONE of the cluster options only - that is, two, three, four or five segments/clusters. You generally should NOT use more than one approach. Do not attempt to analyse the segments across the outputs - just choose the best one. 3 Which one is best to use? A. Look for a good distribution of respondents across the segments - you don't want 90% of all people in one segment. NOTE: This is example data only. loping a strategy. 4 If the segmentation output (means/averages of each variable) don't make sense, then try adding/removing/changing your input variables - try a different mix of variables. NOTE: Trial and error is a normal part of analysis - plus it will help you understand the data better. Output for FOUR Clusters/Segments \begin{tabular}{|l|} In which cluster/segment is the respondent allocated? \end{tabular} \begin{tabular}{|c|c|c|c|c|c|} \hline No. in Segment 1 & 99 & 75 & 76 & 29 & 35 \\ No. in Segment 2 & & 24 & 3 & 19 & 7 \\ No. in Segment 3 & & & 20 & 18 & 20 \\ \hline No. in Segment 4 & & & & 33 & 34 \\ No. in Segment 5 & & & & & 3 \\ \hline TOTAL & 99 & 99 & 99 & 99 & 99 \\ \hline \end{tabular} 1 You need to select ONE of the cluster options only - that is, two, three, four or five segments/clusters. You generally should NOT use more than one approach. Do not attempt to analyse the segments across the outputs - just choose the best one. 3 Which one is best to use? A. Look for a good distribution of respondents across the segments - you don't want 90% of all people in one segment. NOTE: This is example data only. loping a strategy. 4 If the segmentation output (means/averages of each variable) don't make sense, then try adding/removing/changing your input variables - try a different mix of variables. NOTE: Trial and error is a normal part of analysis - plus it will help you understand the data better
