Understanding my customer

[tonyridge]

I'm new to SAS. I have a list of thousands of customers and I have about 2 dozen attributes for each. What is the best analysis to run to determine which attributes they most have in common? My goal is to market to more people like them, but I'm unclear which statistical process is best for this question.

[ballardw]

Do you have any outcome associated with these such as purchase of "product X"?

Do you want single attributes or combinations of attributes?

Are the attribute variables always all populated or are some missing?

Here is a crude example of finding the most common combination of car type, drivetrain and engine cylinders using a data set you should have available.

proc freq data=sashelp.cars noprint;
   tables type*drivetrain*cylinders/out=work.freq;
run;

proc sort data=work.freq;
   by descending count;
run;

The sorted data set will have the first records with the most common combination of all characteristics.

Or

proc summary data=sashelp.cars;
   class type drivetrain cylinders ;
   output out=work.summary;
run;

proc sort data=work.summary;
   by  descending _freq_;
run;

Will have the common combinations of one, two or three charateristics

[tonyridge]

Hi,

These are all customers, but not of specific products. All of the records have attributes, but not all attributes are populated. I want to discover the list of attributes and values that the customers have in common. This will hopefully help me find more customers like that.

So, I need to determine of the couple of dozen attributes that I have, which ones  do most of these customers have in common and then what are the most common values  for each.

Thanks!

[Reeza]

Cluster analysis.

[tonyridge]

Thanks Reeza!