SAS small project

Task 1 - 10 marks) Use retail segmentation data. Build value segmentation with profiling.

Use customers.csv & transactions.csv

Step 1) Calculate 33 & 66 percentile cutoff for total spend & total visits.

Hints - proc univariate data = dataset_name;

var column_name;

output out = output_dataset_name p33 p66 pctlpre=P;

run;

Step 2) Store the percentile cutoffs in macro variables . P33_spend, p66_spend, p33_visit, p66_visit

Step 3) Give Score basis spend & visits

Hint -

If spend > p66_spend then score = 3

If spend < p33_spend then score = 1

else score = 2

Do the same for visits

Step 4) Calculate total score by adding spend & visit scores

Step 5) Create the final segment -

Segment = Champion if score =6

Segment = Losers if score <= 3

Segment = Potential for all other scores

Step 6) Perform the profiling basis numeric variables, by taking the avg of these variables - Total_spend, Total_visit, Instore_spend, Instore_visit, online_spend, Online_visit. Also show Count of each segment.

Step 7) Write a macro program to perform profiling of value segments across the following variables - loyalty, preferred_store_format, lifestyle, gender

The macro should take as input one column name (from the list given above) . It should then generate the cross tab for value segment vs the column name.



Task 2 - 5 Marks) Use the startup company data.

Write a macro program that takes as input 3 values. These values represent cut off for 2015 revenue, 2015 profit & 2015 growth

The program then should the list of company names with 2015 revenue, 2015 profit & 2015 growth more then the respective cutoffs given