SimStore Customer Shopping Times Simulator
Each store is represented by a Customer Segment Matrix (CSM). Its two dimensions are: Lifestage (11 categories), and Behavioral (10 categories about frequency and volume of shopping). The main assumption here is that customers in the same segment have similar preferences in terms of frequency, day of week and time of day of shopping.Rule StructureThe simulator uses 3 stochastic rules linked to the customer segments:
1 - Shopping Frequency (as x times a month). Shopping frequency is provided in a user defined matrix (file freq.txt) (same as the CSM format), as one frquency value for each segment. Individual customers belonging to the segment can vary (+/- 50%) with a uniform probability.
2 - Shopping day preference as a weekend to weekday ratio. This is given in the CSM format (file wk2all.txt) as one value for each segment, (indicating the weekdays to all ratio eg 20%) It is only a rough guide, a value of 20% indicates that that individual prefers to shop on a weekend with 80% likelihood.
3 - Shopping time of day probability distibution. Time of day distributions use the 5 time bands defined by the marketing department. Again it is given in the CSM format as one value for each segment. For weekdays and weekends different matrices are used (files wkmat.txt and wemat.txt). The shopping time profiles referred in these matrices are kept in files wkday.txt and weday.txt. Each line in these files describes a probability distribution of time of day. Users can add new profiles to these files and refer to their line number (index) from the wkmat.txt and wemat.txt matrices. For example number 1 in the segment Life=5, Beh=5 location will tell simulator that customers in the Life=5, Beh=5 segment use the 2nd time of day profile from the weday.txt file on weekends. The system ensures that each customer is unique, following a probabilistic rule
1 - Shopping Frequency (as x times a month). Shopping frequency is provided in a user defined matrix (file freq.txt) (same as the CSM format), as one frquency value for each segment. Individual customers belonging to the segment can vary (+/- 50%) with a uniform probability.
2 - Shopping day preference as a weekend to weekday ratio. This is given in the CSM format (file wk2all.txt) as one value for each segment, (indicating the weekdays to all ratio eg 20%) It is only a rough guide, a value of 20% indicates that that individual prefers to shop on a weekend with 80% likelihood.
3 - Shopping time of day probability distibution. Time of day distributions use the 5 time bands defined by the marketing department. Again it is given in the CSM format as one value for each segment. For weekdays and weekends different matrices are used (files wkmat.txt and wemat.txt). The shopping time profiles referred in these matrices are kept in files wkday.txt and weday.txt. Each line in these files describes a probability distribution of time of day. Users can add new profiles to these files and refer to their line number (index) from the wkmat.txt and wemat.txt matrices. For example number 1 in the segment Life=5, Beh=5 location will tell simulator that customers in the Life=5, Beh=5 segment use the 2nd time of day profile from the weday.txt file on weekends. The system ensures that each customer is unique, following a probabilistic rule