Customer satisfaction model (5 Service quality dimensions)

[Studentsss]

Hi, currently i am doing for the customer satisfaction model regarding to the after-sales service of home appliances. The graph generated looks weird and i didn't know what's wrong with my graph. The customer satisfaction graph supposed to be a fratuated and then stable graph but i get a weird graph. I have no idea what problem for my model, can anyone please help me to solve this problem? Thank you in advance!!

[tomfid]

This looks like a discrete oscillation from having processes in the model that are faster than TIME STEP. A quick fix might simply be to make TIME STEP smaller. Also, check your units, if you haven't already. That may reveal a conceptual issue.

[Studentsss]

This looks like a discrete oscillation from having processes in the model that are faster than TIME STEP. A quick fix might simply be to make TIME STEP smaller. Also, check your units, if you haven't already. That may reveal a conceptual issue.

I have tried to check the units and change a little on the units, but still does not change anything as well as the TIME STEP. I changed to 0.5 still the same oscillating graph. I have no idea what to do right now.

[Administrator]

The oscillation comes from the "Service Quality" calculation.

Fix the unit errors first in the model, there is little point trying to fix behaviour if the model does not pass the units test.

Next look at things like "Responsiveness". Move all the numbers in the equation for this out into separate constants and give them units.

Responsiveness =
IF THEN ELSE(((0.2*Delivery Team+0.16*Installation Team+0.24*Product Quality))*SQ Units>0.1936, 0.1936 , (
(0.2*Delivery Team+0.16*Installation Team+0.24*Product Quality))*SQ Units)

For example, what are these numbers supposed to be? "0.2","0.16", "0.24", "0.1936".

What is "SQ units"? If it's there to make the units test work, delete it as it does not make the model dimensionally correct.

[Studentsss]

What if I changed my model to this? But still the customer satisfaction looks weird as it achieved stable state in the next month which does not make sense. And the service quality dimensions are all straight line. I really do not know which part is wrong for my model. The output of the graph supposed to be fluctuated, increase then stable for the "Customer Satisfaction", all the service quality dimensions as well as the Sales.

[tomfid]

There are two nonstandard uses of stocks here: the "previous" inputs to customer satisfaction and number of customers. I think you should revert to standard notation, where the only things affecting the integrations in stocks are flows (plus the initial state).

For customer satisfaction, you probably want something like adaptive expectations:

satisfaction change rate = (indicated satisfaction - satisfaction)/satisfaction response time
satisfaction = INTEG( satisfaction change rate, initial satisfaction )

[tomfid]

For customers, your INTEG expression is,

IF THEN ELSE((Number of Customers-Previous Customer)>=0, Customer Retention-Customer Turnover-Previous Customer*Customer Units,0)

since previous customer = number of customers, this simplifies to just

Customer Retention-Customer Turnover-Previous Customer*Customer Units

"Customer units" seems like a hack for dimensional consistency rather than a parameter with a real-world meaning. Also, "retention" is not an inflow - conceptually it is the residual of the customer stock after turnover. Normally customers should be formulated as:

Customers = INTEG( new customer acquisition - customer turnover, initial customers )