Have an idea?

Visit Sawtooth Software Feedback to share your ideas on how we can improve our products.

importance manipulation - CBC Study

Hi all,

One of my clients has a 'wonderful' request. Based on our model we get an importance of 30% for price and 70% for brand.

Now he wants to simulate how the market will react in case that the price importance will grow from 30% to 50%. I was trying to explain him with all my arguments that something like this is not possible but he still insists.

Has  somebody ever such a request? How you deal with it? There is anyway to do what my client wants?

Thank you very much.

PS: The study had the following characteristics: CBC Study, Alternative Specific Design, 1 common attribute (brand), 1 primary (type of discount), 2 conditional atributes (level of disounts). The price/discount importance was obtained by summing up the importances of the primary and conditional attributes.
asked Apr 17, 2014 by anonymous

1 Answer

+2 votes
Importance calculations are a strange thing.  They are based just on the extremes of utilities for attributes.  Increasing the price importance from 30% to 50% may or may not result in an increase in price sensitivity of the market (as viewed in the simulator) by 5/3.  

Manipulating the price importance is not a new topic.  Pinnell (https://www.sawtoothsoftware.com/downloadPDF.php?file=multstge.pdf) and Williams/Kilroy (https://www.sawtoothsoftware.com/downloadPDF.php?file=priceaca.pdf) both describe approaches for post hoc adjusting the importance of price after the data have been collected.  It involves multiplying the price part worths by a constant (either per individual as in Pinnell or per segments of respondents as in Williams/Kilroy).

So, it is indeed quite possible to open your utility file and multiply the price part worths by a given constant (different per respondent) to increase the importance of certain attributes from 30% to 50%.  

Whether you should do this or not is a good debate...but it is indeed quite possible.
answered Apr 17, 2014 by Bryan Orme Platinum Sawtooth Software, Inc. (175,190 points)