Have an idea?

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

How is Purchase Likelihood calculated in market simulator?

Dear Everyone,

How is purchase likelihood calculated in Market simulator?
Is the data captured in "Dual Option - None" used for this or something else?

I am using "CBC-HB alternative specific design" for my research.

Kind Regards,
asked Apr 11, 2018 by sacharya (470 points)

1 Answer

0 votes
The "purchase likelihood" simulation approach is an old relic from conjoint analysis in the 1980s (full-profile ratings-based conjoint analysis, or also ACA, where the data were scaled properly for the equation below to return a least-squares fit to the respondents' stated likelihood of purchasing concepts on a 100-point scale).  

The math does not consider the None alternative.  It simply is:

p = e^U/(1+e^U)*100

where p is the purchase likelihood,
e is Euler's constant
U is the total utility of the product concept
(e^U may be computed in Excel as =EXP(U) )

If you want to compute the likelihood of respondents selecting a specific product vs. the None, then you should use "Share of Preference" simulation method, and specify just the product and the None alternative to capture share of choices (share of votes) across respondents.
answered Apr 11, 2018 by Bryan Orme Platinum Sawtooth Software, Inc. (198,315 points)
Dear Bryan,
Thank you so much for such a fast response to my question and providing further directions. Really appreciate it.