Have an idea?

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

Simulator Importance and SOP alignment

Hi Team,

I got one question from client regarding Conjoint Simulator. When we are comparing 2 concepts then SOP and Importance are getting aligned accordingly. If Importance of any attribute's level is high then adding that SOP is increasing and SOP decreasing simultaneously if importance is low. But when i am taking more than 2 concepts i.e. adding 3rd and 4th  then this alignment is getting disturbed and low importance attribute's level is having higher SOP. Kindly confirm if there is any particular reason behind this theory that only 2 concepts can be compared together.

asked May 27, 2021 by Aastha Batra
I think we'd probably need more information to provide a good comment.  You could consider writing in to support@sawtoothsoftware.com with your files and some examples.
Note that attribute importance shouldn't necessarily correlate strongly with preference share.  An example is for an attribute like color...which let's say for argument sake has the highest importance of the list of attributes in the study.  Let's say there are just two colors: Red and Blue...and half the respondents prefer each.  You can change a product's color from Red to Blue without much movement in share of preference for a product.  But, since attribute importance is measured at the individual level (as the range of utilities for the attribute), then a high-difference color attribute (at the individual level) can still cancel out at the aggregate market level when you look at "votes" for respondents of product concepts in the market simulator.

However, I think you are confusing the definition of attribute importance (the range of levels within an attribute) with level utility (relative preference for a level within an attribute).
Oh, as I re-read your question, I think when you say "importance" you mean the UTILITY of a single level.  We define importance as the range of utilities of an attribute and utility as the preference score for a particular level.  

So, your question is why if you have 3 or more product concepts in a market simulation, you are seeing less correlation between utility of attribute levels and shares of preference for the product concepts.  The only thing I can think of right now is that with 3 or more products, issues of product similarity are taken into account and used as penalties for product concepts...if using the default Randomized First Choice market simulation method.  In other words, it's possible to give a product an improved attribute level in terms of preference, but to see its share of preference possibly decrease if by doing so this makes this new product more similar to another product within the same market simulation scenario.
okay, thanks for the reply.

Your solution to the original question

Please only use this to answer the original question. Otherwise please use comments.
Your name to display (optional):
Privacy: Your email address will only be used for sending these notifications.
Anti-spam verification:

To avoid this verification in future, please log in or register.