Have an idea?

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

Feature Pricing Study - Estimates Willingness To Pay

Dear experts,

I am conducting a conjoint study that the aim is to evaluate the consumers' Willingness To Pay monetary for each car feature. There are 20 features to be tested in total, and my current design is that each feature is defined as an attribute, with 5 price points as levels plus "no this feature" level.

Now, I've got the part-worth utilities for each respondent. I'd like to calculate each feature's willingness to pay using these utilities, but after reading most articles as mentioned in other posts, I still cannot work out the proper results. Hence, it is highly appreciated that someone could help to brief the calculating willing to pay process or guide relative articles to read. I am already aware the dangers of calculating the willingness to pay, and will use it with caution.

Thanks very much for everyone's great support.
asked Nov 10 by ericdee Bronze (1,800 points)
retagged Nov 10 by Walter Williams

1 Answer

0 votes
So you have price built into each attribute?  For example Attribute A might have these six levels?
Attribute A, $
Attribute A, $$
Attribute A, $$$
Attribute A, $$$$
Attribute A, $$$$$
Attribute A, not available
Is this correct?

If so I'm not sure how you'll get WTP.  The WTP calculations I've seen require separate utilities for price and for the attributes and I don't see how you can disentangle price and attribute with your design.  

Perhaps you've not shown price with each attribute, but rather summed it across attributes and then shown it as a separate (total) price?  If so, then you should have a price utility and attribute utilities, and you can compute WTP in a standard way.
answered Nov 14 by Keith Chrzan Gold Sawtooth Software, Inc. (48,525 points)
Thank you Keith. Yes, I set the attributes and levels as you described above. The interview programming is to hide price with each attribute, just show with or without this attribute, then summed all prices at the bottom of attribute ccombination.
After estimation, I have price utilities for each attribute, but no separate attribute utilities.  Hence, in this case, can I also work out WTP?
Regarding your comments on " you should have a price utility and attribute utilities, and you can compute WTP in a standard way", may I know what is the standard way to calculate WTP?
Thanks again for your great support, Keith.
For the standard way of estimating WTP (multiply utility of your attribute by the ratio of price in dollars to difference in utility across price range) you would need a separate price utility, not one linked to the attributes.  

I do not know how to estimate WTP if you don't have a separate price utility, but I'm not sure why you can't get a separate price utility from your design.  If you have each variable represented as binary attributes (on/off) and then a price utility (probably treating price as a linear function is the way to go, so that the utility is the slope of that function) you should be able to get a single price utility, right?