Have an idea?

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

How do I get respondent level willingness to pay?

I was trying to get willingness to pay at respondent level in a CBC set-up. I will require this to build a forecasting model to predict customers willingness to pay. I have performed a CBC study without any competitive scenario. I have 8 different products and I want to know respondents WTP with respect to one fixed base product.
asked Jul 16, 2020 by Sahil Arafat

1 Answer

+1 vote
I'm not sure what you mean by not having a competitive scenario - if you're doing a choice-based conjoint you should have 2+ alternatives in each choice set, right?  

Please see this paper for information on estimating WTP - doing it at the respondent level, your options will be limited.

answered Jul 16, 2020 by Keith Chrzan Platinum Sawtooth Software, Inc. (111,275 points)
Thanks for your response. I have already gone through this paper and have noted down its procedure for converting dollar equivalents from utilities but I am getting some practical difficulties as mentioned in the paper itself. I want to know is there any other way of computing dollar values at respondent level?

Also, with reference to competitive scenario I mean that I have used only one brand in my study. All product profiles are for same brand. In short there is no brand attribute in my study. So do you think we can still estimate WTP efficiently? I was referring to another paper to understand it better "USING CONJOINT ANALYSIS TO DETERMINE THE MARKET VALUE OF PRODUCT FEATURES"


Without a realistic competitive environment, I think your willingness to pay estimates are being estimated in a void and I would probably not place much trust in them.  Now, IF the product is something people must but AND there is only one seller, then I suppose calculating WTP makes more sense.  For example if we have a monopoly situation (the product is the city water system and the customer has only one supplier of water) then I don't suppose you need more competitors.
Yes Chrzan, We have a kind of monopoly situation here and still I am struggling to find WTP for each respondent.
The paper I sent you a link two describes an algebraic way of finding WTP, subtracting the utility for an improved level of an attribute from a base level, then dividing that difference in utility by the utility per dollar ( or euros, etc., calculated from your price utilities) to get a WTP measure.  We think this measure is flawed, but is commonly used, and it may be your only option to get WTP at the respondent level.