Have an idea?

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

Material regarding WTP in the Market Simulator

Hi,

I am conducting a CBC study for my thesis with the major goal of estimating willigness to pay for a new product feature.
Besides the algebraic and the two-product market simulation approach, I also want to use the new Sampling of Scenarios method in Sawtooth, as described in the paper "Estimating WTP Given Competition in Conjoint Analysis".
To be able to include it in the thesis, my supervisors asked me about more information how the procedure is conducted in detail and if there is any more research material published, regarding this method. I was not able to find something appropriate via google.

Thanks in advance
asked Oct 13 by anonymous

1 Answer

0 votes
Well, sampling of scenarios is just a way to generalize the simulation approach that involves finding the price indifference point that leads to the same share of preference outcome for a product offering vs. an array of competitors.  The key WTP approach that differs from the two-product approach and the algebraic approach is that notion of having multiple realistic competitors (including a None alternative) and finding the price differential for an enhanced offering that drives its share (simulated share of preference using a Choice Simulator) back to the original starting point.

 Have you found this white paper on our WTP method in our Tech Papers Library yet?  https://sawtoothsoftware.com/resources/technical-papers/estimating-willingness-to-pay-in-conjoint-analysis
answered Oct 13 by Bryan Orme Platinum Sawtooth Software, Inc. (190,565 points)
Oh, and don't forget this white paper from 20 years ago that laid out this long-held approach of using market simulators to estimate WTP, given rich competition.  It didn't specify the SOS trick yet, but the fundamental approach is laid out:  https://sawtoothsoftware.com/resources/technical-papers/assessing-the-monetary-value-of-attribute-levels-with-conjoint-analysis-warnings-and-suggestions
Hi Bryan,

thank you really much for your fast answer, that helps me a lot. I will also have a look on the second paper.

Regarding my results from the WTP Calucation I got a second question regarding extrapolation and convergence, because my result is marked with "*0% of the sampling draws resulted in extrapolated values, and 0.7% did not converge".

I think you already explained in the forum what the extrapolated values mean but I am not sure how I have to understand the convergence. Same question as above, is there some research material where this is addressed?
Is a result that does not converge to 0.7% to be treated with caution or can it be used?  

Thank you in advance
Lack of convergence just means that this particular SOS draw of a test product vs. competitors led to a situation where enhancing the test product and finding the increase in price to drive the share of preference for that test product didn't return a solution.  There can be some strange instances where one cannot find an increase in price that returns the share to its original amount, even without considering extrapolation.  If only 0.7% of the draws of SOS lead to lack of convergence, that's really not anything to worry about at all.  I wouldn't start to worry much until 5% or more of the solutions failed to converge.
Hi Bryan,

thank you again for your answer.  That makes it clear to me.
I was just wondering about your second sentence "even without considering extrapolation". Does this mean that there is no try to use extrapolation in these strange instances?
Usually in these strange instances, extrapolation is attempted as well...but to no avail and no solution is able to be found that drives the share back to the original share.
Okay perfect. Thank you for your answers, they helped me a lot.
...