Have an idea?

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

Usefulness of results from Newton MIller Smith extension to Van Westendorp PSM

Hi,

I am looking to get answers around the application of Newton-Miller-Smith extension extension to Van Westendorp PSM. I have referred to Sawtooth's method (https://sawtoothsoftware.com/resources/software-downloads/tools/van-westendorp-price-sensitivity-meter) and manual.

1. Want to understand how robust/ statistically relevant / useful is this method for deriving a demand curve/elasticity and establishing price point preference (didn't find other relevant or elaborate resources online).

2. What other assumptions/considerations should one keep in mind apart from the ones mentioned (calibration of purchase intent scale and purchase probability for end points - too cheap and too expensive)

3. From what lens should one view the results?

4. How does this compare to results derived from conjoint analysis?

Thanks in advance!
asked Oct 20, 2020 by Prakhar

2 Answers

0 votes
I'm the person who wrote the PDF documentation for our PSM template, which you're familiar with.  I don't have any new information to add to it.
 However, two papers delivered at the Sawtooth Software Conference that test PSM are:

Weiner, Jay (2001), “Applied Pricing Research.”  Sawtooth Software Conference Proceedings.  Download from:  https://sawtoothsoftware.com/resources/technical-papers/conferences/sawtooth-software-conference-2001

Bakken, David, Michaela Gascon, and Dan Wasserman (2012), “Contrast Effects on Willingness to Pay: How Conjoint Design Affects Adult Day Care Preferences.”  Sawtooth Software Conference Proceedings.  Download from: https://sawtoothsoftware.com/resources/technical-papers/conferences/sawtooth-software-conference-2012

Perhaps others will write in with their thoughts regarding your questions....
answered Oct 20, 2020 by Bryan Orme Platinum Sawtooth Software, Inc. (198,715 points)
0 votes
Prakhar,

PSM, even with the NMS extension isn't going to give you a price elasticity curve.  PSM by itself is an attitudinal model about pricing, and with the PSM addition, we make it a little volumetric.  With the extension you can model the elasticity of demand to price, but it creates some odd findings (e.g. at very low prices, demand increases as price goes up).  For this and for the fact that I'm not aware of a lot of empirical support for PSM, I don't view it as a viable alternative to CBC when CBC is possible.

For the calibrations of purchase intention ratings to purchase probabilities there is an extensive literature, some of which is summarized in the great Design and Marketing of New Products book by Urban and Hauser.

Personally I like to use PSM, with our without the extension, only for new-to-the-world products.  When a product inhabits a market with competitive products, I think CBC is hands-down the better way to go.
answered Oct 20, 2020 by Keith Chrzan Platinum Sawtooth Software, Inc. (111,275 points)
...