Knowing When the Price is Right

Research on pricing products and services, obviously popular, can bring successful results through several different research methodologies: find the right one for you

CHICAGO – DECEMBER, 10, 2009—A recent Google search on “pricing” turned up 233 million results. “Pricing research” turned up over 33 million. Pricing products and services evidently are popular topics and no less so here at Sawtooth Technologies Consulting, although you won’t find it referenced 33 million times—yet. Given this popularity, we’ll assume that you know how important research on pricing has become.

What we’ll focus on here is when to use which method. In particular when to use conjoint analysis/discrete choice, Van Westendorp, or the monadic approach to conducting pricing research. Perhaps in a future blog we will delineate within the different conjoint and discrete choice methodologies, but to avoid trying to tackle too much in one entry, we’ll keep them grouped together here and refer to conjoint and discrete choice collectively as conjoint.

The main factors that will influence your decision regarding which methodology to use are:

  • Configuration variability: whether you are considering varying other features of your product or service in addition to price
  • Sample size limitations
  • Understanding of the relevant price range

Product Configuration: When Price is one of many features you are studying

If price is one of many features you are studying, you will want to use conjoint analysis to help you choose the right price. Conjoint forces respondents to make trade-offs among feature sets and, when conducted properly, can provide an understanding of price elasticities and the value of certain features with respect to others. So, if you are determining how best to configure your product, what features to include and how to price those features to achieve a specific goal—such as optimizing market share or profitability—conjoint is the preferred approach.

Pricing a product where the configuration is set

If, on the other hand, you are looking to price a product where the other features are already determined, then it makes sense to apply the monadic approach to pricing research. In this approach, you choose a few prices that you determine most applicable based on your understanding of the market, focus groups, etc. Then you conduct surveys asking purchase likelihood questions to different sets of sample, where one sample set would see the product offered at a certain price, another sample set would see the product offered at a different price, etc. If, for example, you collected four sets of data, you could then plot the curve of the percent of respondents who would definitely or probably buy at each price point.

Pricing if you have a set configuration and the need for a smaller sample size, or if you have no pre-conceived understanding of the relevant price range

As might have been obvious from the monadic approach, since you are basically conducting multiple surveys, the monadic approach requires a larger sample size. If you don’t have that luxury, a possible alternative is the Van Westendorp approach, which needs only sample sizes similar to those for a conjoint study. With a Van Westendorp survey, you would present the product to your respondent and ask them four questions about price: 1) At what price would they consider this product a bargain; 2) At what price would they consider it was getting expensive; 3) At what price would they determine the product was so cheap that it must not be good quality; and, 4) At what price would they consider the product to be too expensive.

Since the responses to these questions are open-ended, Van Westendorp is often used when the product is entirely new or, for whatever reason, the researcher has no pre-existing indication as to the appropriate range of prices before conducting the study. After collecting the data and plotting the responses, a relevant range becomes suggested.


Each of these approaches can provide a wealth of information and each requires an understanding of the nuances to optimize the results of the approach. However, the above should provide a quick understanding of where to turn your attention depending on what you’re trying to accomplish: conjoint/discrete choice when price is one of multiple features you are trying to configure; monadic when it’s not and you have sufficient sample size; or, Van Westendorp if you don’t need to configure other features and your sample size is limited to that of conjoint, or if you have no pre-conceived understanding of the relevant range of prices.

For more information on all of the market-changing research techniques from Sawtooth Technologies, visit

« Back to Blog

>> Follow us on Twitter for the latest from Sawtooth Technologies Consulting.