Guest Blog: Survey-Based Techniques for Optimizing Prices, Product

In our guest blogs we feature leaders in market research sharing their views about best practices, innovation, and emerging trends. Below, hear Sawtooth Software CEO Bryan Orme’s thoughts on best practices for utilizing survey-based approaches for product and pricing optimization. Sawtooth is one of the world’s leading software firms in the area of choice/conjoin analysis, as well as an MRII sponsor. 

Optimizing product features and pricing is a common and valuable deliverable that we as insights and strategy consultants are tasked to give.  While data from actual sales are often available, such data rarely can answer the questions surrounding product and pricing optimization (especially regarding new product variations) consistently well.  Researchers thus regularly turn to survey-based methods to augment expert opinion and existing real-world data.  And, conjoint analysis has emerged over the decades as the most widely trusted approach for product optimization and pricing.

Some survey-based approaches work better than others for product optimization and pricing.  First, the common 5-point and 10-point rating scales are problematic when it comes to directly asking respondents what’s important to them or to evaluate the firm’s performance on various aspects of product/service delivery.  Rating scales suffer from lack of discrimination and scale-use bias.  Respondents can be lazy in ratings-based batteries of questions on multiple items, leading to responses that make it look like everything is important.  Scale-use bias makes it hard to compare groups of respondents, especially results from different countries.  Scale-use bias also leads to artificial correlation among the variables in your data set, making it hard to conduct common analyses such as drivers analysis to discover what’s important to people and/or driving their choices and loyalty.

Van Westendorp PSM (VW) and Gabor-Granger (GG) are two commonly used pricing research approaches that suffer from similar limitations and problems.  A key limitation is that these two techniques usually focus on just one product concept.  Thus, if you are trying to optimize pricing for multiple variations of your product concept, then you’d much rather rely on a better method such as conjoint analysis that efficiently tests 100s or 1000s of product variations within the same survey.  Furthermore, VW or GG ask respondents to propose or evaluate potential prices absent the realistic competitive context that is usually present in the real world.

Conjoint analysis deals better with every issue described above.  Because we’re asking respondents to choose a best product (among a set of competitors), there’s no 5-point or 10-point scale involved that could have problems with lack of discrimination or scale use bias.  Furthermore, conjoint analysis allows you to test 100s or 1000s of product formulations within the same survey, making it more powerful and flexible than VW or GG.  Conjoint analysis also does a better job imitating the context and kinds of decisions that buyers make in the real world: choose among available product options or walk away and not purchase (the “None” alternative in conjoint analysis).  To get realistic data from respondents, we should ask realistic-looking questions that mimic the real-world context and decisions.  When we ask challenging trade-off questions via conjoint analysis, we learn what people truly value.

Sawtooth Software is the established thought leader in the field of conjoint analysis for 40 years now.  For a free trial version of its platform for conducting conjoint analysis, MaxDiff, and general surveys, visit

Share this post:
Share this post
Recent Posts

Subscribe to our Newsletter

By submitting this form, you are consenting to receive marketing emails from: . You can revoke your consent to receive emails at any time by using the SafeUnsubscribe® link, found at the bottom of every email. Emails are serviced by Constant Contact