Using Choice Modeling for Product and Price Optimization: A Basic Overview of Possible Methods
Topics Discussed
- Objects
- Methods
- Pros & Cons
- Examples
Product Optimization
Defining Appropriate Research Objectives
- What set of new product ideas are most preferred or would reach the most customers?
- What combination of messaging and claims would maximize market share or revenue?
- What combination of brand and product features would maximize demand? Revenue? Profit?
- What price structure should I use?
- What bundles of products in a product line will most improve revenue or profit?
- What product packaging would generate the highest purchase interest?
Price Optimization
Defining Appropriate Research Objectives
- What price should I charge for my new product?
- What price change for an existing product will increase revenue/units? By how much?
- What price changes within my product line will increase revenue/units? By how much?
- What price structure should I use for technology or durable goods? Bundled vs. a la carte, base product + additional fees? Installation fees?
- What price structure should I use for services? Flat rate? Usage-based?
Product Optimization Methods
Choice Modeling Without A Competitive Set
Choice modeling without a competitive set respondents see several versions of the product on a screen and selects their most preferred. They then rate likelihood to purchase that product.
Advantages
Delivers accurate measure of:
- Relative importance of attributes.
- Relative preferences among levels within attributes.
Disadvantages
- Respondent task is repetitive.
- Omits competitive effects.
- Omits brand-price and brand-feature interactions.
- Does not measure cannibalization within a product line.
MaxDiff
The next method is max diff. In this method respondents see sets of product features and select the most and least important to their purchase decision.
Advantages
- Works well with a variety of items, such as products, features, ideas, claims, benefits, etc.
- Produces a full ranking of up to 20 attributes without presenting all pairs.
- Simpler and more focused task for the respondent.
Disadvantages
- Respondent task is repetitive.
- Omits competitive effects.
- Omits brand-price and brand-feature interactions.
- Does not measure cannibalization within a product line.
TURF
In Turf (Total Unduplicated Reach and Frequency) analysis, the survey respondents rate several product features on both purchase intent and frequency of purchase.
Advantages
- Inexpensive and quick.
- Face validity.
Disadvantages
- Overstates purchase interest and reach.
- Omits competitive effects.
- Omits brand-price and brand-feature interactions.
- Does not measure cannibalization within a product line.
Choice Modeling With A Realistic Competitive Set
In Choice modeling with a realistic competitive set, survey respondents see several screens of products on each screen. For each screen they can choose to buy any of the products or none of the products. They can be asked to state how many of each product would be purchased. Options presented to the respondents may include a menu of choices.
Advantages
- Gives more accurate market-share predictions.
- Includes competitive and context effects.
- Delivers brand-specific price and feature effects.
- Can evaluate product lines, packages, or bundles.
Disadvantages
- More difficult to design.
- More moving parts.
- More costly to implement.
Joint Stated Revealed Preference (JSRP) Modeling
Joint Stated Revealed Preference Modeling (JSRP), the respondent task is exactly the same as was described previously for the Choice Modeling with a Realistic Competitive Set. The difference with this method and the Choice Modeling with a Realistic Competitive Set is that JSRP combines data from two sources; the choice exercise, and in market behavior in the form of actual sales data.
Advantages
- Gives most accurate measure of price elasticity.
- Measures impact of feature changes on revenue.
Disadvantages
- Requires special software and an experienced modeler to implement.
Price Optimization Methods
Several of the methods mentioned above are appropriate for price optimization. The graphic below looks at the relative accuracy of price elasticity estimates produced by the different techniques
Research Method Decision Tree
The decision Tree below help us to determine which technique is best to use of the different methods discussed.
For example, if the goal is to optimize the elements of a product, (the features and claims that go on the product, the claims to go in the packaging, the price of the product), then the appropriate technique would be choice modeling with a realistic competitive set.
Product Optimization Questions & Appropriate Techniques
Questions
- What set of new product ideas are most preferred or would reach the most customers?
- What combination of messaging and claims will maximize market share or revenue?
Appropriate Techniques
- MaxDiff
- Choice Modeling Without Realistic Competitive Set
- TURF
Questions
- What combination of brand and product features would maximize demand? Revenue? Profit?
- What bundles of products in a product line would most improve revenue or profit?
- What product packaging would generate the highest purchase interest?
Appropriate Techniques
- Choice Modeling Without Realistic Competitive Set
Price Optimization Questions & Appropriate Techniques
Questions
- What price should I charge for my new product?
Appropriate Techniques
- Current Market Choice Task, Concept Test, Choice Modeling With Realistic Competitive Set
Questions
- What price changes for an existing product will increase revenue/units? By how much?
- What price changes within my product line would increase revenue/units? By how much?
- What price structure should I use?
- For technology or durable – Goods bundled vs. a la carte? Base product + additional fees? Installation fees?
- For services – Flat rate? Usage based?
Appropriate Techniques
- Choice Modeling A With Realistic Competitive Set
Presenters
John Colias, Ph.D.
Senior VP Research & Development
As a leader with both university teaching and business consulting experience, John focuses on predictive modeling, prescriptive analytics, and artificial intelligence. As Senior Vice President, Research & Development, at Decision Analyst, John combines academic and business interests to help analytics professionals by offering cutting-edge analytic solutions tempered by business realism. He holds a doctorate in economics from The University of Texas at Austin, with specializations in econometrics and mathematical modeling methods.
Elizabeth Horn
Senior VP, Advanced Analytics
Beth has provided expertise and high-end analytics for Decision Analyst for over 25 years. She is responsible for design, analyses, and insights derived from discrete choice models; MaxDiff analysis; volumetric forecasting; predictive modeling; GIS analysis; and market segmentation. She regularly consults with clients regarding best practices in research methodology. Beth earned a Ph.D. and a Master of Science in Experimental Psychology with emphasis on psychological principles, research methods, and statistics from Texas Christian University in Fort Worth, TX.