Optima® Home-Use Product Testing

  • Optima Product Testing
    Decision Analyst’s principal home-use product-testing system is Optima®, a monadic, normative system comprised of modules of standardized questions.
    The system is tailored to each client’s product category and requirements.
 

The essential features of Optima® are:

  • Monadic Design. Each product is tested alone. This provides the most accurate evaluation of the product and the best diagnostic feedback about how to improve the product.
  • Standardized Systems. The sampling, data collection, data preparation, and data tabulation methods and procedures (i.e., the systems) are standardized for each product. It is essential that every product be tested in precisely the same way.
  • Standardized Questions. The standardized questionnaire is modular in structure and flexible in design. Decision Analyst owns copyrights on all of the questionnaire modules.
 

How Does Optima® Work?

Typically, a representative sample of category users (150 to 200 households) is given a test product to use in home for a few days. Then these consumers are asked a series of standard questions about the product. The Optima® core questionnaire consists of:

  • Overall rating
  • Likes
  • Dislikes
  • Diagnostic attribute ratings
  • Component evaluation
  • Purchase interest
  • Pricing
 

If the product test is a part of a new product's volumetric forecast, then additional questions are added to the core questionnaire.

Based on internal diagnostics, normative data, and analytical models, these standardized questions tell us whether the product is optimal or not and indicate what needs to be changed to improve the product. The primary analytic model is our Pii®, the Product Improvement Index, a unique mathematical model to help optimize product formulation or design.

Pii®—Product Improvement Index

Decision Analyst developed the Pii® mathematical model to help guide developmental efforts for new products and the reformulation of existing products. Pii® was developed because of problems encountered in using various types of regression models in product-testing analyses. Regression models assume that all of the input variables are independent (i.e., not intercorrelated in any way). The reality is, however, that virtually all of the input variables that might explain a product’s performance are typically intercorrelated. The result is a regression equation that omits important input variables. For example, if the color and the sweetness of a product happen to be highly correlated with each other, the regression equation would omit one of the variables. We might think we had a color problem when in fact we had a color and a sweetness problem.

The Pii® model was designed to correct the “missing variables” problem associated with regression. This model is based upon a type of correlation, using “dummy” variables, to examine the relationship between the diagnostic ratings (e.g., too sweet, about right, not sweet enough; or too much salt, about right, not enough salt; and so on) and the consumer’s overall rating of the product. The overall rating is typically measured with an 11-point scale.

The output of the Pii® model is a table of important explanatory variables along with the Pii® rating and the indicated action, as illustrated here.

Diagnostic Variable
Pii® Score
Indicated Action
Too sweet
18.65
Reduce sweetness
Too dark in color
14.72
Make product lighter
Too soft
12.95
Make product firmer
Not enough salt
9.48
Add some salt
Not enough crunch
5.23
Make product crunchier
 

Generally, any Pii® score greater than 4.0 indicates that some modification of the product might be necessary. The greater the Pii® score, the more important that variable is and the more that variable should be modified.

Optimization Methods

In addition to Pii® analyses, response-surface and choice-modeling analyses are available optimization techniques. Experimental designs and simulation models are employed to optimize products. By testing chosen subsets of product possibilities, response surface and choice modeling can simulate and predict consumer preferences for hundreds of product possibilities, as defined by variations in ingredients, features, elements, or packages. The resulting equations are used to build an optimization simulator so that “what if” products can be fully explored and understood. The goal of optimization can vary. It might be maximizing consumer preference, or maximizing the profit margin without losing market share, or maximizing sales potential. The optimization simulator also helps reveal “cause and effect” as inputs are changed and outcomes vary.

Labeling and Shipping Test Products

Decision Analyst operates a large mail-processing facility for the explicit purpose of labeling test products, packing those products in shipping containers, labeling the boxes, and shipping all over North America, Europe, and elsewhere for in-home product testing. This facility’s staff can handle complex study designs, rotation schemes, and labeling requirements.

Product Optimization Services

Decision Analyst is a recognized leader in consumer product testing and optimization. Its staff has evaluated more than 1,000 foods, beverages, and other products during the past four decades. The firms has many staff members with extensive experience in the conduct and analysis of product testing and optimization studies. The company is a leader in the development of analytical techniques to enhance product testing and optimization.

If you would like more information on product testing, please contact Jerry W. Thomas, President/CEO, by emailing him at jthomas@decisionanalyst.com, or by calling 1-800-ANALYSIS (262-5974) or 1-817-640-6166.