Back to top

A retail food brand in a declining category is seeing its margins being squeezed and needed a way to cut costs in the production process. The decision was made to cut costs through value engineering (making the product with ingredients that could save the company money). 

The company partnered with C+R to conduct a product taste test to determine if there can be cost savings gained by altering the ingredients in one of their current frozen pie offerings, without alienating current consumers. 
 

Problem

The client’s team was interested in determining if there can be cost savings gained by adjusting the formula of one of their current frozen pie varieties. The team wanted to ensure the new cost-reduced product was performing as well as or better than the current line and determine if the prototype can be recommended to move forward without alienating current customers. This research would also identify any product optimization opportunities.

Result

The in-home usage taste test did not show positive results for the new pie. While the new cost-reduced prototype generated the same intent to buy and same level of overall liking as the current pie, it did not reach parity on alienation (negative purchase intent). It also demonstrated taste and texture issues with the crust and topping. 

In addition, some risk was exposed through bootstrap resampling analysis. This analysis uses advanced statistical methods to simulate replications of the product test, and measure the risk of replacement. This analysis showed that a majority of the time, across a range of key measures, the two products were at parity, but those simulated cases not at parity favored the current pie.  The resampling approach can reveal systematic risk that is not easy to detect in standard statistical tests.

Overall, the test pie didn’t seem far off the mark from being a suitable replacement for the current pie. Its performance among heavy frozen pie users was encouraging. However, certain clues in the data suggested that the test pie didn’t satisfy all consumers as well as the current pie, and there was clear evidence of a less-favorable perception of the new formulation’s crust and toppings. If our client’s goal was to introduce a cost-reduced pie with zero risk of consumer dissatisfaction, then this formulation did not meet that standard. However, if the cost savings would be substantial, and the brand could afford to lose some sales among lighter category users, then the math could work out in their favor by moving forward with the cost-reduced pie. All information was presented, along with our recommendations, if they should move forward with the value engineered pie. 
 

Solution

C+R used an in-home usage test (IHUT) study to conduct the taste test among frozen pie users. Respondents were pre-recruited online to participate in an in-home usage test. Each respondent ate one pie in a branded monadic study design. Respondents were split into two cells; half evaluated the new cost-reduced pie and half evaluated the current pie. 

Since the prototype was a frozen product, we worked with our packing partner to pack (including dry ice), label, and ship out the products, with oversight from C+R’s Logistics Director who was onsite for each pack-out to ensure the highest level of quality. 

Each product had a product code label and nutritional information attached. We also developed and included thawing instructions and a usage diary for respondents to fill out as they tasted the pie. For extra precaution, we also affixed temperature testing stickers to each pie box that visually showed respondents if the pie maintained the proper level of frozenness during shipping. 

During the in-home usage period, respondents were sent an online quantitative follow-up survey to provide their feedback on the product, capturing overall appeal, purchase intent, expectations, and perceptions of the product. We also asked detailed questions on appearance, aroma, taste, and texture as well as their usage behavior and preparation methods. 

After processing the data and comparing the results of the test pie to the current pie, we also performed a unique bootstrapping resampling analysis to confirm the results, in which some risk was identified. This resampling approach revealed systematic risk that would not have been easy to detect using standard statistical tests – in this case, the majority of time the two products were at parity; those not at parity favored the current pie. Armed with all the data, and our recommendations, the client decided that more work needed to be done before moving forward with the value engineered pie.