It's a simple concept: if a product does not taste good, consumers will not make repeat purchases. Advertising and marketing can get a consumer to try the product, but if it does not deliver in taste, texture or other expectations, chances are the product will fail in the marketplace.

This concept illustrates the importance of sensory evaluation in the product development process. Sensory can determine acceptability of a finished product in the marketplace or determine sensory shelf life of a product. But when it comes to developing a new product, sensory evaluation helps guide development to a more precise goal and can result in cost savings.

"The number one purpose of sensory evaluation in product development is to minimize the chance of putting a product in the market that is not going to be well-liked," says Herbert Stone, Ph.D., co-founder and president of Tragon Corp., Redwood City, Calif. "Sensory ensures that a product will not fail because of its sensory characteristics." Secondly, sensory steers formulators in the direction of what marketing has requested, adds Stone.

Adding Sense to Development

"How do you know when you've reached an acceptable product? How can you decide to use one ingredient rather than another (e.g., sucrose vs. aspartame or low vs. regular fat) unless you can measure the sensory consequences?" questions John Prescott, Ph.D., University of Otago, Sensory Science Research Centre, New Zealand.

Using sensory in your product development cycle answers these questions and more. But what tests achieve the best results in the development cycle?

"I view sensory testing as falling in two categories, either analytical in nature or affective," says Stone. On the analytical side, difference tests and descriptive analysis are the two basic types of tests.

In difference testing, sensory analysts ask panelists if they perceive products as different. Descriptive analysis is far more informative. It tells what characteristics are different and the magnitudes of the differences.

The main difference between the two, says Stone, is time and end results. Difference testing is much quicker. Descriptive analysis takes longer because it provides developers with more focus. "For example, with descriptive, a product development team may find that their product has significantly less fruit flavor than another and the magnitude of difference between the two. Sensory can then ask what formulation was used and recommend how to modify the formula," says Stone.

Analytical tests such as difference and descriptive are often conducted in the beginning and middle of the product development cycle. They can be part of a bigger series of tests to give developers a more precise target to shoot for.

For example, Ann Behen, senior account manager of sensory, Shuster Laboratories, Canton, Mass., notes that the company conducts several types of testing, depending on the project and its stage of development.

"For example, if we are trying to evaluate some concept ideas, we might first begin with focus groups to determine what aspects of a product are important to the consumer," she explains. This generates more qualitative data than quantitative. "We use a couple of consumer groups who would most likely buy the product and place them in a round-table discussion while product developers observe and listen."

Another form of sensory testing involves doing a category appraisal. "For example," says Behen, "a client may want to develop a nutrition bar, but they currently do not make any. We might do an assessment of the top three to five brands and recruit consumers who are users of the product and use acceptance questions to determine what attributes they would like to see more or less of. That gives us guidance as to what to focus development on."

Descriptive panels are used in these types of tests and through the development phase. "We turn to these groups to validate attributes about the product. If it's a development project where we are substantiating the use of an ingredient, we look to see if a descriptive panel is picking up the difference between the two. This also guides the product developers," notes Behen.

Product optimization also provides direction. "Developers create 10 to 15 formulas with varying key attributes. Consumers taste these and through sensory and statistics, we can come out with a more specific picture of what the consumer wants. This may be done in the beginning or mid-point of development to figure out what direction to take next," explains Behen.

Affective tests are used towards the end of the development cycle. Affective testing consists of liking and preference tests. Towards the end of the product development cycle, developers have usually narrowed the alternative product prototypes down to a manageable subset through the use of analytical sensory tests. They are now interested in whether the consumer likes the product, prefers it over another product, or finds the product acceptable based on its sensory characteristics.

Typical sensory affective tests are blind product tests. Affective tests prevent coming out of product development work and not having any idea how the consumer likes the product, comments Stone.

In preference testing, the consumer panelist has a choice. He or she chooses one product over one or more other products. In the measurement of acceptance or liking, the consumer panelists rate their liking for one product on a scale. Acceptance measurements do not require a comparison to another product. Frequently, the most efficient procedure is to determine consumers' acceptance scores in a multi-product test and then determine their preferences indirectly from the scores.

Stone adds that it is important to keep two things in mind with sensory testing. First, there is no universal all-purpose test method. Different methods provide different kinds of information. Secondly, individuals are uniquely different from each other. "The level of sensitivity to using your senses varies by as much as 100- to 1000-fold difference. No amount of screening and training will make people test exactly the same," he notes. Because of this, sensory analysts must incorporate replication in their analytical tests.

Symbiotic Relationship

Sensory and product development should work hand in hand to bring the best products to the shelf. "The ideal relationship between a sensory group and a product development group is one where sensory is brought in the development cycle early on," says Stone.

When product development people meet with the marketing people to set up new ideas and put a product team together, they should contact sensory at that point. If you leave sensory out, they don't know what the background and context of a product is, so they don't have as strong an ability to design a test or series of tests to try and answer questions, he adds. Sensory needs early notification in order to set up panels if you really want precise information about products and people's perceptions.

In addition to early notification, Behen notes that the two groups, although working together, should act as separate groups. "I think it's best to keep them separate groups because there is too much ownership of product when in product development. It can be a conflict with developers if they are too involved in the sensory work."

Sensing the Future

As with all industries and fields, the electronic age will enhance the future of sensory evaluation. For example, the sensory evaluation division of the Institute of Food Technologists (IFT) has established an e-group where anyone involved in the sensory evaluation world can communicate and post questions at the site. "It is a wonderful source of information. I've asked about proper lighting when we were designing our new facility and other technical questions," comments Behen.

With the electronic age, consumer panels can input their responses directly into a program and the sensory analyst can download information and run statistical analyses in shortened times. "You'll have data done and output within 24 to 48 hours," says Behen.

The future and view of sensory is evolving. More and more, food scientists recognize it as an integral part of the product development cycle.

Sidebar: Sensory Goes Global

"With many large food companies now acting on a global scale, plus the increasing receptiveness of many new markets, particularly in Asia, to Western foods, there is a need to ensure that sensory science addresses the special challenges inherent in global product development," says John Prescott, Ph.D., University of Otago, Sensory Science Research Centre, New Zealand. Prescott offers some advice when using sensory in the development of products for global markets.

Define the important parameters for product success in a particular market. Sensory studies should not only assess the relative importance of sensory characteristics in influencing overall acceptability of foods, since these may differ between cultures, but also recognize that non-sensory factors (familiarity, expectations, attitudes) also influence preferences. Assess appropriate parameters for products in advance of product development. If the aim is to develop products for a specific export market, then you need to develop sensory techniques that you can use to guide product development. One approach may be to assess the "flavor principles" for particular foods by assessing appropriateness or expectations for sensory characteristics in a particular market. For example, you may ask what are the acceptable flavors for biscuits in your export market. Utilize sensory techniques that are not influenced by cultural differences in language or conceptual framework. Discrepancies between Australian and Japanese panels in assessing complex sensory qualities such as fruitiness and creaminess raise the issue of whether another culture perceives or conceptualizes sensory qualities equivalently. Since ratings of overall liking are least likely to present such interpretation problems, they can be combined with detailed sensory analysis by panels in the home market to provide understanding of the basis of liking. Alternatively, products with variations in key ingredients can be assessed in factorial designs.

Prescott can be reached via e-mail at john.prescott@stonebow.otago.ac.nz.