A delicious discipline: sensory analysis

Have you ever asked yourself how a certain company managed to develop that fantastic beer that you just can’t help but buy every time you see it on a shelf at the supermarket or at your regular bar?

Or that savory cheese that goes so perfectly with your favorite crackers?

Or that chocolate, with notes of toasted hazelnut that fill the room and that you munch on between breaks at work?

How did these companies manage to develop products that so perfectly satisfy your taste?

The answer is undoubtedly a complex one, and many factors come into play in the process of developing a new food product.

Nevertheless, there exists a scientific discipline whose objective is to measure, interpret, and understand how people perceive a certain food product through their senses of sight, hearing, smell, and taste (Lawless, 2010), seeking to isolate the unique sensory characteristics that could potentially cause the target consumer to choose or reject it.

This discipline is known as sensory science, which, with a multi-disciplinary approach, takes advantage of other scientific disciplines like statistics, psychology, neuroscience, chemistry, and so on.

Sensory analysis is the tool that applies this scientific knowledge, developing specific methods which are useful in evaluating food products (Sacchetti, 2008).

In sensory analysis, three primary methods can be identified which are used to investigate the relationship between human perception and the characteristics of a product: discrimination tests, descriptive tests, and hedonic tests.

When was Sensory Analysis created?

Since ancient times the evaluation of the quality and enjoyment of a food product has been based on the sensations that this product elicits. To evaluate the quality of a piece of merchandise, and have confidence that they were purchasing a re-sellable product, shopkeepers would judge a portion of the food to be purchased (hoping that it was representative of the whole), evaluating the color, smell, taste, and consistency.

It was in response to this very need that, at the beginning of the 20th century, specialized figures were created, true tasters, whose job was to evaluate the quality of a product and ensure that it remained consistent over time.

In the 1940s and 1950s sensory analysis received a huge push through the U.S. Army Quartermaster Food and Container Institute, which supported research on the acceptance of food by the armed forces. Through this research, it became clear that simply formulating a complete and nutritious meal wasn’t enough to guarantee that military personnel would accept the food. In short, people weren’t eating everything they were given!

The objective of this study was to understand which food products were more or less preferred and to develop a method with which to measure a food’s degree of acceptance.

The development of sensory analysis made it possible to understand that the sensory information collected on a food product was more reliable when taking advantage of a panel made up of 10-15 people, all equally trained to evaluate a list of sensory descriptors and a specific food matrix. Furthermore, specific tests were theorized and applied, aimed directly at consumers in order to understand their perception of this same product.

In the second half of the 20th century, the first PhD programs and the first companies specializing in sensory analysis began to emerge in the United States. Many scientific journals (Chemical Senses, Journal of Sensory Studies, Journal of Texture Studies, Food Quality & Preference) can be consulted in order to remain up to date with the developments in this field.

In practice, what’s it for?

Sensory analysis is a tool that can help reduce a company’s risk and uncertainty regarding the development, testing, and consumer perception of a food product. The methods mentioned briefly at the beginning of this article, are an example of the tools that are adopted to reduce this uncertainty.

During quality control, a discrimination test can be fundamental to verifying whether or not the changing of an ingredient has caused a perceivable sensory modification in the product.  The most commonly used tests are the triangle test and duo-trio test. In both tests, panelists must identify, also forcibly, the product that they perceive to be different from the others or how a product compares to a reference standard. In this way, it’s possible to statistically evaluate whether or not the difference perceived between the old and the new formula is significant.

During a product’s development phase, it can be useful to apply a descriptive test in order to understand and quantify the sensory characteristics of various pilot samples proposed by the R&D team. In this case too, various methodologies exist, all of which use a numerical value to quantify the intensity of the sensory attributes.

For marketing, it can be useful to study the level of enjoyment or disgust of one’s product as compared to that of various competitors. Likewise, a catering service (in schools, hospitals, restaurant chains) could be interested in knowing which pairings on their menu are more or less appreciated and thus reduce the risk of waste. In this case, hedonic tests can be used, which require the direct involvement of the consumer and use a scale from 0 to 9, in which each number is associated with a level of appreciation (ex. 0 = I don’t like it at all, 9 = I like it a lot).

We use it too!

Even at Vinhood we use sensory analysis in what we do, pairing it with research in the field of neuroscience and the development of a machine learning model. Our objective is to help you understand, based on your personal taste, what product is more or less right for you.

What methodology did we use in this case?

We chose to use descriptive analysis in order to quantify the sensory characteristics of various products, which are subsequently grouped together using clustering methods to highlight those with similar characteristics. In this way, we created the #Characters which you’ve probably already seen mentioned on our website and social media channels.

To use descriptive analysis, we put together a panel of at least 10-12 people trained in recognizing and evaluating key sensory descriptors for the studied food matrix. The analysis of each product is repeated at least twice and the analyzed sample is statistically significant for the category being considered.

Every cluster of products has certain characteristics that coincide with the preferences of some people, but not of others, and thanks to our predictive model of machine learning we’re able to give you personalized suggestions!

If during the wine Taste Test, for example, you end up with an #Outgoing character, you know that all fresh and savory white wines will match your taste, and you’ll even be given some reference labels in order to help you start discovering your palate!

Bibliograpfy

Lawless, H. T., & Heymann, H. (2010). Sensory Evaluation of Food: Principles and Practices (Vol. 2). New York: Springer.

O’Mahony, M. (2017). Sensory Evaluation of Food: Statistical Methods and Procedures. Routledge.

Stone, H., Bleibaum, R. N., & Thomas, H. A. (2012). Sensory Evaluation Practices. Academic Press.

Webography

https://agriregionieuropa.univpm.it/en/content/article/31/15/qualita-dei-prodotti-agroalimentari-e-scelta-dei-consumatori-il-ruolo-della