Since the publication of our article Sensory analysis, a misunderstood discipline, a number of questions and reactions followed. They have been shared in Facebook groups, however, as we believe this might be useful to non Facebook users too, we are trying to respond here as well.
“A tasting can only be subjective.”
The answer to this is in the very first paragraph of our article about misconceptions around sensory analysis:
“Sensory science is a relatively young research area which, up to now, often had to deal with establishing itself as an “objective” discipline. At its core, it desires to be a science that, while dealing with subjective data, produces reliable, exhaustive and repeatable results.”
And this is why there is so much reluctance and confusion about the sensory analytic method. After all, the language used for the outcome of all of this rigorous work, that is done in the back scene, looks like the language used by any consumer or solo expert in their fully subjective write ups. The difference is in the backend, in the process. And that is why there are so many misconceptions about the benefits that sensory analysis can provide. It all looks the same, right? My subjective sensory evaluation is not less worth than the outcome of a trained panel, according to some people.
So why should you trust it? And how do you measure the value of sensory analysis in the cigar industry?
You can’t measure it by simply comparing cigar analyses with reviews from all different media sources. You’ll get headaches because everyone says something different. Except if you have a structured vocabulary, a structured lexicon, and you map everything against it. Then you can maybe find some agreement across various sources, even if they are single reviewers.
But, if you really care for how a cigar tastes and don’t want to waste your time and money, we highly recommend to pay attention to the source accuracy. The ultimate goal is to be satisfied by the cigar you chose based on the description published by the source you are looking at. If the source achieves its goal consistently, it does not matter how subjective the review is. What matters is that the cigar you were suggested to try satisfies you and your palate. You may have found the Robert Parker of wines transposed to cigars, an individual whose palate is very similar to yours. It’s perfect if you find that. But if you are in the situation in which Franca has been for a long time, if you are only confused by all of the different reviews that say something different every time, then what do you do?
“If you tell people what they will find in a cigar, they will because of the confirmation bias.”
The predictive confirmation bias is a very wide-spread cognitive error which consists in letting our mind predict what you will experience based on some wrong assumptions or on some past experiences that don’t have any reason to be replicated. Preconceptions and cliches are also forms of biases.
One example is when some people hear or read some others (that they consider as one level up vs. themselves, so more knowledgeable) describing a cigar with certain attributes and, rather than effectively experience and focus on what they sense, they just say they are experiencing what they remember they read or heard from others. They either deliberately ignore the evidence that their senses is delivering or they are not able to recognize what they are sensing. It is like finding your way in San Francisco, looking at a map of New York.
This happens with consumers, unfortunately when they are unable to identify accurate sensory information sources. In fact, if they gave credit to accurate sources, they might realize that the evidence corresponds to the predicted aroma or taste or whatever. They might use such sources as educational resources.
Instead, what sadly happens today is that people are persuaded to buy by biased reviews which cannot keep the promise of what they write. This reinforces the idea that taste is subjective anyways, so consumers accept to just being disoriented and buy whatever they are told to buy.
Finally, the worst application of a predictive confirmation bias is generally found among reviewers rather than among consumers, who are often just trying to understand what they are buying.
“You can quantify nicotine strength but not flavor.”
This is not true. You can measure nicotine strength as well as aromas, tastes, tactile sensations and any other attribute that you define are important in your model.
The challenge with flavor components has less to do with measurability, more with the training of the tasters.
“There is no such thing as a blind tasting.”
We understand that it is very difficult to ask someone to cover our eyes before we choose the cigar we need to taste from the batch of samples.
There are a few things that can be done within a panel to reduce the visual effect in a blind test. One could be using a colored light bulb that filters the real cigar wrapper color. The masking is not total, but the wrapper colors are confused. Another, even more critical, way to deal with such challenge is the training of panelists. The understanding of biases (and the demonstration of this through the data) is extremely important to reach an acceptable level of panelist performance.
“The best critic is the one who does not necessarily have the best palate, but the one who has the best way to bring the information to the consumer.”
From a marketer point of view, this is totally true. From a consumer point of view, we believe it is necessary to clarify based on the consumer’s goals: if a consumer wants to be entertained and persuaded to buy, this might be true. If a consumer wants to make educated purchases and cares for a cigar’s flavors, it is not only necessary to have a trained palate, but also to work within a panel and not as a standalone critic.
“For me “to taste” means “to judge”, whether a single or a panel does it makes little difference (in the definition). Obviously the method changes. Do you understand it differently?”
A tasting is an exposure to stimuli that involve our senses. And typically this is non casual, it’s something that needs to be somewhat organized. For most of us, it’s difficult to avoid hedonic reactions, meaning it’s difficult to avoid to say “I like it”, “I don’t like it”, and that naturally leads us toward judgment. And, indeed it is up to the framing and the definition of the methods to distinguish the tasting and make it become a tool, a tool which can serve different purposes.
“Sensory analysis applies more to food products. In extending it to the cigar, what factors should be added or excluded to the sensory analysis process?”
In rigorous sensory tests, the question is not so much what you evaluate, but it is the objective. Why are you evaluating? The testing location and the parameters are almost identical for all contexts. It’s the behind the scene, the preparation and the presentation practices that change, not really the task itself. In the case of the cigar, of course, because we don’t have access to a lab with special lights, we need to consider the element of “at home testing”. And immediately this sounds like it’s a hedonic approach, but the personal routine is important. That’s why we need to focus on training the panelists. We have to introduce rigorous guidelines that actually also insist on the prejudice and the distractions that we need to avoid.
The context in which you smoke every time you test needs to be as repeatable as possible. We also need, for instance, to take into account our saliva flow and we need to consider what are the best times, and what is the best physical condition at which it’s best to taste, and maintain that as a routine. If I test a cigar after dinner, it’s not the same as testing a cigar late morning or late afternoon if I haven’t had coffee breaks or so. There’s a lot of things that are taken into account.
“The results of a panel are obtained by applying statistical techniques to the data provided by the panelists. And one of the things that is done is a cut of the data that deviate from the average. Can this be seen as experience education, but also as an adjustment, a sort of adaptation of the panelists to the group? And also this phenomenon could lead the panel’s results to flatten out. If a producer favors this trend, he risks a drift in the characteristics of his product.”
Adaptation and conformity are well-know phenomena that are driven by panel-interaction – a good reason to have panelists tasting individually and blind. It is true that the list of attributes, for example, may be restrictive, which is why two aspects are continuously emphasized: 1. revisiting the vocabulary to determine if it needs modification, and 2. Encouraging each individual panelist to always report additional perceptions they may have. The statistics are not there to exclude, but to verify: If an individual is reliable and consistent in finding a specific attribute (that nobody else finds), that is a very important finding also. The initial trouble is only in determining that it is indeed a recurring, reliable attribute.
The whole of the data is not generally seen by each panelist and the objective is description, not judgment. Descriptive analysis is fundamentally different from judgements and point ratings, as there is not right, wrong or “best”.
The fallacy you describe is fundamentally one of marketing focus-groups: Consumers themselves, when given the opportunity to ask for EVERYTHING they like, do not make the best decisions (see Pontiac case).
But descriptive analysis as a tool is neutral until additional data is introduced to focus on a particular aspect, but it is still fundamentally non-judgmental by design and this enhances the product rather than driving it adrift, because a manufacturer must always take into consideration at least the degree of incidence and the actual importance that the consumer gives to a characteristic of the cigar, with its dry evaluation and the correlation with the measurements of the panel.
If you only take data from consumers who individually described the cigars in multiple ways, it’s very hard to translate all of that into a language that you can make actionable, that you can understand, that the consumers understand, that everybody understands.
“You affirm that the applications of sensory analysis concern the world of producers, who are facilitated the path towards consumer tastes. Like all market phenomena, however, the individual is drowned in the average (this seems an existential discourse, I am aware of it, but it is the famous discourse on “medietà”, on conforming, which sociology deals with).”
Companies use sensory internally for consistency. The value of having consumer data is NOT to trend towards the mean, but rather to validate existing products and identify unoccupied niches – product spaces that would have consumer liking, but no products catering to those yet. It also helps with understanding the segmentation of consumers, as there is no universal preference beyond very basic minimum requirements. There is the unresolved, elusive topic of aesthetic/harmony etc. that we do seem to have an innate sense for, but attempts at exploiting that have generally failed due to turning out as mediocre or not distinctive…
The courage of “medietà'” could be called discernment. And it is for this reason that Cigar Sense personalizes the advice and does not publish generic judgments that speak to a wide range of consumers. In fact, how can you guarantee a consumer the pleasure from a product if you don’t listen to it, if you try to dictate what is good and what is not good according to your personal standards? This denies the variability of the personal tastes of each individual. Obviously there is always a certain level of conformity that arises, after all we do not live in airtight compartments, but we are social and intelligent animals. So if I didn’t like a cigar but I smoke it again and reconsider my position, I conform to those who love that cigar. What matters to me is that this update of my preferences is useful for making better and meaningful choices in respect of the fact I am an individual.
The purpose of all these methodologies is to recognize and improve outliers. A taster who reproducibly recognizes a unique attribute generates valuable information.
There is a reason why all unique, local, regional and protected origin products are pursuing similar paths to accumulate technical-analytical results and associate them with sensory studies: identify their uniqueness, their standards and codify / protect them.
“What do you think of the fact that learning the functioning of the senses, mental and social mechanisms can be useful for the individual to preserve his individuality, without becoming data to be mediated? I know, it’s a strong image. In the end, I believe, learning to taste, or to understand a cigar and judge, is a way to be more yourself and you can say: “I enjoy the cigar”.”
We agree with you on avoiding making the consumer a datum to be mediated. Truth is people want to be empowered, but need information, context, education, and just overall tools to become knowledgeable and capable/confident. That is why Cigar Sense offers training to all cigar lovers (consumers and professionals) who want to increase their awareness and recognition of the factors that influence their free choice, with the aim of optimizing the pleasure of the cigar.
If you really want to talk about data, there is a difference between big data and small / deep data. The latter offer a breath of fresh air when you want to do your own disintermediation, i.e. you want to move away from generic – (or manipulated by profit-motives) product finders of large e-commerce sites and social influencers – to connect producers and consumers in a personalized way (if you can sell me what I really like, I’ll be back tomorrow and buy more). In any case, here we are talking about a (still) little explored niche.
“If sensory analysis could be used in the tobacco world, it could only be for industrial products because they would provide an homogeneous product.”
While we appreciate the desire to preserve the “mystique” of the cigar, sensory exploration does not aim to discard it.
The same discussion must be had, and has been had, for all other emotional products: no two steaks are the same, even off the same cow, and no two batches of milk made into cheese on two subsequent days are technically the same, but all these industries have established internal parameters to control a basic character of the product. Even highly protected, unique foodstuffs are using sensory analysis to enhance the definition of what makes them unique in order to protect them.
Any cigar manufacturer, from the tabaquero to the factory manager, tells us their main goal is production consistency, because otherwise the mere notion of “marcas” and brand profiles would be completely meaningless from the onset. The same cigar vitola inside a brand will have some variation among samples, but the producer strives to preserve the core profile. Objective sensory analysis seeks to track this variability as part of the bigger picture of what a cigar is, and can only do so by having multiple people assess multiple samples of the same vitola. This also serves to illustrate and utilize the inherent differences of the taster – a basic vocabulary and training are necessary, but the unique subjective impressions should also be accounted for, and enhance our understanding at that point in time.
This will, by necessity, require of the taster to disregard – or at least consciously reflect on – the inherent emotional response of the experience, as this is deeply personal and cannot be reliably translated for others.
“Tobacco, given its natural state, can ferment in the box or in a conditioned place. Even after being harvested it can change its molecular structure until decay: this can happen even when temperature and URL are taken into account.”
On the subject of the variability through storage and aging, we are talking about a deeply beloved niche that is unique to a very small group of connoisseurs who find added enjoyment and value in this aspect. These are, by definition, NOT the consumers who seek guidance or advice and understanding through other people’s evaluations. Sensory science could provide a snapshot of even this elusive pursuit by assessing vintage cigars, or even doing consecutive tastings of the “same” vitola/marca over a span of production years. However, these would have become so variable that the experience could not be reproduced. In that same vein, most readers of a “100 point” rating for one vintage cigar will hopefully assume that a cigar of the same age will/may not provide that very same, sublime experience.
It is like the average wine consumer vs. the collector: every day consumers of wine usually buy a bottle off the shelf and will consume it at home that same night. They will usually be looking for guidance on certain characteristics they enjoy, without any interest in how a collection of that wine will taste with age. You can also see bottle to bottle variation within the same case of wine, and packaging also imparts differences, which itself is the amalgam of millions of individual grapes, which were all slightly different from each other, and then turned into a dozen or so individual barrels that each were noticeably different from each other before being blended to make a final product.
Either we deem the entire pursuit of understanding the big picture of experiences obsolete, or we recognize that a large part of that spectrum may be tackled for some degree of understanding – even just the understanding of differences, or the ultimate admission of science – “we don’t know exactly”. It is up to the individual and the individual project to determine what is worthwhile, but the fact of the matter is that there is a large need of consumers seeking insight and guidance to their pursuit of pleasure – to then freely explore all the fascinating fringes that we cannot representatively grasp.
“Recent studies on the molecular orientation of the smoke produced during smoking have shown that a molecule can, oriented in one direction, cause fruity-floral olfactory complexes and specifically apple notes , plum, rose, raisin, tea, black currant and tobacco; if oriented differently, it will generate wood, cedar, violet-raspberry. That is, an extra puff and the increase in temperature generate different perceptions. How do you code all this?”
Chemical analysis helps detect the presence of components in a matrix. You can chemically analyze individual leaves across the different manufacturing processes up to the point in which they are rolled together into a cigar and later. But the machines cannot tell, in a human understandable language, what the final product provides from a sensory perspective. You need sensory analysis to translate the expression of the molecules into a language that other humans can understand.
Also, it’s not just the orientation of the smoke that changes the molecular structure. In tobacco blends, there are molecules that have the ability to inhibit others, and if they are together, they will either not be synced distinctively, not in another way, but only one of the two will be sensed. These are all the games that make sensory analysis challenging and very interesting at the same time.
“On one hand, there is the scientific measurement of sensations, and on the other hand there’s the necessity to believe that human perceptions work like a machine with always the same type of stimulus and the ability to not be influenced by the outer world. Can anything like this work with a cigar?”
As mentioned in our article and in our A new profession in the cigar industry, you need to include training and statistics as key tools. You need to reduce the gaps among tasters in terms of stimuli, sensations, and perceptions and training needs to focus on that. You need to make the panelists aware of all the influences and biases they can be subject to when preparing for and tasting a cigar. And at the same time, physical and mental processing abilities are verified through the data.
At Cigar Sense we measure the success of our method by asking our members. How many of the cigars that we recommended you did you smoke and did you like? We do not claim to be a flawless predictor of what any given consumer will experience when smoking any particular cigar. There is no flawless predictor, for all the reasons we’ve observed above and in our reference article. However, we do claim to be the most reliable predictor thanks to our process and to our trained panel. And our consumer feedback bears this out as we have always maintained a 90%+ consumer satisfaction feedback.
Obviously this comes from our (thousands of) members who regularly use our service. We want to keep improving and we value any observation and suggestion from our members and whoever is familiar with and uses our service.
Franca Comparetto and Constantin Heitkamp (sensory scientist)
You can also listen to our related podcast episode.