Products change over time. Those subtle, gradual changes that occur in a product’s sensory attributes are known as flavor drift.
Flavor drift often occurs after a producer makes slight, undistinguishable changes to a product. This new product eventually undergoes another slight change, creating a third version of the product that seems undistinguishable from the second product but is distinguishable from the original. While a number of slight changes can go unnoticed, they can eventually lead to multiple products that appear to be the same but taste drastically different from one another.
A good illustration of this in the spirits industry is the difference in flavors between bottles of bourbon produced in the early 1900s and those produced today. Although the processes and products are ostensibly the same, the differences in aroma and taste are undeniable.
Almost all food and beverage products experience some form of flavor drift during their lifetimes, and the phenomenon can be a major concern for producers. That’s because flavor drift has serious effects on consumer trust and confidence, causing previously popular products to lose sales. In some extreme cases, producers may need to declare recalls because their products are no longer recognizable.
This is why it’s important for distillers and producers of all sizes to understand what flavor drift is and how to prevent it.
[PAYWALL]
Creating Baselines
The first step that a distiller needs to take when trying to detect flavor drift is to establish their products’ baselines. A product baseline is a detailed description of what a product is, including all attributes both acceptable and unacceptable.
When creating baselines, it’s important to get as detailed as possible because those baselines are how you’ll evaluate batches of a product. For example: If you want to create a baseline aroma description for corn whiskey and you simply describe the aroma as “corn and oak,” that is not a detailed baseline. Corn can have many different aromas, such as cooked, dried, or green; likewise, oak aromas can be described as new-cut, toasted, or even charred. Therefore, simply describing a whiskey as smelling of “corn and oak” does not give an evaluator enough information to accurately determine whether a given batch has the correct aroma descriptors.
Instead, a good baseline aroma description for such a product might read “cooked sweet corn top-note, with fresh-cut oak logs accompanying.” This descriptor could also be associated with an intensity rating as well as a brief list of associated aromatic chemicals, such as dimethyl sulfide.
That example may be a little exaggerated, but the point is that baseline descriptors should be precise. Otherwise, evaluators may have trouble accurately detecting flavor drift.
The best way to produce detailed baseline descriptions is through a combination of descriptive analysis and analytical testing. Using a trained descriptive-analysis panel and then comparing their results to that of gas chromatography (GC) and high-performance liquid chromatography (HPLC) testing can help distillers formulate a unique product “fingerprint.”
You can then use that fingerprint to create a product baseline that evaluators easily understand.
Detecting Flavor Drift
After creating an acceptable product baseline, the next step is to begin determining whether your products have undergone, or are undergoing, any flavor drift.
You can accomplish this by simply asking a group of trained evaluators whether a new batch of your product matches your baseline descriptors. That’s an easy and moderately effective way to catch flavor drift—however, it doesn’t provide much information about the flavors causing the drift.
For a distiller to precisely measure and understand what’s causing flavor drift, the first thing to do is determine what exactly the differences are between the baseline product and subsequent batches. The best method of doing this differs among products, but it normally involves a combination of descriptive and discriminative testing. For example: A distiller could first use a discriminative test, such as the triangle test, to determine which batches have a noticeable difference from the baseline. They would then undertake descriptive testing on the different batches to define the exact differences. This gives distillers enough information to begin formulating a plan.
Once you know the differences among your products, it’s next important to establish the scale of those differences. You can do this with a form of descriptive testing known as ranking testing. Ranking testing involves having evaluators taste a series of known samples that have been spiked with an increasing amount of the flavor in question. They then taste an unknown sample and describe it as either higher or lower than each of the known samples. This testing should include both the baseline product and the batches that have been identified as different. The results of these tests allow distillers to understand the exact scale of the differing flavor in each sample and to identify what actions may be needed to fix the problem.
It may be necessary to repeat the ranking tests multiple times, depending on the number of different flavors you’re trying to identify and the level of accuracy you need.
Finally, you’ll need to analyze the results from the above tests, along with any other data, to try to accurately quantify the drift that has occurred. The preferred method of analyzing the collected sensory data is up to personal preference, but common statistical methods include Analysis of Variance (ANOVA) and Principle Component Analysis (PCA).
These analyses are complicated and usually require specialized computer programs and expert training to conduct them correctly. However, when used effectively, they can give distillers important information about the kind of flavor drift occurring and how they can fix it.
(Note: While analytical testing can help detect flavor drift, it is generally considered less effective than sensory testing because of the synergistic flavor effects of many compounds. Also note that hedonic testing should not be used in the process of detecting flavor drift—that type of testing is only designed to measure a consumer’s liking of a product, not any specific sensory aspects.)
Factors that Can Cause Flavor Drift
Once you determine that flavor drift has occurred in a product, you’ll want to ascertain the cause.
Unfortunately, a myriad of factors can cause flavor drift, and many of those may be out of the distiller’s control. It also tends to be difficult to nail down any one exact cause. Nevertheless, it can be helpful for distillers to know and monitor some of the most common causes of flavor change.
The first common source of flavor change in spirits—and perhaps the most controllable one—is alteration in how a spirit is made. The process of making distilled spirits is a constantly evolving science. It seems almost daily, for example, that companies are developing new enzymes or methods to increase efficiency and improve sustainability. That’s led to large transformations in process for many distillers.
An example of this is the relatively recent introduction of heat-stable alpha-amylase and glucoamylase enzymes, which have allowed distillers to shorten—or in some cases entirely skip—the once-vital saccharification step during mashing. Although that can speed up a distiller’s workday, there are often more enzymes at work than just amylases during the saccharification step. Many distillers have found that shortening or eliminating their “sach rests” with these enzymes has caused noticeable changes in the flavors of their finished products. What may seem like a small change to the distiller may represent a serious change to the overall mashing process and chemistry.
Another example of how small process changes can affect spirit flavor involves barrel storage. In the past, almost all barrels were stored on their sides in large rickhouses or rackhouses. This was because distillers did not have ready access to heavy machinery, and the easiest way to move barrels around was by rolling them. In recent years, though, forklifts have become more accessible and widespread, allowing distillers to begin storing barrels standing upright on pallets. That’s good news for many distillers because palletized barrels take up less space and are safer to handle—but there’s a trade-off. Barrels that are stored standing up experience different environmental conditions than those stored horizontally. Those differences cause them to taste slightly different, illustrating why it’s important to be aware of any and all process changes—no matter how small. Something as simple as changing the orientation of a cask may have a significant effect on the final flavor of a product.
A second common (and somewhat less avoidable) source of flavor change is alterations to process equipment. Equipment breakdown is an unavoidable part of operating a distillery; pump and mixer motors give out, heat exchangers become irreversibly damaged, and fermentors eventually succumb to general wear and tear. Replacing worn-out equipment is common, and it’s human nature to look for chances to upgrade or improve any gear being replaced. Unfortunately, that tendency can sometimes backfire—distillers often fail to account for the strong effect that equipment has on the final flavor of many products.
This idea is perhaps best understood in the context of stills. Stills, by their very nature, are unique. Each has its own pot geometry, heating coefficient, lyne-arm angle, and other eccentricities that can affect how the still runs and the resulting flavor of the distillate. If a distiller were to buy a new still that had drastically different characteristics from their previous one, it could have a serious effect on the flavor of their products. That’s why some distillers actually get 3D scans of their stills—so that when they order new ones, they can make sure that the manufacturers produce them with exactly the same characteristics; otherwise they risk changing their product.
A still is just one piece of equipment that can influence flavor. When replacing or upgrading equipment, it’s important to consider all possible consequences before trying something new.
Finally, another common source of flavor drift in spirits—and possibly the least controllable one—is change in raw-material inputs.
All spirits are made from agricultural products, which means they’re subject to the same pressures and seasonal variations to which crops are vulnerable. For instance, many barley-growing regions recently experienced an unexpectedly long heat wave coupled with an unprecedented drought. Those adverse conditions caused much of the barley grown that year to have higher protein content than normal. Besides the process issues that presented to distillers, it also resulted in slight changes in the flavor of new make spirit. To combat this, some distillers altered their collection points, while others changed their barrel entry proof; still others chose to do nothing at all, confident that they could blend out any differences once the product had been aged.
That sort of variation in spirit caused by incoming raw materials is incredibly common. Innovations in farming, changes in planted grain varieties, and even miniscule alterations in regional microclimates can all affect the quality and properties of raw materials received. Furthermore, the yeast that produce alcohol may react unpredictably to new environments and pressures.
Ultimately, distillers can’t take consistency for granted, and they should be prepared to adjust.
How to Fight Flavor Drift
In some ways, flavor drift is inevitable because human taste is an ever-moving target. However, there are ways to prevent and minimize it.
The first and foremost is to use regular and consistent testing. Any batch produced should undergo a minimum set of standard tests. Traditionally, that means discriminative testing; in some case, it can also include a descriptive test. A producer should also be using those tests on raw materials and any other items used to produce the spirits. That level of testing can help ensure that only the most representative products go to market.
Another way to forestall flavor drift is to consistently refresh evaluators’ and operators’ baseline palates using “gold standards.” Gold standards are internally approved samples of a product that have the ideal sensory characteristics of a specific product. If evaluators and operators can regularly sample gold standards, it helps them to more clearly define the acceptable parameters of the product in their minds. Then they can more readily identify when there are problems in a process or a current batch, lessening the chances of an inconsistent product.
Finally, a last way to beat flavor drift is join it—that is, lean into it with regular brand refreshes. Consumer preferences aren’t static, and it’s common for customers to change tastes over the years. Distillers who want to stay competitive can take advantage of flavor drift when it’s time to create a new product or line extension. In that sense, flavor drift can become an innovation that helps invigorate a brand whose products have become stale.
Ultimately, there are far too many factors involved in flavor drift to eliminate it completely—nor would you necessarily want to. Flavor drift represents an opportunity to learn more about what makes your spirits unique and how small adjustments can lead to larger changes.
No product should remain stagnant, lest it be left behind by an ever-evolving market. Nevertheless, a product should remain consistent and true to the distiller’s original intentions. Establishing good baseline descriptors and always monitoring for flavor drift can help you preserve that consistency.
After all, no distiller ever wants to hear a customer utter the words, “It used to be so good, but then it changed.”
