Has there really been recent premiumization in the wine industry?

I don’t necessarily claim to know much about the wine industry, but I do know something about numbers; and sometimes I wonder about the numbers discussed in the wine industry. Indeed, it seems to me that the numbers I am presented do not always lead to the conclusion that is also being presented.

One such example is the oft-repeated claim in recent years about the so-called premiumization among modern wine drinkers. For example, we have headlines such as this: “Consumers in established markets are consuming less frequently, but pay more per bottle.” This claim comes from noting that the average price of wine bottles has increased recently, while the number of drinkers has decreased.

Premiumization of everything

This claim is apparently worldwide (see, for example: The US Wine Industry in 2019; UK shoppers trade up to higher priced wine; Proportion of Australians drinking alcohol down). It is an important conclusion for the wine industry:

Every number, every study, every expert and every trade group says the same thing: premiumisation has become a part of the wine business, and all will benefit as the consumer spends more money for a bottle of wine.

However, in all cases the conclusions are based on summary data, which can (unfortunately) hide what is really going on. Sometimes, we need to dig deeper. In the case of premiumization, we readers are not given enough information to do so.

Background

Many of the the problems that I see refer to the word “average”. This is not a reference to anything that might exist, such as an “average person”. Oh no — this refers to things like an “average price”, which does not exist. The price of each bottle exists, especially if you are the one paying for it, but the combined price does not.

Averages are figments of the imaginations of mathematicians. You take a bunch of numbers that do exist (eg. prices), and perform some mathematical calculation, in order to produce another number (eg. the average), which does not exist in the real world.

The reason I am intrigued by averages is that they represent patterns in data that can be created in many different ways; and knowing about the average does not necessarily tell you about the underlying patterns and causes. After all, if there are two patterns in my data, one going up through time and one one going down, then the average pattern will appear to not change at all (ie. the two patterns will cancel each other out). I might think that “nothing is happening”, when in fact two things are happening, both of which it might be important for me to know about.

So, let’s look at our specific example: premiumization.

Does premiumization really mean that drinkers are moving up-market?

It seems to me that there are at least four different ways that the average price of wines might increase through time, some of which have nothing to do with premiumization. These are illustrated in the four graphs here. In each case, we are shown how many drinkers there are within each of two wine-price categories: those who drink Everyday wines, and those who drink Premium wines. These groups change in size through time across the graph, sometimes increasing, sometimes decreasing, and sometimes staying the same.

The first graph illustrates what we might call “True premiumization”. Here, the number of people drinking Everyday wines decreases while those drinking Premium wines increase. That is, the Everyday drinkers have become Premium drinkers. This is the interpretation that we are currently being given.

The second possibility, as illustrated in the second graph, is which the Premium group increases in size but not the Everyday group. That is, some of the Everyday drinkers have become Premium drinkers, while new drinkers have moved into the Everyday-drinking group. This might be called “Semi-premiumization”, because there is less movement of people but more overall drinkers.

Another possibility is simply “Inflation” of wine prices. Here, wine drinkers keep drinking the same types of wine they always have, but the prices of those wines go up. After all, wine prices do not usually go down, not even when the grape quality varies from vintage to vintage (which the price used to do, once upon a time).

Finally, there might be a net loss of people as wine drinkers. If this happens, it is likely to be those who drink Everyday wines who depart, not the Premium drinkers. After all, at heart wine is simply one type of flavored alcohol, and there are plenty of alternative, and often cheaper, ways of getting that. One can buy high quality craft beers and ciders for much less than the cost of wine. One can also buy flavored coolers, and even flavored spirits, although the latter are not necessarily cheaper than wine. These days, the grape-growing industry is also competing with cannabis producers. There are thus many reasons why the pool of wine drinkers might decrease through time.

So, that makes at least four ways that we could observe an increase in average wine prices through time. We might call all of them “premiumization”, but I think that this would be very misleading. The last two possibilities, in particular, seem to have nothing whatever to do with wine drinkers moving up-market. Indeed, the response of the wine industry should probably be very different in these two cases.

Conclusion

So, which of these four possibilities are we being shown when the media discuss wine drinkers moving up-market? We need more data in order to decide.

Jeff Siegel has also high-lighted this issue (The downside of premiumisation), mainly in the context that premiumization may be nothing more than the third and fourth options listed above. If this is so, then the industry is looking at the future through rose-tinted spectacles. He concludes:

Given all of this, shouldn’t it be time for the industry to put an end to premiumisation? If wine is in a fight for its future, shouldn’t it focus on selling well-made and affordable products in response to the competition from craft beer, spirits and all those hard ciders and seltzers? ... So expecting an end to premiumisation because it’s a dead end once the last Baby Boomer dies is probably hoping for too much.