Millenials do not exist: so why market to them?

We often read in the wine press about the (supposed) difficulty of marketing wine to Millenials (eg. The maddening business of marketing to Millennials; Are Millennials a threat to fine wines?). However, the main reason Millenials are hard to market to is because they don't exist. The idea of "generations" is a marketing concept — there is no such thing as a discrete generation, because human reproduction is continuous, and thus all generations overlap — their distinction as groups is arbitrary.

Nevertheless, marketers often recognize cohorts of people who behave in a similar manner, because they can then market to the group rather than to the individuals. This does not change the fundamental idea that cohorts do not exist in reality, because generations overlap. We may thus be making life difficult for ourselves, by trying to do something that is biologically illogical. The groupings are nothing more than a convenience, and the individuals could alternatively be grouped based on any criteria you care to choose (the pie can be sliced in many ways).

Where does the concept of discrete cohorts come from?

By definition, my daughter is a different generation to myself, as are my parents. However, this does not necessarily make any of us part of a cohort (a group of people with a shared characteristic), just because we were born within an arbitrary set of years. So, strictly speaking, we need to distinguish generations from cohorts — the recognition of "generations" for marketing purposes should actually refer to "cohorts", and I will use this word below.

You can read about the conceptual origins of cohorts and generations at Wine Opinions. There are two organizations who seem to be responsible for defining cohorts – the editors of the book American Generations: Who They Are and How They Live (currently in its 9th edition), and the Pew Research Center.

The cohorts, believe it or not, are based on changes in birth rate in the USA (see the discussion at Defining generations: where Millennials end and Generation Z begins). For American Generations the cohorts look like this:

1933-1945 1946-1964 1965-1976 1977-1994 1995-2009

Silent Generation Baby Boomer Generation X (originally called Baby Bust) Millenial (originally called Generation Y) iGeneration (originally called Post-Millenial or Generation Z)

There was an increase in the U.S. birth rate immediately after WWII (ie. a boom), followed by a decrease for a decade, followed by another increase for a decade and a half, etc. The changes in rate are assigned to arbitrary years, which are then used to group the generations into cohorts.

So, the definitions come from the reproductive behavior of the parents. What this has to do with the consumer behavior of the children is unclear — attitudes are formed by your upbringing, sure, but not by the birth rates of your parents. As Wine Opinions puts it: "While generational comparisons and analysis are often useful in consumer research, these broad age segments are not always key drivers of behavior." There may well be better ways to group consumers than this, in terms of marketing to them.

I say this in spite of the fact that I am obviously a Baby Boomer, my parents were classic Silent Generation people, and my daughter is an archetypal Millenial. My daughter does, indeed, explicitly reject what her parents did — that's why her life is so different from mine. However, I do not recall, growing up in the 1970s, that I rejected my parents' goals — that was the 1960s not the 1970s. By the 1970s, young people had gotten over that, and instead concentrated on the idea that life was fun (disco music, flared pants, platform shoes, etc), rather than being for contemplating your own navel and the seriousness of the world — sadly, the latter idea returned in the 1980s and has not left us since.

Fuzzy groups

The idea of breaking continuous variation (eg. generations) into groups (eg. cohorts) is not new. Indeed, we have always done it with people, for example, by recognizing different "races" — this idea was originally related to genetic variation (not assumed cultural superiority!). This genetic idea may once have worked, because most people stayed in the same place, and thus genetic differences arose in different geographical locations. However, in the modern world, with mass transportation, people are much more genetically mixed than they have ever been. "Races" was once a practical genetic concept, but not any more.

At best, we are now dealing with what are called "fuzzy groups" — there can be a core part of each group that is distinct from other groups, but at the boundaries the groups grade into each other.

The difference between groups and fuzzy groups can easily be seen by considering geography. For the USA, recognizing an east coast and a west coast makes perfect sense, because the distinction is not geographically arbitrary — these are not fuzzy groups. However, dividing either of these two coasts into a north coast versus a south coast would be completely arbitrary — there is no obvious (non-arbitrary) geographical point where the distinction might be placed. At best, we would be dealing with fuzzy groups.

Clearly, cohorts are fuzzy groups. A person born in 1965 may well be very different from a person born in 1995, but a person born in 1995 (Post-Millenial) is not likely to be different from one born in 1994 (Millenial).

I have written before about the inherent dangers of putting people into categories in the wine business, which is often done for Big Data (because it is otherwise hard to deal with): The dangers of over-interpreting Big Data. Marketers should be wary of generalizations, even when they are of practical use.

Confirmation bias

This is another problem. If we pre-define a set of groups and then compare those groups we will surely find differences between them. However, this is called "confirmation bias" — we use our data to confirm the existence of the groups that we pre-defined.

Consider the following graph, which shows Wine Purchasing Decision Cues by Generation. There seem to be differences among the cohorts for most of these bases for deciding which wine to buy; and they are all statistically significant (based on contingency chi-square tests).

However, this is simply confirmation bias. We took 109 people, with continuous generations, divided them into three cohorts based on age, and now we are showing that these cohorts differ. However, we already know that people of different ages behave differently, and this is all we have shown from our data. We have not demonstrated that there are three cohorts, because we pre-defined them before we started. The cohort "Millenials" exists in the graph solely because we put it there, not because it is necessarily a group of people who have similar behavior.

Marketing

Marketing is the term used for the process of telling consumers to buy something that they didn’t know they wanted. If they already know they want it, then you don’t have to market it to them — they will buy it, you don't have to sell it.

A classic example is the Yugo car, which was considered hard to market because it did the only thing that cars were originally designed for — transporting people from A to B. It did almost nothing else, but it fulfilled its design purpose cheaply — that is why it had the longevity it did (selling 800,000 cars in the USA from 1978 to 2008). However, it was persistently mocked because it didn’t do any of the other things that marketers had persuaded the car-buying public are essential — it is the available options and the perceived image that sell cars, not the transportation activity.

So, perhaps it is time to stop trying to find out how to market wine to Millenials. Marketing is about the packaging and the presentation (telling a story), not the contents. There is little point to learning that Millenials like unique packaging, or that they use wine to relax, etc. If they are interested in the contents of the packaging, then sell them the wine, don't try to market the packaging to a group that does not exist.

Conclusion

Looking for time trends in the wine industry is a useful activity, especially possible future trends, because it helps all of the people who rely on the wine industry for their livelihood. Using confirmation bias to categorize fuzzy groups, on the other hand, will get us nowhere. It seems that marketing to Millenials might be an example of the latter activity.

---

Selected literature for wine generations / cohorts

Marianne McGarry Wolf, Scott Carpenter, Eivis Qenani-Petrela (2005) A comparison of X, Y, and Boomer generation wine consumers in California. Journal of Food Distribution Research 36:186-191.

Janeen E. Olsen , Liz Thach, Linda Nowak (2007) Wine for my generation: exploring how US wine consumers are socialized to wine. Journal of Wine Research 18:1-18.

Joanne Teagle, Simone Mueller and Larry Lockshin (2010) How do millennials’ wine attitudes and behaviour differ from other generations? 5th International Conference of the Academy of Wine Business Research.

Polymeros Chrysochou, Athanasios Krystallis, Ana Mocanu, Rachel Leigh Lewis (2012) Generation Y preferences for wine: an exploratory study of the US market applying the best-worst scaling. British Food Journal 114:516-528.

Jasmine B. MacDonald, Anthony J. Saliba, Johan Bruwer (2013) Wine choice and drivers of consumption explored in relation to generational cohorts and methodology. Journal of Retailing and Consumer Services 20:349–357.

Catalina Chivu-Draghia, Arina Oana Antoce (2016) Understanding consumer preferences for wine: a comparison between Millennials and Generation X. Scientific Papers Series Management, Economic Engineering in Agriculture and Rural Development 16:75-83.

Madeleine Rose (2017) The difference in wine purchasing behavior between generational cohorts. Report.

Madeleine Rose (2018) New research study highlights motivational differences of wine consumers by generation. Report.

Marianne McGarry Wolf, Lindsey M. Higgins, Mitchell J. Wolf, Eivis Qenani (2018) Do generations matter for wine segmentation? Journal of Wine Research 29:177-189.

Prohibition still exists in the USA

It seems to surprise many Americans that there are still prominent remnants of Prohibition throughout the U.S.A. This is not "Neo-Prohibition", as some Americans like to call it — these are the remnants of the same old Prohibition from the early 20th century.

This situation arises from the fact that the USA differs from every other country on the planet, in being a country created from a collection of states rather than a country divided into states. Everywhere else, the national government has all of the power, and it delegates some things to the states or provinces (if they exist), which in turn further delegate some things to local administrative areas. In the USA, on the other hand, the states retain most of the political power, except for those things they have delegated to the national government (pertaining principally to international relations), or which they have delegated to local administrative areas (ie. counties).

Pretty much the only way for the national government to regulate the states is by changing the U.S. Constitution (or, to be precise, by adding things to it); and they also have a lot of power over what happens with national taxes, too.

This creates the unusual situation of the United States being distinctly non-united on a number of topics. Some of these relate to alcohol, since alcohol laws are currently a state matter. For example, there is much current discussion about that fact that some states allow the sale of alcoholic beverages by retailers based in other states, and some do not — that is, inter-state shipping is prohibited by some states. Here, the European Union is much more united than the United States, because any retailer can send anything from anywhere in the EU to anywhere else in the EU, as this is a basic tenet of the original creation of the Union — there is free trade.

However, the USA once went much further than this, and decided nationally to ban the sale of all alcohol-containing products, except those intended for medical or sacramental purposes — this was done via the 18th Amendment to the Constitution. Most Americans think that this situation was changed again in 1933, when the 21st Amendment repealed the 18th Amendment. However, this is not quite so, nationally. There are still prominent remnants of Prohibition throughout the country, with complex state-by-state differences in the status of alcohol availability.

Since this seems to surprise many Americans, it worthwhile to summarize the situation here.

When was Prohibition repealed?

Section 1 of the 21st Amendment (added to the U.S. Constitution in 1933) explicitly repealed the 18th Amendment. However, Section 2 banned the importation of alcohol into states and territories that have laws prohibiting the importation or consumption of alcohol. This means that each state was allowed, in practice, to decide for itself whether Prohibition should be repealed. Prohibition had been an admitted failure, in terms of addressing alcohol consumption, and so most states simply allowed alcohol sales as from 1933. However, the following states did not do so until later:

Texas North Carolina Tennessee Kansas Oklahoma Mississippi

1935 1937 1939 1948 1959 1966

So, there are currently no states that completely ban the sale of alcohol (but there were when I was born!).

Alcoholic beverage control states

Since individual states retain the right to regulate alcohol as they see fit, there is what has been called "a dazzling array of confusing alcohol control laws under seemingly arbitrary regulatory agencies."

At the moment, there are 17 U.S. states that currently control the sale of alcoholic beverages to one extent or another. They are shown in the following map (from Wikipedia).

Green: State control of beer, wine, and spirits

Utah — state-run stores (aka State Liquor & Wine Stores) for sale of all alcohol over 4% ABV

Dark blue: State control of wine and spirits

Pennsylvania – Pennsylvania Liquor Control Board stores (aka Fine Wine & Good Spirits) control the sale of spirits, while wine sales licensees are limited based on county population sizes

Light blue: State control of spirits

Alabama – liquor stores are state-run or on-premises establishments with a special off-premises license
Idaho – state monopoly over sales of beverages with >16% ABV
Iowa – spirits are sold to privately owned retailers by the Iowa Alcoholic Beverages Division
Maine – spirits sold in state contracted liquor stores
Michigan – state monopoly over wholesaling of distilled spirits
Mississippi – spirits sold in state contracted liquor stores
Montana – spirits sold in state contracted liquor stores
New Hampshire – spirits sold in state-run liquor stores
North Carolina – state Alcoholic Beverage Control Commission controls wholesale distribution and oversees local ABC boards, which own the liquor stores for sale of spirits
Ohio – state contracts with private businesses to sell spirituous liquor
Oregon – spirits sold in stores operated and managed by state appointed liquor agents
Vermont – spirits sold in state contracted liquor stores
Virginia – Virginia Alcoholic Beverage Control stores control sale of spirits
West Virginia – state monopoly over wholesaling of distilled spirits
Wyoming – state monopoly on wholesale importation

Orange: Other alcoholic beverage control states

Maryland — by state law, Montgomery County operates under a control model.

Alcoholic beverage control counties

Alcohol regulation occurs even more commonly at the level of counties, which is what creates the "dazzling array" referred to above. There are currently at least 200 counties nationally that remain Dry, in that there is some ban or other on the sale of alcohol. Most of these are in the east and south-east, as shown in the following map (from Wikipedia).

Blue: Wet
Red: Dry
Yellow: mixed (according to city or alcohol type).

This example from among the Yellow counties is from the American Association of Wine Economists' Facebook page:

As of 2013, there were eight completely dry towns in Massachusetts: Alford, Chilmark, Dunstable, Gosnold, Hawley, Montgomery, Mount Washington, and Westhampton. Plus, there are many partially dry (moist) towns. For instance, in Rockport, alcoholic beverages may only be served to patrons who are consuming a full meal.

Import of alcohol from out-of-state retailers

Some states currently ban alcohol import direct from from out-of-state retailers. This is the so-called Three Tier system, which was one of the practical outcomes of Section 2 of the 21st Amendment — states control alcohol by imposing a middle tier for alcohol movement, separating the customers from the source. That is, a local distributor must import the out-of-state alcohol, and pass it on to local retailers, who are the only legal suppliers to local customers.

At the moment, it seems that inter-state courier companies are delivering wine only to the following territories:

Alaska
California
Idaho
Louisiana
Missouri
Nebraska
Nevada
New Hampshire
New Mexico
North Dakota
Oregon
Virginia
Washington D.C.
West Virginia
Wyoming

According to the National Conference of State Legislatures, as of 2016:

Alabama, Oklahoma, Utah — specifically prohibit the direct shipment of alcoholic beverages to consumers
Mississippi — does not have a statute that specifies that direct shipments are allowed
Rhode Island — allows intoxicating beverages to be shipped when purchased on-site
Delaware, Massachusetts, Montana, North Dakota, Ohio, Oregon, Vermont, Virginia — allow the direct shipment of beer and wine, as specified
Arizona, Florida, Hawaii, Nebraska, New Hampshire, District of Columbia — allow direct shipment of all spirits, as specified.

The remaining states allow direct wine shipments, to one extent or another. For example, Tennessee has laws requiring liquor store owners to be TN residents for two years, which is currently being challenged in court (Wine shipping gets its day in court).

Many states in the U.S.A. now allow alcohol import direct from U.S. wineries, although for most of these states this is only a recent phenomenon. For example, Kentucky is the most recent (2019) state to pass a bill to allow inter-state shipping (Kentucky opens doors to wineries).

Note: Pennsylvania (2005) and Massachusetts (2010) have had statutes ruled unconstitutional by state courts. Delaware has a statutory provision that requires orders to be processed and shipped through licensed wholesalers.

Update:  Tom Wark has recently tackled the same topic: The state of discrimination in wine.

---

This is just like Australia in the 1950s and early 1960s. Even as late as the 1970s, alcohol could be served on Sundays only to travelers — that is, a public house (ie. a pub) in the countryside could serve alcohol but not one in a town or city. I lived on the outskirts of a big city, so my mates and I only had a short country drive to get to the nearest pub. Current licensing is much more lenient.

Australia's biggest wine companies

We often read in the media about a particular company being "the biggest" or perhaps "second biggest", often without specifying the criteria. Which wine companies are Australia's largest depends very much on which criteria you use — the biggest sales, for example, do not necessarily generate the biggest revenue.

According to Winetitles Media, there were 2,257 wine produces in Australia in 2018, with a 7.7% drop from 2017 in the number of small producers (crushing less than 100 tonnes). However, the top 5 producers accounted for an estimated 87% of the total export volume, so they may be the only ones you have ever heard of. Circa two-thirds of Australian wineries produced fewer than 5,000 cases, which means that direct-to-customer sales are important for the majority of wine businesses — 90% of wineries have a cellar door, accounting for half of all direct sales (Cellar door survey shines light on direct sales opportunities for wineries).

Various features of these 2,257 producers are summarized on the page The wine producers in numbers 2012-2018, including what types of wine are made, usage of social media, and export destinations. However, I will focus here on the biggest companies only.

The table below lists the top group of companies, showing their size ranking for each of five wine-producing characteristics. The data are taken from Winetitles Media, which lists only the top 20 companies for each of the characteristics. Therefore, data are not listed for a company if they do not make it into the top 20 for any particular characteristic.

Treasury Wine Estates Accolade Wines Casella Family Brands Pernod Ricard Winemakers Australian Vintage De Bortoli Wines Warburn Estate McWilliam’s Wines Group Kingston Estate Yalumba Wine Company Zilzie Wines Andrew Peace Wines Brown Family Wine Group Qualia Wine Services Idyll Wine Co. Angove Family Winemakers Berton Vineyards Calabria Family Wines Tahbilk Group

Total revenue 1 3 4 2 5 7 8 6 12 10 15 17 9 14 18 13 19 20 11

Sales of branded wine 2 1 3 4 5 6 7 9 16 11 19 8 13 10 15 12 17 14

Total wine production 3 1 2 4 5 7 10 11 6 13 9 12 19 8 14 17 16 15 20

Wine-grape intake 3 1 2 4 5 7 11 9 6 13 10 12 16 8 15 17 19 14 20

Vineyard area 1 10 2 6 3 8 7 11 4 12 15 16 13 14 17 19

The conglomerates

Treasury Wine Estates (TWE) has one of the most complex, and embarrassing, histories of any Australian company (see Southcorp Limited history). It was founded as a brewing company, which started acquiring wineries in the 1980s, followed by other businesses not related to alcohol. It then sold off the original brewing business (to another brewer), and changed its name to Southcorp. It then sold off the non-alcohol businesses, and focused on wine. In 2001 it merged with Rosemount Estates, at the time Australia's most profitable wine company. This was a reverse takeover — Southcorp paid the money but the Rosemount executives ran the new company. It was all downhill from there, as every decision to rationalize the joint company turned out to be wrong, leading to record financial losses — it is a case study in dysfunctional decision-making (see the book Contemporary Issues in Management and Organisational Behaviour). (Trivia: one of the decisions was to sack one of the two US distributors, who immediately joined forces with Casella Wines to make Yellow Tail the no. 1 imported US wine brand). The remnants of Southcorp were bought in 2005 by a brewing conglomerate, Foster's Group, who tried to run it like a beer business. Beer and wine do not mix, because their modern business models are radically different; so Foster's eventually spun TWE off as a separate company in 2011.

This company was pretty much non-functional because of all of the historical baggage — it has taken until very recently for a turn-around, having discovered the benefits of selling premium wine rather than mass-market wine (TWE delivers strongest organic growth rate in net sales revenue in its history). TWE's recent sparkling results have been boosted by exports to China (Treasury sees ‘tremendous opportunity’ in China to continue to grow), although it is not clear what the future of the Chinese market will be (China takes a shine to Aussie wine). TWE also has wine activities in California and New Zealand, and has announced plans to invest in French wineries (Penfolds owner to buy French wineries to satisfy Chinese tastes). They own Australia's 4th largest winery (Lindemans Karadoc Winery, processing 110,000 tonnes per year), which was the first one to be modeled on an oil refinery (see picture below). Their best-selling brand is actually Beringer, from the USA (9th in the world); but their most famous Australian brand is Penfolds, among many, many others.

Accolade Wines is now owned by the private equity giant The Carlyle Group, having recently been bought from another equity firm, CHAMP, plus Constellation Brands in the USA (the world's 2nd largest wine company, which is itself embarking on a premiumization strategy: Constellation brands to discontinue 40% of wine & spirits portfolio). Accolade is a classic example of the aspiration to become a virtual wine company, one that manages the sales and marketing of consumer brands but no longer owns many vineyards (similar to Constellation: Why you don't need land for a winery) — it has been steadily selling most of the ones that came with its various winery acquisitions. It has also recently announced a major restructuring plan (Carlyle Group takes knife to Accolade Wines), including outsourcing much of the administration. It recently acquired the head of TWE's Asian operations (Treasury Wine's top Asia executive departs to rival Accolade Wines), which probably indicates a new direction.

Accolade owns Australia's 2nd largest winery (the Berri Estates Winery, processing 230,000 tonnes), so it does actually make wine — indeed, it is Australia's largest producer, as well as having the greatest grape intake. Accolade has a big portfolio of premium Australian brands, but their best-selling brand is Hardys (8th best-selling wine brand in the world). It also has brands from Chile, Italy, New Zealand, North America, South Africa and the United Kingdom.

Pernod Ricard has wineries in Argentina, California, France, Georgia, Mexico, New Zealand, Portugal, South Africa and Spain, so Australia may not play a big part in its global plans. However, its main brand from Australia is Jacob's Creek, the no. 1 bottled wine brand in Australia, both by volume and by value. It was recently ranked Australia’s 8th strongest brand (Jacob’s Creek ranked in Australia’s top ten strongest brands 2019), the only wine brand to make it into the top 10. Oddly, when Pernod Ricard decided to re-brand all of its Orlando branded wines, it changed all of the premium wine names to Jacob's Creek, which was its mass-market brand name — one does not normally replace a higher-status name with a lower-status one. The company has recently announced that it will be using 100% renewable electricity in all Australian sites by mid-2019.; and it is reported to currently be considering the sale of its wine division (Pernod considers offloading $500 million wine unit).

Australian Vintage is the descendant of a conglomeration formed from a merger of McGuigan Wines and Simeon Wines. They own a lot of vineyards, but don't really have any premium brands to speak of — they target the lower-middle market at best. Their current financial report is optimistic (AVL’s profits up but predicts vintage yields will drop). They own Australia's 3rd largest winery (Buronga Hill Winery, processing 150,000 tonnes per year); and they recently invested in a fancy new packaging facility (Australian Vintage invests AU$11m in packaging facility). They seem to be very proud to have been the first wine producer in Australia to run 90% of their operations on solar and wind power (Australian Vintage signs ‘landmark’ renewable energy deal).

The others

Both the Brown Family and Tahbilk specialize in premium wines — note that they have a lower ranking on intake and production compared to their higher position on revenue. They are both still family owned, and have been for a very long time (see Keeping the family wine business is often hard). The Tahbilk winery still has some of its original buildings, including its famous wooden tower, where tastings are held. I well remember one visit in which I had to cup the tasting glass in my hand for several minutes in order to warm the wine sufficiently to taste it! The Brown Family, on the other hand, have probably the biggest tasting room I have seen. They are located on the road to one of Australia's skiing fields, and the overlap between skiing enthusiasts and wine enthusiasts is large. This has allowed the winery to be one of Australia's greatest experimentalists, because they can try out new ideas on a large and inquisitive customer base.

Other big producers focus on bulk wine or cheap wine, including Warburn Estate, Kingston Estate, Zilzie, Andrew Peace, Calabria and Qualia Wines. They are located in the large inland Murray-Darling irrigation areas, originally populated by returning soldiers but thereafter dominated by southern European migrants — Warburn, Kingston, Zilzie, Andrew Peace and Calabria are all still family owned. Most Australian wine drinkers will never have heard of these companies, but if they drink inexpensive wine then they are very likely to have tasted their wines — Warburn, for example, owns the Coolabah bag-in-box label as well as the AC/DC celebrity brand. These wineries tend to be large exporters of unbranded wine.

Idyll Wines and Berton Vineyards did not start out in the irrigation areas, but have instead moved into them from the premium end. Idyll, like Qualia Wines, offers "end-to-end solutions" for customer own-brand development, as well as having brands of its own.

Both De Bortoli and Casella originally started out as immigrant bulk-wine producers from the irrigation areas — Casella owns Australia's biggest winery (at Yenda, processing 250,000 tonnes), while De Bortoli own the 5th largest (at Bilbul, 100,000 tonnes). However, both companies have been trying to break into the premium market over recent years. Both are still family owned, with De Bortoli being the older, being founded in 1928 compared to 1969 for Casella, which is much the larger company. De Bortoli came to fame as the result of its Noble One (Botrytis Semillon), still recognized as Australia's premier Sauternes-style wine; and its Black Noble rivals any Pedro Ximenez wine from Jerez. The company used this springboard to start buying wineries and vineyards in the premium areas of south-eastern Australia; and it is now the archetypical large family producer covering all brand segments.

I have discussed Casella before (Yellow Tail and Casella Wines). They have been using their rise to the top position for US wine imports (with Yellow Tails as the 5th best-selling wine brand in the world) to fund the acquisition of wineries and vineyards in the premium areas of south-eastern Australia. This is clearly (as it is for De Bortoli) a recognition that you don't survive long in the wine industry by selling bulk wine at slim profit margins, especially with the modern trend to premiumization among drinkers. However, Casella have also been trying to maintain their no. 1 position in the American market, by forestalling any drop in buyers. Their Super Bowl ads over the past three seasons were extremely expensive, and there is little evidence that they produced any increase in sales. The ads are, instead, all about maintaining "brand awareness ... in a declining category and outpacing all of our competitors" (Yellow Tail returns to Super Bowl for third year).

There are some very long-standing family firms in the table, including McWilliam’s, Yalumba and Angove (see Keeping the family wine business is often hard). These are the archetypal survivors: mid-sized family companies targeting the premium market, who have not succumbed to corporate takeovers. The latter is the fate of almost all mid-sized wine companies in Australia — once they attract attention, the corporate predators descend, they are bought, stripped of their assets, and their wine brands are devalued. I have spent my entire life as a wine consumer watching this happen time and time again to premium-wine companies — stay small or be eaten. One has to respect those families who have resisted this onslaught. [Note: McWilliam’s went into voluntary administration in January 2020.]

Other companies

There are also a couple of companies missing from the above list, but which have large vineyard areas.

Seppeltsfield Wines (9th largest vineyard area) may well be the first winery I ever visited, back in the late 1970s. The Seppelt brand is now owned by TWE, but the property was bought from TWE's predecessor a decade ago; and the new company has been been buying other premium vineyards, as well, some of them from TWE. My main memory of my visit is the impressive arrival along the Avenue of Palms (see the picture above) — a 5 km driving avenue of more than 2,000 Canary Island Date Palms, which were planted by Seppeltsfield workers during the Great Depression (to keep them employed).

Duxton Vineyards (5th largest vineyard area) is a massive consolidation of pre-existing vineyards in the inland irrigation areas (including two of Australia's largest vineyards), planning to produce 5% of Australia’s wine-grape harvest each year. It sells grapes, and has a few wine brands of its own, but it specializes in bulk juice and wine.

Finally, it is worth mentioning the rapidly increasing investments from China, both by buying established companies (Grape expectations: The Chinese investors buying into Barossa Valley wineries) and by establishing new vineyards (eg. the Weilong Wine Grape Company's activities in the irrigation areas).

Global bulk wine routes visualized

In a recent post I discussed the use of heat maps to display patterns of data that are often shown in a tabular format (US 2018 grape harvest prices visualised). In the current post I will discuss an alternative type of graph, one that shows flows of things from on place to another — the Sankey diagram. This has obvious relevance to the wine industry, in which grapes and wines move globally as well as nationally.

In this case I will look at the movement of bulk wine between countries. I have compared world-wide imports and exports in an earlier post (Where does all of this wine come from and go to?), but not shown the actual flows of the volumes.

The data come from Comtrade, the United Nations International Trade Statistics Database. I accessed the data available for 2017 in the category: "Wine; still, in containers holding more than 2 litres" (code 22042). This may include pretty much anything (bulk or otherwise), except import/export of single bottles of wine, but excludes sparkling or fortified wines.

For an example of a Sankey diagram, let's look at Spain, the world's largest bulk-wine exporter. Spain actually exported to 116 countries in 2017, according to the database, but I have shown only the top 15 countries in the figure.

The diagram shows flows from left to right, so in this case the exporting country is on the left and the importing countries on the right. There is one line connecting to each importer, with the thickness of the line indicating the relative volume of wine. So, in this case France is far and away the biggest importer of bulk Spanish wine, followed by Germany. The countries tail off pretty quickly, past Italy and Portugal.

That France is a massive importer of bulk wine has been commented upon before, and it is usually assumed that it is being added to cheap French wine, without being mentioned on the label (the latter is only required if the addition exceeds a certain percentage). In this case, though, the Comrade database show that the reported Spanish exports exceed the reported French imports by 4 million litres, which is rather a large discrepancy (0.9%).

As an alternative, we can look at Germany, which is the world's largest bulk-wine importer. Germany actually imported from 28 countries in 2017, but once again I have shown only the top 15 countries in the figure.

The diagram shows us that, at the moment, Italy slightly exceeds Spain as the main source of bulk wine for Germany, with a pretty rapid decline after that.

We can also look at both exports and imports. For this, we might look at the USA. In 2017, the USA exported to 58 countries but imported from only 15 countries, when looking at large containers. These are shown in the next two graphs.

Clearly, the United Kingdom is far and away the largest recipient of American bulk wine, while Chile, New Zealand and Australia vie to be suppliers.

The Sankey diagram also allows us to combine imports and exports into the same graph. The data for the next figure are taken from the American Association of Wine Economists' Facebook page: World's Top 30 Bulk Wine Routes 2017. So, the Sankey diagram illustrates the 30 largest movements of bulk wine during 2017. As above, wine is exported from the left of each line and imported at the right, so that importer / exporters will lie in the middle.

In this case, the diagram indicates that Spain exports to Italy, which in turn exports to France, which exports to Germany. Spain also exports directly to France and Germany, and Italy also exports direct to Germany. Any flow of bulk wine opposite to these flows is not among the top 30 routes. Note that it would be impossible to show all of the routes in a single diagram using this arrangement, because flows can only go from left to right.

Perhaps the most interesting outcome of this arrangement of the diagram is what is shown about Canada, which exports more bulk wine to the USA than it produces! Clearly, Canada is re-exporting wine, mainly imported from Australia, while there is also a direct route from Australia to the USA.

Another version of this same diagram is actually shown in the comments section of the AAWE page linked above, created by Yuta Kanazawa. I have included it here, because it is such a pain to directly access old Facebook posts. In this case, countries appear at both the left and right of the diagram, depending on their role as either an exporter or an importer (eg. Canada, France, Italy, USA). With this arrangement, rather than the one above, we could, indeed, show everything.