Origin, Timing, Provenance and Quality: Is Data the New Wine?


In association with Teads.

In this exclusive article for ExchangeWire, Jamie Toward (pictured below), head of data, UK & NL, Teads, discusses how data is perhaps less like oil, as it is commonly compared with, and instead is more analogous to wine.

“Data is the new oil” is an analogy that’s gained traction since it was first used by The Economist in an article published in 2017.

Since then many people have sought to prove exactly why data isn’t the new oil, focusing on flaws in the analogy, comparing data to gold, soil or even bacon.

As some of us look beyond ‘Dry January’ to February and the prospect of uncorking a bottle of red or perhaps white, it seems to me that data is more comparable with wine than any other material.


1. It just keeps growing
Jamie Toward

Jamie Toward, head of data, UK & NL, Teads

Unlike Oil, which is only extractable from finite resources, data just keeps on growing. It’s not only “mineable” it’s cultivatable – individuals are like vines that keep growing and producing fruit, in the form of data, on an ongoing basis.

2. Where it comes from is really important

Many variables are important in dictating the qualities of wine. The vines it’s grown on, the type of grape, the amount of sunlight and rain, when it’s harvested, even the skill of the grower, all contribute to what a wine tastes like. And, consequently its value gets dictated by all of these factors. The combination of the natural and human aspects of growing grapes form together to describe “terroir”, a slightly ill-defined but much regarded concept in viticulture. And, terroir is a vital clue in the quality and value of wine.

And the same is true of data.

Think of an single data point as an individual grape. That grape has already been influenced by a lot of natural factors that determine its colour, size and sweetness. Similarly, a single data point has already been influenced. If, for example, we said that a data point had an attribute of showing intent to purchase a new Volvo estate car it may have been influenced by many factors (the branding efforts of Volvo, the product itself, the fact that an individual has just had their first child, the reading of reviews, seeing the same car drive past on the street). Each of these influences, and many more variables and combinations, have coalesced to produce that single data point.

Much like the grape is a product of the terroir, the data point is a product of the real-world influences on an individual that the characteristic is attributed to.

3. Its harvesting

Wine can be harvested by machine or handpicked. Many wine producers use hand picking for their flagship wines, while scale production is most effectively conducted mechanically.

And there are similarities in the world of data. Sometimes the richest data is the most difficult to harvest. Qualitative data can, sometimes, provide the greatest insight and by its nature requires delicate extraction in the form of in depth interviews, active form filling, or human observed behaviours.

Quantitative data can provide digital observation at scale but tends to capture less rich, less insightful data.

4. Timing the harvest

Grapes are picked at varying points of ‘ripeness’ to ensure they produce the right characteristics for the type of wine a producer wants to make. Similarly, data points need to be harvested at different points in a customer journey to match the characteristics that the user of the data wants to achieve.

If you want data about somebody buying a product then you need to harvest data points that demonstrate that attribute. In short, data purchasers need to be sure that the data they are buying is appropriate for the kind of outcomes they set off with as their objectives.

5. Juice extraction and quality assurance

When wine is made there needs to be an assurance that the grapes being used are healthy. Further, as they’re being crushed and pressed to turn them from fruit into juice, there needs to be an assurance that this all takes place in a clean and hygienic environment.

And data is again the same. Data collectors and processors need to be certain that what they are collecting is ‘clean’ data – that the data point being collected is actually what it says it is – you don’t want your intent data getting contaminated by your interest data now do you?

In more and more markets there’s an absolute legal requirement to ensure that consent for data collection was given right at the start (while the grapes are growing on the vine and being picked). Ensuring consents are in place and that the data stays clean all the way through processing is a vital component of the wine/data industry.

6. Pairing

Many people will tell you that certain wines should be “paired” with certain food.

And again, there are comparisons to data use in advertising and media. Buying media against audience data only is simply not as powerful as buying a combination of audience and context.

Combining the power of targeting people with the right audience attributes is great, but it’s effectiveness can be improved by making sure you’re communicating with people inside content that indicates they’re in the right mindset to receive the message.

7. In Vino Veritas

As a final thought, recall the Latin phrase “in vino veritas”… in wine there is truth. Sometimes, the use of data will lead you to find things you didn’t expect and sometimes the answer it gives is not the one we wanted.

It’s also worth considering that when dealing with data in marketing. People can believe that data is a ‘truth’ and often it is, but what kind of truth can be difficult to ascertain. Is it deterministically or probabilistically derived? Is the ‘truth’ that somebody ascribes to a data set the same ‘truth’ that the buyer was looking for.

The bottom line here is that using data as a marketer is a bit like drinking a bottle of wine. If you bought it from a reputable provider, it tasted good and does what you wanted it to do then it was a good buy.

But always keep in mind that getting advice and guidance from an expert is going to be a wise course of action. For wine that might mean getting the advice of a sommelier – to achieve the Master Sommeliers diploma can take 6 months of full-time study and less than 300 people have achieved the title in over fifty years.

While there’s no equivalent distinct qualification in digital marketing, it might be time to consider that there should be…

So, is data the new wine? Probably no more so than it is the new oil. But it does need to be a better understood part of the marketer’s armoury and 2020 is definitely a great vintage to start learning from.