It is sometimes said that “data is the new oil”. The expression is almost ten years old – according to Forbes, it was coined first by Michael Palmer.
But is data really the new oil?
Well, let’s see why data was being compared to oil in the first place.
The most important reason is because oil has been the essential fuel and commodity of the 20th century economy.
Oil is not only used to generate energy for transport and heating. It is also the basis from which a large scale of other essential products are produced: hydrocarbons are used to create plastics, lubricants, asphalt and wax, and petrochemicals are used to create fertilizers, pesticides and many other materials and substances.
In short, oil and its derivative products are everywhere – we could not imagine our world without it. And we continue to invent and develop new products or uses based on oil.In addition, for most of the last 100 years, oil has been relatively cheap.
So it’s safe to say that oil is an essential commodity and driver of economic development throughout the 20th century.
Similarities between data and oil.
The similarities between oil and data are based on these characteristics. Like oil, data is becoming the essential new commodity for the 21st century economy
The amount of data produced by humanity doubles every year. As I’ve written earlier, this phenomenon has many consequences, some of which we are only starting to see.
But just like with oil, we are finding many different and new ways to use data. We create data, we share data. We analyse data. We mine data. We create derivative products, such as information and intelligence, leading to superior levels of understanding and knowledge.
Data are fast becoming both the basic commodity and the fuel of the information age, and are becoming a major driver of many different kinds of economic and social activity, from software development, education, over public policy and governance to more intelligent and efficient ways of – indeed – drilling for oil.
Differences between data and oil.
But there are also some very interesting differences between oil and data. And some of these differences are, in my view, more relevant than the analogies set out above. They are the reason why data is not the new oil – not in the way oil worked for the 20th century economy.
The first, and most important, difference is that oil is a rivalrous commodity and data is non-rivalrous. That’s economic code for the fact that you can burn a liter of fuel only once. The benefit of use of oil is captured only one time, by one person. (recycling of e.g. plastic does not alter that statement from an economic perspective, btw).
Data are non-rivalrous. When I use a certain data set, that use limits or prohibits in no way the same or a different use by someone else. Rather, it is the opposite. The more people use data, the more the data become valuable for all those users. Opening up and sharing data makes most data more valuable, not less.
That is exactly the opposite of what is true for oil.
A second important difference is in the production costs.
For oil, production cost is now between US$ 20-100 (including “up-front” or “sunk” costs) a barrel, and the marginal cost (this is the cost to producing one extra barrel) tends to go up, not down.
The opposite is true for data – the marginal cost of producing additional data tends to go very quickly to zero.
As a result, the economic dynamics of data are fundamentally different from the economic dynamics of oil.
An oil based economy runs on principles and methodologies of managing scarcity.
A data based economy runs on principles and methodologies of managing abundance.
And, interestingly, most of our economic thinking (such as it is) is based on the theoretical assumption of scarcity, not abundance.
An economy with a core commodity that is both non-rivalrous and abundant has not really existed before, so we have some obvious challenges in trying to come to terms with this new reality, and what it will mean in practice.
As the data-driven economy is starting to develop further, and becoming more and more important, we will see many changes, most of which we cannot predict quite yet.
But we can already make some predictions, and see some patters of what will emerge.
First, business models will often be wrong – not just a little bit, but wildly so. Any business model that presupposes scarcity of data must be looked at very critically – about as critically as a business model based on free oil.
While some data will remain scarce, particularly as a result of artificially created scarcity (the most obvious examples that spring to mind are legislation on privacy and security aspects, both of which aim to block or limit access to data), we will have to adapt our mindset in many business models based on data usage or exploitation.
Sharing data may create more value than not sharing data. Combining data with other data tends to create more value than keeping data separate. Organizing data in a network environment tends to multiply potential use and value.
Second, new ways of dealing with abundance will emerge. Most of our economic habits are based on the absence of abundance, and we will have to unlearn many things, in order to find new ways of creating, capturing and distributing value.
In such an environment does it make sense to “protect” data? If so, what is the cost of such “protection” – which is often not more than denying ourselves the ability to capture additional value from data?
And what will be the evolutionary and market pressures on how we deal with data? Are traditional market systems, with their information exchange system in essence based on pricing scarcity, still the best way to find out what’s going on in our economy?
Many questions, with new answers to find.
But one thing is certain: data is not really the new oil. It’s much more interesting than that.