The price of data

01 Aug 2018

We leave data trails across everything we do - on social media, when shopping and on the internet. Now the pressure is growing on tech titans to give something back to reflect the value of this resource.

“You mentioned your name as if I should recognise it, but I assure you that, beyond the obvious facts that you are a bachelor, a solicitor, a Freemason, and an asthmatic, I know nothing whatever about you.”
 
Thus does Sherlock Holmes amaze his new client, “the unhappy John Hector McFarlane” in the opening pages of The Adventure of the Norwood Builder. Holmes’s declaration is of course based on his core methods of observation and deduction, all from a cursory glance at McFarlane’s appearance. 
 
These days we leave such a trail of digital information behind us, in virtually all of our interactions, that the modern McFarlane’s LinkedIn profile and Tesco Clubcard records would probably betray far more about him than even Holmes could deduce.
 
A phrase often bandied about these days is that ‘data is the new oil’. 
On the whole we provide that information (with varying degrees of willingness) for nothing, either because we have in practice no choice or as part of a contract under which we receive a perceived benefit – typically ‘free’ goods or services – in return. There are, however, significant downsides including the risk of privacy breaches, identity theft and fraud, as well as the inevitable exposure to yet more (not necessarily welcome) marketing messages.
 
At the same time, data is the stock in trade that has made companies such as Alphabet and Facebook among the largest in the world. These issues raise the question of whether we are selling ourselves short; pressure is growing on the tech titans to give something back to reflect the value of this resource.
 
A phrase often bandied about these days is that ‘data is the new oil’.  In fact the saying was coined more than 10 years ago by Clive Humby, a co-founder of Dunnhumby, the company behind that great source of consumer insight, the Tesco Clubcard.  As we discuss later there are important differences between data and oil but Mr Humby’s original point was a rather narrow one – that like oil, data only really has value when it has been refined, and to that extent at least there are indeed similarities.

Data: Simon Lapthorne, Investec Wealth & Investment’s Senior Research Analyst, examines the power of data, and how it is used by large technology companies.

Data: Simon Lapthorne, Investec Wealth & Investment’s Senior Research Analyst, examines the power of data, and how it is used by large technology companies.

Two broader ideas that are often used to support the data=oil argument are
  1. That data is the fuel or the current, fourth industrial revolution just as oil was for the second; and
  2. The business of data is dominated by a small number of firms that exercise excessive power based on the exploitation of a common good. The sub-plot here is that ‘something must be done’, such as breaking them up, just as happened to the oil industry.
 
The climate of distrust and dislike of the big data companies has been building for some time, fuelled by a variety of concerns. These include their tax arrangements, the hosting of content that some find offensive and their alleged role in promoting fake news. The recent Cambridge Analytica affair has only served to strengthen this feeling, and the companies now seem to be well and truly in the crosshairs of politicians.
 
The climate of distrust and dislike of the big data companies has been building for some time.
The point about the industrial revolutions is broadly valid as far as it goes, but is really rather vague and warrants little more than a shrug of ‘so what’ in response. The question of market dominance (and possible abuse) and what if anything should or can be done about it is more interesting and on examination highlights some fundamental differences between data and oil that point to much more complex remedies than simple ‘trust-busting’.

 
First, data is not a commodity in the same way oil is. Each piece of digital information is different from the next whereas oil (subject to the differences between grades – e.g. Brent versus West Texas Intermediate) is all the same. Thus data is not fungible as oil is.
 
Second, data is non-rivalrous; in other words the same piece of data can be used multiple times for multiple purposes whereas a single barrel of oil can be used only once. Third, oil is a finite resource (albeit one whose demise within 30 years has been wrongly predicted since the 1970s) whereas our reserves of data keep on growing.
 
Around 90% of the data in the ‘digital universe’ was created within the past two years and the total is forecast to grow to 180 zettabytes (that’s 180 x 1021 or 180,000,000,000,000,000,000,000 bytes) by 2025. And unlike oil, data becomes more valuable the more of it there is, particularly in a world of artificial intelligence, where algorithms are constantly refined and improved by being fed more and more data (for example IBM’s Watson and Google’s AlphaGo Zero).
 
For all these reasons it is difficult to put a monetary value on data and they are part of the reason that it tends not to be traded for money once the original ‘contract’ has been completed (mailing lists being one exception).
 
There is another - paradoxical – complication to putting value on data, which in part goes back to Mr Humby’s observation that it is only worth something when refined. 
90%
Proportion of the data in the ‘digital universe’ that was created within the past two years
180zb
Forecasted increase in data by 2025 (that's 180 with 21 zeros)
 
We, as consumers, provide unrefined data and companies such as Google and Facebook turn it into something useful (note: other technology behemoths are available). Why, then, should we as consumers expect to extract value from something that is inherently worthless? Similarly, the amount of data we each contribute as individuals is relatively minor and in itself of little value. Again, why should we expect to attract value?
 
Herein lies the paradox; individual pieces of unrefined data are worthless but large (and the larger the better) quantities of refined data are worth hundreds of billions of dollars. It is an extreme example of a business in which the whole is worth more than the sum of the parts, but without the parts the whole would be nothing. And yet – notwithstanding the free apps and services and all the cat videos you can eat – the vast majority of that value accrues to the companies, not the providers of their raw material.
 
There are, however, factors that suggest traditional antitrust solutions are not appropriate or even desirable. Companies whose stock in trade is data rely on network effects, the virtuous circle of offering services, gathering data from users, using the data to improve services, attracting more users and gathering more data.
 
Breaking them up would probably only have a short term effect before one of the resulting smaller companies grew back to occupy a dominant position. In the meantime the effect could be to impede progress in artificial intelligence, a field that is highly dependent on ‘big data’ and which has applications that most would agree are of great societal benefit, such as medical diagnostics and prevention.
 
This is already hampered by the fact that an estimated 70% of all data is privately held with only 30% in the public domain. This raises questions about data sharing. Clearly there are many applications that would benefit from it, but the problem of assigning value again arises, as does the risk of security and privacy breaches which is elevated every time data moved.
 
The recent General Data Protection Regulation had tightened the rules governing data protection, but security is likely to remain a significant concern, particularly with issues such as the recent Equifax hacking incident still fresh. It serves to strengthen yet further the argument that data providers are inadequately rewarded.
 
We, as consumers, provide unrefined data and companies such as Google and Facebook turn it into something useful.
A second approach, rather than breaking up the companies, is to return a degree of control of data use to people from whom it originates by requiring more explicit permissions to use, limiting the uses of data and allowing people more control over where their data is stored. In theory, as things stand our data is not used unless we give permission, either explicitly or implicitly, and we have every right to withhold that permission. In practice this is less straightforward. In many cases, to do so would require us to forego many useful services, some of which are impossible or virtually impossible to access otherwise.
 
Most of us have been guilty at some time or another of clicking on the ‘I Agree’ box without reading the interminable and impenetrable small print.
 
Experiments have shown that people pay such scant regard to terms and conditions that they have inadvertently agreed to give up their first-born children in exchange for free services. Purists may shout caveat emptor, but measures could perhaps be taken to make the process less opaque at least. 
 
There are also financial routes to compensating the providers of data, notwithstanding the difficulties of valuation. There are companies – Citizenme and Datacoup among them – that claim to enable consumers to monetise their data directly by selling to information brokers or advertisers. Another approach – which would require regulatory intervention – is to force companies to pay royalties each time they use somebody’s data. And then there is appealing to their better nature, which might not be as outlandish as it seems.
 
The CEO of Facebook, Mark Zuckerberg, has expressed his approval of a system in Alaska under which all citizens are paid dividends out of the income arising from the state’s oil production. He has been challenged to use the value of Facebook’s data to a similar end, and in the current climate he may end up having no choice. Perhaps data really is the new oil.

Download the Vision 2018 Brochure

Book one

Book two

Book one

The first of a series of three Vision 2018 books, covering three key topics - War, Water and Investor Behaviour.

Book two

The second of a series of three Vision 2018 books, covering personalised medicine, the great taskmaster, and the price of data

Discover how Investec Wealth & Investment can help you and your clients

Find a Vision event near you