The consequence of valuing data
For years theorists, economists, and even tech-evangelists, have all been arguing over the value of data. Notwithstanding the hype, even the Economist in 2017 suggested in a leader article that data was the new oil.
Gartner has been writing about ‘data as an asset’ for years. We even ratified calculations to help determine and quantify the value of data assets. See Applied Infonomics: Why and How to Measure the Value of Your Information Assets. And one of our number wrote a book in the topic (Infonomics): a career pinnacle.
But as most readers will know and appreciate, there is no legally binding method to value data and more importantly, accredited accounting rules around the globe do not even recognize data as assets. As such organizations are not permitted to bolster their balance sheet with the value of their data assets.
There are some interesting related angles, such as things like goodwill which is an intangible asset. Likewise, intellectual property (IP) can be valued (and is, as a capitalized assets) such that patents can be bought and sold; and listed as an asset in a balance sheet.
If your company goes broke you might find out that your lawyers or administrators might be able to sell your customer list, or perhaps some other data from your now defunct internal business systems. Such data might be sold for money, to help pay the bills. Yet this possible fire-sale, clearly showing that some people do value data, has no bearing on the value of the data when you are in business as a going concern. Sure, you could insure the database or system in which the data is stored; so, there is some protection there. But that protection generally only goes so far as covering the replacement cost for such data.
I have read widely in information theory and despite the rhetoric, there remains ample disagreement in research circles that data has an implied value independent of use or context. I myself tend to struggle with the point: From 2013, a blog: What is the (business) value of data, anyway? And more recently in 2017, another blog: More on Data as an Asset and Ownership.
Does data at rest or in motion, with no context, have value that can be quantified? If there is an implied context or use, how can that drive the value of data given that context can change and more importantly, the outcome is not guaranteed? The value of a bond change over time when interest rates fluctuate? Of course, they do – but that is a known context.
These are all vexing questions and exploration of the calculations at the heart of Infonomics will help. But their value is in fact based on assumptions, like any other asset. And this leads me to the reason for today’s blog.
In [a recent] US print edition of the Wall Street Journal, I read an article titled, “Water-trading market runs into trouble”. The article tells the sad story about how the Australian government has tried to manage the shortening supply of water in a region west of Sydney (Darling river) by introducing a cap and trade system. The process, being evaluated for use in the US, has not worked.
The article explains how the introduction of a price for a commodity, such as water, has led to unintended consequences. The idea was good: if water was valued and quantified with a price, farm use would shift to more profitable and so more sustainable levels. Water might be traded such that not every use would be exhausted, leading to a balance in water consumption and use. However, by valuing this asset more agencies became interested in buying up water; and some of those were not interested in farming at all but in making a profit on trading a commodity. Investors started to trade water since they could turn a fast profit. Theft and hoarding have also become common.
The conclusions are broad. First how a market is defined can be problematic. Public sector agencies don’t have a great reputation here and if you are a staunch follower of Milton Friedman, you would know why. But even the example reported highlights some obvious policy changes that might have helped, such is only permitting farmers to trade. But that would require yet more regulation and monitoring.
But the most interesting angle for me pertains to the value, the market price, of an asset and what it might mean should data really be treated as an asset. There are huge differences between water and data; water cannot be replaced once used; data can be copied, and use has no bearing in its subsequent availability (ignoring license rules).
Water is hard to transport and store; data is effectively free to transport, and storage costs are relatively cheap, compared to water. But the point is that assets of all types and classes can be stolen, hoarded, and used to otherwise drive prices up.
Every day we hear of breaches and corrupt uses of data. If data were to become an official asset, quantifiable and reported on balance sheets, won’t human corruption and misdemeanor naturally increase? Quite probably. Perhaps that’s the costs of doing business….
(This post originally appeared on Andrew White's Gartner blog, which can be viewed here).