Everyone says data is an asset, but few know what kind of asset data is
Data assets are talked about a lot, but few really understand what kind of asset data is. When you do understand it, you realise why most people don’t, and how we, the data community, can help them.
The word “economy” comes from the Greek roots oikos (household) and nemein (to manage). A smooth running household manages things well. In the economy, a smooth running business does the same with what we call assets.
The earliest human societies had to manage assets essential for survival, like food, water, clothing, shelter and weapons. With the agricultural revolution, humans developed new assets. Some of these assets were essential items, like crops, livestock and furniture. Others served as abstractions of value that made trading different assets much easier, like money and gold.
As societies have become more complex and technologies evolved, so have our assets. Things like government bonds, debts, intellectual property and company stocks.
Once an item has been recognised as an asset, a range of disciplines emerge around it. There are best practices for optimising and preserving the asset, standardisation processes, legal regulations, valuation techniques, markets and more. These then make economic activities involving the asset more insightful and efficient. Creating more value for the traders of the asset, and the economy as a whole.
Today we talk about a new asset class – data.
Data has long been collected and analysed by the scientifically minded (Aristotle wrote multiple treaties on biology, from his systematic observations of zoological data). Data’s unique properties and its ability to generate insight in an increasingly high-speed world make it not merely a new asset, but the asset that organisations need if they want to thrive in our digital era.
The cliché “data is the new oil” stands for good reason – just as the industrial era’s factories and global trade networks were powered by fossil fuels, so will the digital era’s production and distribution networks be defined and made possible by data and data analysis. This leads us to two problems. One, data often gets confused, or subsumed into the technologies it fuels. Two, we don’t have helpful metaphors to accurately describe it and so avoid confusion. Despite what we wrote earlier, data is not like oil, and it’s not really like renewable energy either.
Previous generations of merchants, financiers and investors knew well the properties of the assets they traded in. And did very well because of it. Today, the most successful organisations are the ones that have, and understand, their data assets. This means they make considered data investment plans – prioritising data investments to get the best ROI from this unique asset.
Data isn’t subject to the same forces as other assets
Daniel Moody’s classic paper Measuring the Value of Information draws our attention to one of the most important properties of data, referred to by economists as being “anti-rivalrous”. Unlike capital or oil, data is not limited to a single owner or user. Easily shared and copied, and undepleted by usage, it does not obey the same competitive dynamics that traditional assets do, and around which historic markets have emerged.
One organisation possessing and analysing a particular data asset does not prevent another from also possessing and analysing the same data asset. Intellectual property laws and other protections may seek to minimise this, but in principle, and “left to its own devices”, data can be copied and used indefinitely. Consequently, organisations need to take special care to protect their proprietary data, insights and competitive advantages from data breaches and leaks.
Combining data can create value, so how accessible it is matters
The anti-rivalrous properties of data also open new avenues for inter-organisational collaborations, which will prove essential in addressing global 21st century problems. During coronavirus, for example, national and international data sharing between public health and medical communities has enabled a coordinated scientific response to the pandemic.
In practice, organisations will find different ways to allow or restrict access to their data assets, depending on the sensitivity of the data and what it is used for. The Bennett Institute notes that data assets can be divided into three categories: open, closed or shared access. Open access data is available to all potential users. Closed access data is available only to a certain organisation, typically the one that collected or produced it. Shared access data is shared amongst multiple organisations, though is not openly available. The graphic below explains this in more depth:
Data’s seen as a sunk cost
Unlike traditional assets, valuing data assets is not straightforward
A data asset that is extremely valuable to one organisation may be worthless to another, or indeed to all others. Investments in physical assets like stock and machinery can be recuperated, at least in part. But liquidating data assets can prove tricky, if not impossible. Furthermore, even when a data asset is clearly creating value, it is notoriously difficult to calculate this value. Analysing data is a complex process, requiring capital investment, computational power, human labour, and potential investments in knowledge and training, to note just a few of the variables. Even when final outcome of data analysis is evident (like a notable improvement in efficiency), it is difficult to isolate the contribution of the data from the whole process leading to this outcome. This all causes headaches for accounting teams, which means data assets don’t get treated as they should. A core part of our business is data valuation, which we do by identifying the role data plays in the activities that create value for stakeholders.
Unlike most assets, using data creates more value and a new asset
Data quality and quantity is subject to the law of diminishing returns, oftens significantly
Some data is perishable
To sum it all up
Haskel and Westlake sum up the properties of data, and intangible assets in general, with 4 S’s: scalability, referring to its copyable, anti-rivalrous property; sunkenness, referring to the difficulty recuperating expenditure on data; spillovers, referring to the possibility for competitive advantages to be noticed and copied; and synergies, referring to new opportunities for collaboration and combinations of different data assets.
In conclusion, data is anti-rivalrous, enabling it to generate exponential returns without being depleted, but also making it a competitive and regulatory liability. Its value, while evident, is difficult to calculate, and may be impossible to recuperate via traditional methods like liquidation. It presents new opportunities for collaboration, and organisations that make use of its synergies with other data and emerging technologies will be well rewarded. All this is difficult because, as an asset class, not all data is the same.
Crops, livestock, commodities, currencies, financial assets and fossil fuels were assets underpinning the economies of previous eras. The traders who were able to understand and capitalize on the unique properties of these assets were the ones who succeeded. The asset underpinning the digital economy is data, and organisations that understand its properties well enough to extract its value will be the ones that thrive. Just look at Amazon.