The phrase “data- the fuel of the 21st century” has often been bandied about and has become a catchy slogan for the dawn of the digital revolution. A huge amount of data is being generated, collected, and processed every second with every single click. But what is the data even worth? Who is paying for such data and who receives remuneration? Or has data even become a “currency” for services perceived to be “free of charge”?
In her Duet Interview with Dr Andreas Zechmann, a specialist in corporate finance, Dr Caldarola, author of Big Data and Law, discusses the valuation of data and its current and future market price.
With the rise of the Internet of Things and the incorporation of sensors in objects, companies are in a position to collect more and more data. Consequently, the amount of data is increasing. Back in 2018, the total data volume collected worldwide was about 33 zettabytes. Many data analysts cautiously predict that this data volume may rise to a level of 175zb by the end of 2025. So, one could say that there is an interest, to put it mildly, in data. Is data perceived as an asset collected from data subjects and/or companies? If so, is that asset being identified in accounting regulations?
Dr Andreas Zechmann: Due to digitalisation, the role of data has changed in both the private and economic life of human beings since data has permeated all areas of life and industry.
You mentioned the development of the worldwide data volume. In my opinion, this trend shows the increasing impact of data on all of us.
The added-value chain of companies and their resulting business success depend more and more on data. This is not only true for modern digital business models, but also applies to traditional areas, such as, for example, infrastructure companies. I would even go so far as to say that data has become a strategic success factor for companies.
To answer your question whether data is today perceived as an asset, I unfortunately have to reply in the negative: No, not to the degree that it should be. There is a strong imbalance, when looking at data as an asset. Why is that so? The answer lies in the nature of data.
When we speak of digital data, we have to remember that data has no physical substance which in turn means that data is intangible. Data represents information on objects and subjects – for example, customers, vendors, materials, machines etc. and follows a life cycle (collection – use – deletion). The question of how well data represents information on objects and subjects can be answered by looking at the quality of its data.
The discussion concerning the status of data as an asset is dominated by the future prospects of customer and consumer data, because I think companies are well aware that their customer and consumer data are undoubtedly assets, on which their future monetary success depends. Trade secrets, such as recipes like the formula for Coca-Cola, or production methods to create drugs, are perceived in the same way. But, when it comes to corporate data other than customer data or trade secrets, my opinion is that companies, for example, do not consider inventory data in their enterprise resource planning system as an (intangible) asset – despite the inventory itself being considered as an (tangible) asset. I believe that the perception of data as an intangible asset is more pronounced in digital industries than in traditional (production) industries. It will be exciting to observe whether data is perceived in the same way as production halls, machines and facilities in the years to come.
I definitely still need to answer your question as to whether the current accounting rules perceive data as an asset and, once again, I must reply in the negative. The reason is a historical one, since current accounting regulations were established in the early 20th century when there was a strong emphasis on tangible assets, such as machines or properties, and intangible but “traceable” assets, like licenses, rights, concessions and bonds. The concept of data, as we know it now, was not yet conceived of within structures which had evolved much before the digital revolution.
In addition, the International Financial Reporting Standards (IFRS), that came into effect much later, namely, in the early 21st century, balanced intangible goods only as intangible assets when a company could prove that the asset would provide a future economic benefit stemming from its use. And that is the critical point as far as the topic of data is concerned: to evaluate and quantify production and acquisition costs. Consequently, a company’s data is not viewed as an intangible asset either because its future economic benefits cannot be proved or because the monetary production cost of self-generated data cannot be quantified, as is the case of computer-generated mass data, e.g., created by the use of sensors. I think can is relative because it is simply not done in practice.
In my PhD dissertation, I considered how to assess the value of data generated by a company in order to render it into a concrete asset for the company in question. My main argument was that only data that is actually used by the relevant company and is of high data quality and provides a positive value proposition for the company at the end of the added-value chain, can thus be said to have a financial value.
This then sets the stage for data becoming intangible fixed assets which can be balanced and depreciated just like other (tangible) assets. I deliberately refer to “self-generated” data because acquired data has a purchase price.
The mere presence of data does not automatically ensure its financial worth. Only data that is found to be of some use in the future can meet the criteria of being a data asset and be worth something. A good example would be data concerning a customer who hasn’t purchased anything for 10 years, where the invoice address is incorrect, or where the last purchased product is already outdated. I think this thought is consistent with the IFRS because, from the company’s point of view, this data will most likely not generate any future monetary benefits; perhaps we can compare such data to perishable goods from the food industry.
My opinion is:
“The mere presence of data does not insure their financial value. Only data, which are in some way and at some time found to be useful, fulfill the requirements of being data assets and thus can be said to be worth something, financially speaking.”Dr Andreas Zechmann
In order to define data as an asset in accordance with accounting regulations, data needs to have a market value. How is or how can data (be) evaluated and do companies’ controlling departments actively valuate and monetarize the data? Is the process either tied to obtaining additional profits or to cost savings?
As mentioned earlier, the acquisition or production costs are relevant for the financial value of an intangible asset. That means that the valuation either results from the price that has been paid when the good was purchased or the cost that accrued when the good was manufactured, for example, personnel costs involved.
Let us consider software as an example and the rationale behind this procedure becomes clear: either it is the price of software when purchased or the cost for the programming of in-house generated software; such things can be quantified because the programming started and ended at a certain point.
The discussion becomes more difficult when data is involved: The easiest and most objective data value would be the market price of a non-monopoly market with several suppliers. This would be feasible for customer and consumer data since such markets exist, so that purchased customer data could be valued in line with the purchase price, as with any other tangible good and, of course, could depreciate over its life cycle. One could even evaluate customer data in accordance with a potential purchase price that a third person would have paid (so-called “market test”).
But what happens with other data that is typically self-generated by the company and not acquired from a market, such as screws of a plant manufacturer? For self-generated data a market test is not applicable, so that we must find alternative ways of determining a financial data value. One possibility is to quantify the costs in the process used by a company to collect and maintain the data in question.
But, at this point we leave the realm of objective valuation and enter into subjective valuation with the result that there is no plausible monetary value for that data. It is obvious that, for example, “customer base” data sets can be appraised based on their cost by dissecting the collection process of a data set. For example, how many fields are included in a customer data set? How much time is needed to fill in these fields? How much is the hourly rate for the employee in charge of filling in the customer data? I would like to add this thought: if we have obtained a cost-based valuation in this way, we still don’t have a continuously increasing financial asset just by collecting data; meaning that, even if a data set is copied a million times, the financial value of the data set and its copies would not increase by a factor of a million. On the contrary, it remains constant. In my opinion, cost-based financial valuation is only partially applicable for data.
Consequently, if a market-price-based valuation of data does not seem to be suitable, the company should then, in my opinion, attach a value that is in line with the advantages arising from the data use. Simply stated, data that is often used and that therefore substantially contributes to the added value has a higher worth than data that is not or only used sporadically.
In my PhD thesis, I compared different valuation concepts and I came to the conclusion that the financial value of data should arise out of a use-based valuation approach considering the revenues that a company could earn and the investments that a company could save on through efficient processes. Use-based means in this context that a company focusses explicitly on the use of data in selected business processes, e.g., the role of customer address data in an invoicing process of a company. High-quality data has greater potential to be used and be profitable through its efficiency than lower quality data. It must be admitted, however, that data value derived in this way is subjective and not comprehensive because of the specific process being focussed on. By looking at financial data values in this way, we have the advantage of the value now being strongly linked to its use and thus being separated from the discussion that merely collecting data results in an intangible asset because of accruing production costs.
I would like to add that assessing the financial data value is not a means to an end. Valuation is only a first step leading to post-processing activities. All methods for evaluating data should support any post-processing valuation purpose. At the moment, data valuation hasn’t really become common in the financial community which is currently operating according to traditional financial Key Performance Indicators. There are a number of reasons for this, including, for example, the importance of accounting in capital market communication.
Personal data stems from data subjects and non-personal data from objects owned by a proprietor or a group thereof. However, only companies are collecting data – whether of a personal or an impersonal nature. They are investing in infrastructure for collecting and processing data and they are using the data of others for their internal and external purposes. Is there any monetary exchange going on between the “data-owner”, on the one hand, and the data-collector and data-exploiter, on the other. After all, we are familiar with trade of this nature for other raw materials that companies purchase.
Your comparison between data and raw material for business processes is, in my opinion, very apt. Data has developed from simply being a raw material to becoming an enabler of processes and, from there, developing into an enabler of products and services and, even beyond that, to data as a product. This is a new dimension which does not exist for traditional tangible raw materials.
This development leads to an increasingly large debt imbalance with regard to monetary compensation between data collector, data owner and data exploiter. In my view, when we omit social media platforms, where private persons offer their data free of charge, and there is a strong divergence among these three roles, then the problem does not really arise in a corporate context. Here the companies themselves occupy, in my opinion, all three roles, so that the problem of financial compensation does not arise. Nevertheless, with regard to social platforms, such as, for example, Facebook, the data owner / data subject could be compensated financially, even if we are only talking about trifling amounts.
Basically, when discussing data, the challenge is that data can easily be copied without accruing additional costs. Therefore, determining data control is more difficult than is the case for tangible goods that can be secured (secure in the sense of not letting third parties into a warehouse) and cannot be copied. An essential feature of tangible assets is their shortage: they are “valuable” because they are not accessible to everyone and are somehow scarce. If one looks, for example, at freely accessible data, such as “open data”, the question arises as to who, if anyone, owns such data. To be sure, this type of data has a data subject, but it still represents, in my opinion, a free asset available to all and belonging to no one. Therefore, I am not surprised that there is no monetary compensation being distributed among the different roles.
Perhaps an example will make my answer clear: A photographer taking a prizewinning photo of the Matterhorn in Switzerland does not pay the Canton of Wallis compensation for the revenues he makes because his photo has been included in a commercial photo calendar. Everybody could take such a photo if s/he wants to. I view the internet in the same way: A company markets data which everyone has access to. Everybody could do the same thing, but only a few actually do, thereby earning money with its usage.
Assuming that data has a value attached to it, what are the actual market prices for data? Do special categories and respective monetary ranges exist? Are they stable or do they increase over time? Can you provide some examples?
I think the topic whether data is traded on markets is only relevant for data, for which sufficient prospective buyers exist, such as, for example, for customer and consumer data. In my opinion, the going market prices for customer data sets offered by large information service providers are in the range between € 0.5 and € 2.0, to name one example. But let me give you a counterexample with respect to our market discussion: Only a few prospective buyers exist for data related to load-bearing strength of constructions. Consequently, this data is unlikely to be sold or purchased. However, for railway companies this data is crucial for maintaining their systems and, therefore, for their added value chain. No train would move if information from construction data were missing. My point here is that the discussion on market price of data is only part of the whole discourse concerning financial data valuation.
With regard to financial data values, I would like to add that I have applied cost-based and usage-based approaches for companies. The cost-based approach generated around 38 CHF for each material master data at Swiss industrial companies and between 11 and 30 CHF for each customer master data. When applying the usage-based approach, the financial value amounted to about 36 CHF for customer data of a power tool supplier and roughly 5 CHF for construction data of a railway company for each data set. It is interesting to note that the values for each data set quickly reach an asset value in the range of double-digit millions in cases of 100 000+ customer data sets. These are significant values in the balance sheet of an enterprise.
But coming to your question about price development, one must be clear about the fact that data is strongly connected to its timeliness, meaning that obsolete data, such as incorrect information about real objects, isn’t worth anything. Nobody would buy data that has “expired”. But I also believe that, as the data trade becomes more important, more markets will grow, and the topic will receive more attention, as digital business models become more essential.
Let me add the following thought with regard to the distinction between price and value. The value corresponds to the price only when the latter corresponds to a multitude of expectations of the potential buyers. If this is not the case, and, if only a few buyers exist, then the market price is one for a product offered on the market, but this value is not a reliable indicator for an objective market price.
Many digital business models work in the following way: a data subject allows access and processing of his/her data and receives, in return, a “free” service. For example, a lot of people use “free” messaging, email services or cloud space and, in return, let the provider use their data. Has data become a currency to “pay” for certain services? Or is data the cheapest, if not the smartest, “stolen” asset?
The idea that data is a stolen asset simply muddies the water in terms of what the value of data actually is and what should be paid for it.
The fact is, data can be copied an unlimited number of times if enough storage space is available. Reproduction costs remain negligible because capacities are so big.
If users provide their data without remuneration and make use of a service at no charge to themselves, an exchange has taken place, with the provider having gotten the better deal since s/he can generate revenues on his/her website through advertising. In the strictest sense of the word, it is a fair “exchange” and not a theft. But if I go back to your question on data being the “smartest” stolen asset, then I have to agree that the user has most likely been enticed into providing his/her data at no charge for a “free” or discounted service. I assume the user is not aware of the fact that s/he has agreed to a barter transaction or, if s/he is cognizant of the implications, that s/he has consciously accepted the deal.
An owner of a mine usually sells his/her raw material or leases his/her mine to somebody so that the leaseholder can exploit the mine/raw material. The owner of a mine, therefore, receives money either for the raw material or for the lease. If we transfer this example to data, are data subjects receiving money for their data? Or differently put, should a data subject consciously put a price on his/her data? Are there any business models or current trends where data subjects can actively monetarize their data? Or do only companies – data collectors – makes these sorts of commercial deals?
In my opinion, there is no conscious data monetarization taking place on the part of data subjects / consumers / customers. We cannot even observe a trend leading to any change in this situation because our herd behaviour always seems to be on the alert for free disposable internet-based services (e.g., mailing, YouTube), for which the provider earns advertisement revenues.
I could imagine consumers receiving credit vouchers for trifling amounts for their data being further processed beyond the original transaction. Also possible is that consumers deliberately offer their data on platforms for marketing and advertisement purpose, but also here we would still be talking about pocket change. In my opinion, data trade remains a bulk business where a huge amount of data is purchased with the knowledge that a certain percentage is useless because it is obsolete.
For this reason, the professional monetarisation of data will probably remain in the B2B sector, even if the data owner / data subject does not participate in the revenues accrued.
Not only private companies collect and process data. In some countries, governments are also collecting and using data – and not just to fulfil their services and their security obligations. Do government also “pay” or should they be offering some form of remuneration for data?
I imagine you are alluding to the Social Credit System in China. The system serves, in my opinion, to monitor the population and to reward some people for loyal behaviour. This is a special case that works in a “microcosm” of a totalitarian state, such as China. The Chinese government does not wish to generate revenue through the trade of data.
Otherwise, I cannot think of any other examples where governments deliberately collect data to monetarise or trade data. In my view, governmental activities should focus on services, such as education, safety, infrastructure, maintenance of public life as well as medical and sanitation services.
During the pandemic, we have experienced in a positive way how governments can become a sort of data octopus that spreads its tentacles and sucks up all the data it can find for a concrete purpose, in this case, for the Corona-Warning-App. My point is that even this example is based on citizens being willing to provide their data, which they do partially because they believe their data is in safe governmental hands. I don’t think the Corona-Warning-App would work if citizens thought governments generated trade by using personal data. A breach of trust between citizen and government would put the system of voluntary data provision at risk and lead to governmental data trade no longer functioning.
My final thought on this matter is that governments should not trade data so that the question of paying data subjects never even comes up.
Yuval Noah Harari stated in 21 Lessons for the 21st Century that: “Big Data algorithms might create digital dictatorships in which all power is concentrated in the hands of a tiny elite while most people suffer not from exploitation, but from something far worse – irrelevance.” and “It is much harder to struggle against irrelevance than against exploitation.” Should a data subject, therefore, start to monetarise his/her data in order to actively shape his/her participation in the digital market and to fight against the exploitation and irrelevance of his/her data?
The question seems, in my opinion, to be rather sensationalist. You stated, first of all, that there was a market for data subjects to monetarise their data and, furthermore, that data subjects have access to this market. In my view, this situation does not hold true due to the nature of internet-based services, since they are formed in a huge radius involving registered users, visitors and clicks. I do not think data subjects can participate in this bulk business- particularly because data subjects profit from this mechanism, for example, posting of WhatsApp at no charge as opposed to paying for texting. The spirit of the internet – namely speed, simplicity and flexibility (e.g., Uber)- has its price: namely, the end-user providing his/her data at no cost; on the plus side for the end-user: services provided at no charge. Meanwhile, what does the provider get out of it? Possible revenues from the data trade minus the cost for the service provided.
Dr Zechmann, thank you for sharing your insights on the value of data.
Thank you, Dr Caldarola, and I look forward to reading your upcoming interviews with recognized experts, delving even deeper into this fascinating topic.