With the advancement of communication technology and the rapid spread of the mobile Internet, every individual has become a producer and user of data. As a new type of asset, data is considered the “oil” of the fourth industrial revolution and is expected to bring about significant changes in productivity and production relations. Unlike traditional assets (properties, securities, etc.), data assets have some distinctive characteristics, such as the high degree of privacy of certain data types, and the separation of data ownership and control. Currently, we observe the following two data business models.
In the traditional data business model, data producers usually do not get paid for producing the data. Take the massive amount of data generated by users through social media networks as an example. Service providers have actual control and usage rights of the user data. They can mine the value of data and profit from it (e.g., analyzing user behavior and pushing marketing ads) without paying users rewards.
As more users realize that data is an asset, how data owners and controllers can cooperate in a win-win situation has gained more attention, i.e., how data producers can profit from their own data in a privacy-preserving manner. Data trusts is a novel data business model proposed recently to address the above challenges. The organizations that collect and hold data (e.g., service providers in the above example) act as trustees. They assume responsibilities (e.g., increase data value) and obligations (e.g., protect user privacy) for user data on the premise that they have reached a legal entrustment agreement with users. Meanwhile, they will charge a certain service commission from users as their income.
Like traditional trusts, the aforementioned data trusts model has some problems in aspects of data assets analysis and trust transaction settlement, such as high processing fees, complicated and slow settlement processes, low trust information transparency, etc. To reduce the trust cost, accelerate the transaction process, and increase the data transparency, we envision two promising models in this article, namely Agent-based and P2P-based data monetization. Both models employ blockchain as the underlying architecture since blockchain can bring the following benefits:
- Decentralized Network—Blockchain builds a peer-to-peer self-organizing network by establishing a set of highly transparent open network protocols. The processes of validation, storage, maintenance, and transmission of data in blockchain networks are implemented based on distributed system architecture that can provide data monetization with a high fault-tolerance.
- Enhanced Trust of Data—Consensus protocol and hash encryption algorithms assure data recorded on blockchain is immutable and traceable. Therefore, the use of blockchain technology can make data registration, transfer, and settlement more transparent and credible.
- More Efficient Business Process—Smart contracts can write any business logic (e.g., access control, authentication, pricing, and settlement) in the form of computing code on blockchain. The events in the contract are automatically executed when pre-defined trigger conditions are met, which ensures that the entire data monetization process can be automatically managed and executed.
TWO DECNETRALIZED MODELS
(1) Agent-based Data Monetization
Lacking sufficient time and knowledge to operate their own data assets, data owners can choose an agent-based data monetization model to benefit from the value of their data assets. Similar to data trusts, the agent here refers to a legal entity that can offer independent data stewards. It can be the data actual controller or the external independent third party. It provides expertise to help clients unlock the value of their data and charges commissions from clients.
Compared to the centralized data trusts model, the proposed model builds a more efficient and transparent data monetization infrastructure for data owners by employing blockchain and smart contracts in the following aspects: (1) Legal rights of the data controller. The data controller has the legal authority to manage and use data in accordance with the agreed terms and conditions signed with the data owner in the smart contract. The management authority includes but is not limited to data access, data mining, and data modelling. (2) Obligations of the data controller. The obligations are mainly manifested in the duty of care, loyalty, and confidentiality. More importantly, the data controller must not harm the fundamental interests of the data owner. Each obligation term and the corresponding consequence can be clarified in the smart contract. (3) Income and commission arrangement. Smart contracts can design and allocate the value-added portion of data assets according to the principal’s wishes. By flexibly adjusting the commission ratio of agents, a balance of the equity structure between data owners and data controllers can be achieved, resulting in long-term and win-win cooperation.
(2) P2P-based Data Monetization
Peer-to-Peer (P2P) based decentralized data monetization enables data owners to enter into transactions directly with the party (can be an individual or organization) that demands data, without relying on a central authority. Badreddine et al. have explored the feasibility of a similar solution to trade user real-time data streams. Firouzi et al. propose an AI-driven data monetization approach that brings intelligent analytic and predicting strategies to profit from IoT data deluge. Compared to the agent-based model, data owners are the actual controllers of data and independent subjects in an open data marketplace. They can flexibly adjust their strategy of offering data according to the market supply and demand. Meanwhile, they can earn more revenue by saving the cost of agency commissions. However, data owners have to take the risk for fierce market competition. It is possible that they cannot even complete a transaction since buyers are always looking for high-quality and low-priced data.
In this model, smart contracts act as trusted intermediaries for transactions between data owners and data buyers. It mainly brings three benefits: (1) A high degree of autonomy. Once the buyer and seller reach an agreement, the contract deployed on the blockchain will be executed strictly and accurately in accordance with the embedded code logic (e.g., data port access permission opening, transaction settlement, etc.). Throughout the life cycle of the contract, people neither need to monitor the transaction process, nor can they interfere with the execution of the contract. (2) Fairness. Data buyers and sellers have equal rights/status throughout the transaction. The fairness is achieved from the beginning of the execution of the smart contract, where they have reached a consensus on the data and its price. Nonetheless, they are both able to terminate the trading at any time (e.g., the data seller cancels the transaction). Consequently, a penalty will be made in accordance with their agreements. (3) Trust. Smart contracts are transparent to all peers in blockchain. With the immutability and traceability of the blockchain, every transaction can be verified and audited, thus establishing trust between anonymous data buyers and sellers.
Although data monetization is a trend in the future era of big data, we should clearly see that at this stage, both the legal and technical aspects of data monetization are yet to be matured.
First, at the legal level, governments need to establish a more robust system in terms of data use restrictions, security and privacy protection, data transaction taxation, and transaction audition. Current data regulations such as General Data Protection Regulation (GDPR) in the EU and Consumer Data Rights (CDR) in Australia are still facing challenges. For example, in the implementation phase of the public policy, e.g., weak government supervision and law enforcement, controversial detailed rules, corporate boycotts, etc.
Second, at the technical level to enable the implementation of data monetization, the following aspects need to be further studied and improved:
(1) Data privacy. Sensitive and data personally identifiable information (PII) must undergo necessary privacy inspection and desensitization before entering the circulation data marketplace. If the data owner adopts the agent-based model, agents are responsible for balancing the tension and conflict between the privacy protection of the data owner and the tradable value of the data. There are many privacy protection techniques, such as differential privacy, multi-party computing, homomorphic encryption, etc. However, the key to employing them in our envisioned models is to find the balance between cost and utility.
(2) Scalability of blockchain. As limited by the consensus protocol, block interval time, and the size of block, current blockchain is not yet able to support highly concurrent Dapps (e.g., P2P-based model). On-chain and off-chain scaling are popular scaling solutions, while both are currently encountering bottlenecks.
(3) Data pricing. Data as a commodity has some unique properties that make the price of data required additional considerations. First, the marginal cost of data is extremely low. Second, the value of data is not necessarily related to the volume of data. Third, the value of data is very difficult to quantify. There is a need for reasonable and practical data pricing schemes to help both data owners and buyers evaluate the value of data.
In this article, we envisioned two promising models of data monetization by employing the technologies of blockchain and smart contracts. Meanwhile, we pointed out a few challenges as well as potential directions. We expect that this article can inspire IT professionals to generate more new ideas and innovations regarding data monetization in the future.