By José Malpica, Oracle Insight
Blockchain is one of the hottest topics in IT. Although public attention peaked in December 2017 (along with most cryptocurrencies), companies big and small are looking to apply the technology at every opportunity as if it were a silver bullet. On the other hand, the cryptocurrency crash has encouraged harsh criticism to the point of questioning the value and applicability of the technology to any real business problem.
How can a company sidestep the hype and the gloom to determine whether a specific use case is better implemented using blockchain rather than another data management solution?
At Oracle Insight, we have developed a simple framework that can be used to evaluate whether a blockchain solution is superior based on six qualifying criteria. For each criterion, a blockchain sweet spot is defined against which the use case is evaluated. The final score depends on the weighting (including knockout characterization) that your company must define. Here are the criteria:
1. Data sharing and contribution scheme: Blockchain was conceived to enable multiple, potentially anonymous, participants to exchange data without a central, trusted intermediary. Therefore, blockchain solutions are most applicable when large numbers of interdependent parties with limited knowledge of each other must update a shared dataset.
2. Trust and verification requirements: When parties in an exchange trust each other, the need for verification is not acute. Conversely, low-trust environments require intensive verification before a transaction can be validated. Blockchain creates the most value when conventional solutions require expensive and/or slow verification to validate transactions.
3. Intermediation: Using a trusted party, such as a clearing house, is often the only feasible way to guarantee the terms of an exchange. Blockchain was conceived to eliminate the need for intermediaries under certain conditions—for example, when the irreversibility of transactions is acceptable. If existing intermediaries are expensive or untrustworthy, blockchain solutions are likely to bring very substantial gains.
4. Time sensitivity of transactions: Some interactions must be completed in milliseconds—for example, trades in the stock market—while others have much-less-stringent requirements. Most consensus protocols used to validate transactions in a blockchain require a nonnegligible, and often widely variable, time to complete—sometimes minutes or even hours in peak traffic. Blockchain works best if latency requirements are lax.
How can a company sidestep the hype and the gloom to determine whether a specific use case is better implemented using blockchain rather than another data management solution?“
5. Transaction (meta)data: A blockchain can be an expensive mechanism for exchanging information, because data is replicated to and validated by multiple (sometimes all) nodes in the network, multiplying processing and storage needs. If each transaction requires a lot of data to be exchanged and stored for it to be validated, it is likely that centralized solutions will have a very significant performance advantage. However, there may be ways to reduce the data storage requirements on the blockchain. For example, a company can store only the metadata of a transaction as a means of validation. In any case, a blockchain will work best when all data related to a transaction can be stored in it.
6. Automated decision-making: Information is most often exchanged as the basis for executing some business logic. Conventional client/server architectures have made it possible to do this efficiently for decades. Blockchain, though, brings a crucial innovation: the smart contract (SC). An SC executes business logic in a trusted and decentralized manner within a blockchain, providing predictable results on any and every node of the network. An SC can, hence, be used to encode and automatically enforce the terms of a transaction between the parties, eliminating the need for interpretation and enforcement by an external party. This wonderful benefit is accompanied by heavy trade-offs. The SC code must be severely constrained through a restrictive programming language so that it can run securely on every node without exhausting computational resources. Knowledge of SC programming languages is still very scarce, which compounds the problem that once released, an SC cannot be updated. Therefore, any programming errors can be very costly. In practical terms, this means that current SCs are somewhat limited in the complexity that they can handle. In addition, SCs results are meant to be irreversible, which may make them inadequate for many real-life situations. Despite these limitations, SCs provide a powerful programming framework that can replace many current business arrangements by executable code that, when invoked, embodies and automatically enforces a contract between the parties. At the current stage of development, an SC-enabled blockchain will create the most value when making simple, nonrecourse business decisions based on transaction data that is slow and/or expensive.
Let’s apply this framework to two apparently similar use cases in which blockchain-based solutions have been widely discussed: domestic payments and international money transfers. The following figure summarizes the evaluation of each criterion and highlights which ones are critical for a real-life implementation.
In both cases, the number of participants potentially sending or receiving money can be very large (1). Verifying the validity of transactions (2) is rather more difficult in international transfers where a chain of banks is involved, as opposed to simpler domestic payments structures. However, the first key difference arises when comparing the current cost of intermediaries (3). Although domestic payment solutions in some countries are notoriously expensive, the Chinese market shows that this need not be the case: WeChat and AliPay have extremely low costs while providing a high-quality service. International transfers traverse multiple interconnecting banks, accumulating fees that add an average of 7% to the cost of transaction, according to the World Bank.
Leading providers execute thousands of payments per second with millisecond latency (4). This is a critical requirement, because payments require very fast confirmations to complete the checkout of the goods. This is much less so in international money transfers, where customers often wait for several days to see the transaction completed. Very little data is usually required to completely describe an international transfer (payer, payee, sending account, receiving account, amount, currency, timestamp), so it can all easily be recorded in the blockchain (5).
A payment may require some more information, including invoice ID, goods ID, purchasing terms and conditions, licensing requirements, and shipping address. In terms of making simple, nonrecourse decisions (6), legal requirements in many countries require payment transactions to be reversible in many situations such as online purchases. International transfers have much-less-stringent requirements to this effect, while the current chain of mediating banks makes it complex and expensive to automate decisions.
The conclusion of this brief analysis is that domestic payment value chains are unlikely to be disrupted by blockchain-based solutions, but international-transfers providers will probably be facing serious challenges from blockchain competitors in the near future.
Our simple framework must be adapted to the concrete circumstances of your company’s operating environment. These examples show how it can be useful in identifying threats and opportunities, as well as in making blockchain investment decisions on the basis of objective analysis rather than on the vagaries of the hype cycle.
Work with our Oracle Insight strategy consultants to uncover the power of technology.
Watch this on-demand webcast: “Introducing Oracle Autonomous Blockchain Cloud Service”