How much to decentralise?
Steve Pomfret, CEO of Cygnetise, discusses the landscape for fintechs providing blockchain solutions when deciding on the optimum degree of decentralisation for products and markets.
Decentralisation is the name of the game. DeFi, Web 3.0 and Central Bank Digital Currencies are but a few significant examples of blockchain technology being explored and applied. At the other end of the scale, startups are creating solutions using distributed ledgers that target specific business problems or sectors.
While there’s a trend for decentralisation, there are different degrees, arguments for and against these, and varying suitability for different use cases. For startups utilising blockchain for SaaS solutions there’s a host of choices to be made, and when you’re selling into financial services companies, things become more complicated.
Balance needs to be struck between practicality and innovation, and with technology changing so rapidly, an eye kept on progress to be aware of where it’s heading to avoid technical debt. Products need to stand the test of time.
Nodes
The COO of one blockchain startup described 100 nodes, hosted by 100 people on five different cloud services as their ideal world. In reality, the hosting cost of this would be prohibitively expensive and there’s no financial incentive for people to take ownership of nodes. It’s not necessary to live in an ideal world, though. Clients, want effective, trustworthy products at a reasonable price. If it’s secure, reliable and solves their problem, you have a market.
Involving more than one party in the ownership of nodes helps mitigate the risk of a single owner being unable to maintain payment for the nodes, resulting in the loss of the blockchain. Still, one individual may have the power to switch it off or forego a payment. Multiple owners could collude to take over a ledger, an issue shared by permissionless blockchains as we see from the numerous discussions about Bitcoin and a 51% attack. There is no perfect world here, but solid terms and conditions provide protection for customers.
Cloud Hosting
Splitting hosting across different cloud services offers greater decentralisation, added cybersecurity, and reduces reliance on a single cloud provider. On the other hand, it increases costs and complexity. While the risk of the likes of AWS, Microsoft Azure, etc. going under is slim, it’s still a consideration.
Sectors and jurisdictions have their favourites, though. Banks generally prefer AWS or Microsoft Azure over Google Cloud. The Middle East tends to prefer Microsoft Azure over AWS and Luxembourg prefers Azure and their own closed service. Geographical and industry preferences are things to bear in mind.
Latency
Depending on the problem being solved, speed is more or less of an issue. For transactions that can be queued, latency is less of an issue. The more nodes and the more separated they are the slower they work, so for time-critical instant use cases like payments or trading, this is inappropriate. However the fewer the nodes and the closer they are, the less decentralised they are, stacking a few nodes on top of each other isn’t decentralisation.
Consensus Protocols
Back in 2016, there were a handful of consensus protocols to choose from. Today there are more proof-of-somethings than you can shake a memory stick at. While consensus protocols don’t directly affect levels of decentralisation, nodes do, and the number of nodes can be influenced by the consensus protocol.
While we’re here, with ESG considerations of growing importance, the energy consumption of different protocols matters. Proof-of-work is outdated and uses eyewatering amounts of energy, but newer mechanisms can use minimal energy, and certain DLT solutions also help environmentally, by reducing paper usage, for example.
Governance
Financial institutions need robust network governance models that safeguard who can enter, read, and write on a blockchain and who has control over nodes, prohibiting full decentralisation to safeguard sensitive information and ensure regulatory compliance. There’s a growing array of choices when designing the system of rules and practices that control a network and its participants.
Open-source enterprise distributed ledger frameworks provide an increasing number of ‘pick’n’mix’ options to help simplify application creation and tailor it for each use case. Some are more decentralised than others, and all are constantly developing new features. As strong governance is such a vital component for blockchain fintechs, should the level of decentralisation of enterprise frameworks be the highest priority when selecting which one to use? Safeguarding who is granted permission as an administrator is arguably the key factor here.
So, the overarching trend is for decentralisation, interested parties may ask about decentralising tools further, and purists may balk at private blockchains, but it’s about being fit for purpose, not decentralising for the sake of it. Balance lies in meeting your specific demand at a viable cost. Why build a spaceship and offer trips to the moon when nobody is even airborne yet and it’s prohibitively expensive? People are far more likely to pay smaller amounts for more modest, lower-risk trips closer to the ground. No doubt demand will increase to get closer to the stars over time, but it’s not always necessary to push for ever greater decentralisation.