Supply Chain: You (Probably) Don't Need Blockchain For It, Part II.
In the first part of the series, I explained the basic principles behind blockchain and cryptocurrencies, that anyone who considers implementing these concepts should understand. This is the first of a few follow-ups, which will apply these principles to different domains.
Supply Chain is unsurprisingly one of the most quoted examples of possible applications of the blockchain, because of its traceability of all transactions. It is much more difficult to grasp the cryptography aspect of the blockchain, let alone to understand the concept of the proof of consensus, but any layman gets (or thinks they get) the idea of being able to trace every transaction to its origin.
Walmart boasts here about their partnership with the big blue and how thanks to blockchain they’ll be able to trace all of their goods all the way back to their origin. However, if that’s their goal, there is absolutely no reason to use blockchain and actually many reasons why not to use it.
In their case, the real challenge isn’t in maintaining the ledger, but in
- Integrating all parties involved in the chain and;
- Tracing each of their product as it breaks down from a commodity (a ton of mangoes) to an SKU.
Both are process and technological challenges. The integration requires the agreement of thousands of parties to join a single platform, which means facing corporate bureaucracies and resistance of various vested interests. From a technological point of view, it is a fairly straightforward problem and a good centralized integration platform with well-documented APIs will do the job much more efficiently than slow and expensive-to-run-and-maintain blockchain. The only advantage blockchain might bring is, that it probably sounds hot and sexy enough to remove some barriers, which have accumulated over a decade since SOA was the buzzword on the street.
Juraj (whom I quoted extensively in the first part) pointed out to another possible explanation - while some vendors might be reluctant to integrate what they (rightfully) perceive as “Walmart’s database and integration platform, giving them access to a blockchain, which then gets replicated to their node might change the perception and make then think of it as a jointly owned solution, notwithstanding the fact, that Walmart still gives them specifications of the blockchain and permission to access it. If that’s really their thinking, they’re in fact shooting themselves in the foot because they have to bear the costs of running the node without any added value of truly permissionless distributed blockchain.
As I pointed in the introduction to this series, much bigger challenge - as in any blockchain application - is in the connection between tangible stuff and digital records, i.e. how to record every move, handover and repackaging of every good as it flows from farms through commodity traders, shipping companies, wholesalers, warehouses and stores and is combined, split, removed for not meeting quality requirements, labeled, etc. and all of that without imposing ridiculous burden on the people involved and with acceptable error rate. That’s where IoT might come in, but unless Walmart is able to force every supplier to put NFC chip on every mango the moment it is picked at the farm, it is again about mapping and redesigning yours (and mainly your vendors’) processes and their efficient reporting. Blockchain doesn’t help in this at all.
Where cryptocurrencies, even without blockchain can really help is in lowering transaction costs. Bank transfers are slow and suppliers often cannot be paid as fast as required. One example Juraj shared with me was a case of a friend of ours, who has a company that wanted to manufacture a machine in China. There were four steps quality verification checkpoints in the manufacturing process, followed by a payment. The supplier would not continue until the payment went through, which in the legacy systems took a month between Europe and China. By using Bitcoin, he saved 4 months, which translated to a lot of money, because the machine, that had been manufactured was a mean of production and as a result, it could produce 4 months earlier. When capital is not locked-up in slow wire transfers, the supply chains can become more efficient.
I don’t want to dismiss the case for blockchain altogether. Realizing its function as a distributed ledger, and attaching a necessary cryptocurrency as means of exchange, one can see more clearly how it could be applicable. Open and distributed blockchain tailored to an industry could indeed minimize the necessity to trust a new party one engages with. If it’s immutability of transaction records I’m looking for, a blockchain upheld by a large group of incentivized miners is indeed something I can put my trust in. If I’m looking to engage with a new customer, having an immutable settlement mechanism clearly lowers my cash flow risk.
Even if Walmart’s goal was to increase the trust within the whole supply chain, the implementation does not adhere to that goal. As I explained in the introduction, the necessity of trust is removed by consensus and that is achieved by distribution. The distribution needs to be allowed and incentivized. Private blockchains with no cryptocurrency attached don’t allow, let allow incentivize that to happen.
The winners in proper applications of blockchain and cryptocurrencies to the supply chain will definitely be smaller vendors and customers, but also Walmarts of the world. The losers will be likes of IBM who don’t get the concept of openness and will be caught by surprise from the smaller guys who do.