Why It Doesn't Make Sense to Put Your Supply Chain on a Blockchain

Michael Ippolito | July 10, 2018

One of the most touted use cases for blockchain outside of payments is supply chain. There have been a couple of high profile supply chain pilots you've probably heard of, like Walmart tracking its pork or Maersk teaming up with IBM to create a blockchain-powered trading platform.

The idea is that in today's era of global supply chains, it is extremely difficult to track assets as they move through different players in the value chain. It is also a challenge to verify authenticity of certain assets, like making sure the tuna you bought in your local supermarket actually comes from where you think it does.

In some ways, at a high level, blockchain does make sense for supply chain. Blockchains remove the need for trust between transacting parties and it's immutable.

There isn't one COO in the world who wouldn't understand the benefits of increased transparency and accountability in her supply chain. The problem though is that this is high level, and in operations the devil is in the details.

And, as it turns out, there are just a host of issues with this entire use case.

It really goes beyond the current technical constraints of blockchain technology (i.e scalability, interoperability, etc...) as well. To me, it seems like a fundamental misalignment between the purpose of a blockchain and a supply chain. So let's begin by breaking down what that means.

 

The Fundamental Misalignment

Blockchain as a technology balances the efficiency of centralization with the security of decentralization. The reason bitcoin is such a secure store of value is that every block of transactions is validated independently by the roughly 10,000 nodes on its network. In other words, security is achieved by a (pretty massive) redundancy in computing power and resources.

There are other blockchain protocols (such as EOS) that have sought to strike a more even balance between security and efficiency by having fewer nodes in the network to validate transactions.  And to be fair, you absolutely could do this in a supply chain scenario. The point though is the fundamental innovation introduced by blockchain technology is transactional security that necessarily is achieved at the loss of some degree of efficiency.

Now let's take a look at the purpose of supply chain within a company. Operations as a function lives and dies over cost and efficiency. It's true that this varies slightly between industries (healthcare and pharma, for instance, might place a higher premium on quality) but generally speaking the marching orders are get the product where it needs to be, as quickly as possible, for the lowest possible price. Supply chains have generally achieved this goal over the past decade through centralization. It's often said in supply chain that there's a triangle between cost, speed, and quality, and you can only pick two.

The majority of companies have (and I know I'll get push back for saying this) gravitated towards cost and speed, with cost being by far the number one priority. Don't believe me? Just look at the number of organizations that have outsourced their manufacturing to China and Mexico. Again, there will be industries like food & beverage and healthcare that will take great pains to ensure quality, but for the most part cost is king.

Now let's look at this excerpt from an article in Forbes on how blockchain will transform supply chain. Here's a quote for you:

"There is not one central authority over the blockchain, and it is extremely efficient and scalable."

There are a lot of things to unpack here, but basically every part of that statement is untrue. Scalability is debatably the most significant problem for blockchain as an industry right now. Blockchains are also far less efficient than their centralized counterparts - that's exactly what makes them so secure. And in terms of lacking one central authority, is that really necessary here?

I guess my question is will a decentralized database recording transactions between players that are already cooperating result in so many benefits that companies will be willing to sacrifice cost and efficiency? I don't think they will.

So that's the fundamental misalignment. Companies will (for the most part) prioritize cost and efficiency, especially as it pertains to supply chain. Cost and efficiency are not what blockchains do best. And yes, if you had 100% transparency and accountability you wouldn't need to cost in counter-party risk. You could also save on time and cost associated with reconciliation. Theoretically, a blockchain could take care of these issues, but I don't know how you really make the argument that they're increasing efficiency or improving cost in a meaningful way.

Now let's move past the whole misalignment issue and address another part of this argument: the massive technical problems in a blockchain supply chain solution.

 

Some Massive Technical Problems

Let's begin with the most pressing issue, which we'll call the 'physical to digital problem.' In other words, you hear a lot about putting real world assets onto a blockchain but absolutely nothing about how this can actually be accomplished.

To illustrate this challenge, let's return to our previous example of Walmart putting its pork on a supply chain. How, exactly, is Walmart going to execute here? Is each pork-chop going to have an RFID strip that corresponds to its digital representation on the blockchain? Let's say that's the case - has anyone given any thought to how much that's going to cost? Can Walmart build in a personalized tracking system for each individual piece of pork and still hit their required gross margin?

Now let's give Walmart the benefit of the doubt: say they can build a system that accurately tracks each piece of pork at cost. How are they ensuring that the data is entered into that system correctly? Some studies estimate the error rate of manual data entry at 30%. You could argue that the real point of failure from an accuracy perspective in a supply chain is improper data entry, a problem blockchains could potentially exacerbate.

But (again with the benefit of the doubt) let's say that companies were capable of near perfect data entry. You would still be faced with the oracle problem, which is the idea that you would need some sort of list capable of verifying that your real world assets match up with their digital copies on the blockchain. That list represents a central point of failure, which pretty much cuts the value of a blockchain solution in half.

Finally, there are still no established standards for industry participants to work together on. Lack of interoperability is one of the most cited issues facing blockchain adoption today (particularly within supply chain) so I wanted to address it here, but in my opinion it's less concerning than the other issues mentioned above. Industry participants will create standards if there's enough incentive, they just need time. There's good progress being made on this front already actually. But a list of issues wouldn't be complete without at least mentioning standards so I wanted to include it.

 

I'd Love to be Proven Wrong

I'm a big proponent of the blockchain & crypto movement, so I would love to see a supply chain solution that makes sense. However, I think for a blockchain solution to be viable you would at least need to solve these issues.

1. The Digital to Physical Problem - What is the exact mechanism (some combination of IoT and RFID?) that you're using to digitally track all of the physical assets moving through your supply chain. You'd also need to figure out a workable solution to the oracle problem.

2. Cost - You've got to create a solution to the first issue that works at cost. From an ROI perspective, no executive is going to implement a supply chain solution until there's a business case. In this instance, I would imagine you'd need to be able to demonstrate that elimination of counter-party risk and reconciliation driven cost savings would outweigh the implementation costs of a blockchain plus the cost of putting those assets on said blockchain. I'll be honest, this one seems really far out to me.

3. Buyer Education - I think the right way to conceptualize this problem is through the eyes of a an executive who would buy a blockchain solution for their supply chain. That's really how these types of software technologies tend to get integrated into corporations. So someone who's selling a supply chain solution like Corda would need to A) find a company officer who understands blockchain, B) be able to convincingly demonstrate a business case and C) convince them that they could get their vendors on board.

4. Standards / Interoperability - Again, I think this will be solved with time. There's no denying though that right now it's an issue and it's standing in the way of any implementable blockchain solution.

Even if you nail all four of those though, you're still looking the misalignment between security and efficiency dead in the eyes.  Blockchain is really trust as a software primitive, and it's best use cases will be in environments where no trust exists (i.e payments and banking).  In a supply chain, some level of trust does exist (yes it could be better) but to get to the level of trust you need there are other database and software solutions that are far more efficient and cost effective.  I just never really see blockchains being more than a 'nice to have' solution. 

All that being said, I do look forward to hearing all about why I'm wrong in the comment section!

Join our no fluff newsletter

View All News Posts

Michael Ippolito

Co-Founder of BlockWorks Group