Read views of Infosys experts on how blockchain technology offers an unprecedented opportunity to transform the transactions of the future, how its adoption will create newer value propositions and what is required for its integration into larger IT systems.

« How Machine Learning and Blockchain Can Improve Supply Chain Management? | Main | Business Rationale for "Paper Boat" Projects in Blockchain »

Analytics on Distributed Ledger (Reducing noise in the Value Chain)

I was reading a 2016 HBR article from Daniel Kahneman (et al) about the hidden cost of inconsistent decision making (refer notes for details), on my flight back to India last week, and wondered whether Blockchain could reduce disparity of view-points held by different stakeholders within a firm and across the entire value chain in which it operates. Firms are operating in an increasingly connected ecosystem with shared value; having consistency of view-points or decisions may have a big impact on driving multiparty collaboration.

One of the reasons why different stakeholders of the same value-chain come up with different outlook about the same situation is that, they are referring to different data sets for their analysis to start with.

Take for instance a re-insurance case in which the insurer has spread a particular risk across three different re-insurers. Now, each of these parties- insurer and re-insurers maintain their own ledger summarizing- claims, premiums, tax etc.

Each of these parties maybe using different tools to summarize their financial position, but more importantly each of them may also be potentially using different datasets to generate their financial summary, leading to inconsistent status for the same risk.

There could be several reasons for having multiple versions datasets to start with-

They may have been generated for different time periods; may have been extracted with different filter conditions; may have been consolidated differently by using different cleaning, imputation and aggregation techniques; or could be that one party had a more recent information than anybody else (transient information asymmetry). In short, the cause is largely on the ETL (extraction transformation and Loading) side.

Here is the case for transient information asymmetry- the insurer, being in direct contact with the insured will always have firsthand information about any new claim; this information may take time to permeate down to the re-insurers. In-the-mean-time their claims summary report will not tally.



At t1 a new claim comesàInsurer updates its ledger. But all re-insurers disagree. (Information asymmetry)

At t2 Re-insurer 1 and Reinsurer 2 gets updated information, but Reinsurer 3 is still having the old version. (Information asymmetry)

At t3 Reinsurer 3 updates its ledger and everyone is in sync.

Similarly, two Bank agents working on different versions of a customer's data will come up with different credit score or compute different applicable interest rate for the same customer. This discrepancy will cause severe damage to the trust and loyalty bestowed by the customer.

In global supply chain, a customer may get different status of goods in-transit from the supplier and other multi-modal logistics partners. This discrepancy may eventually get resolved but not without significant reconciliation effort- administrative overheads and payment delays (often leading to detrimental cash-flow pressures).

Having a source of data that gives a consistent view to all the participating value chain players is key to consistent decision-making. Enter Blockchain (or for that matter any DLT)...

In this Blog I have used Corda to demonstrate this idea of consistent reports for all value chain partners (but you can use any framework or platform).

Corda allows point-to-point messaging- only those members, who are party to the same flow have access to the shared ledger. This shared ledger forms the basis for Business Intelligence reports and Analytics that are uniform across participating members of the Value chain. [It is not necessary to comprehend the architecture of a corda node to appreciate the essence of reporting and analytics on Blockchain/DLT. Feel free to skip this section.]

A node (could be a computer maintained by individual participants of the network) is a collection of processes, a Vault, which contains the output state relevant to a party and a Transaction Storage that has key-value store for attachments, transactions, and serialized state machines (SSM).



The figure (modified from Corda documentation) depicts the Corda node and where BI reports can be plugged.

Re-insurers and the insurance company sharing the same Risk will have a common ledger on which they will query.

The nodes can choose to put this shared information (for a given point in time) in its Vault. An API can expose this vault data to the outside world to be consumed by BI or Analytics tools such as Tableau, R or Python.


The figure depicts a sample Tableau report created from the Data extracted from Corda node. As the report is generated on the shared ledger, every participant (with the privilege to view the data) will always have matching numbers about status of claims on common policies.

Here is a graphical representation of the flow of materials from a Supplier to a Customer through a multi-modal logistic channel. The nodes/circles represent the participants; and the weights of the connected arrows represent the volume of goods passing through them (or handled by them). These kind of visualizations can easily be made using tools such as R/Python.

Stakeholders can use these insights to determine critical transport partners and formulate a incentive strategy or evaluate integration options.


The above figure has been drawn using power point for illustration purpose only; similar visualizations can be easily created in R.

As pointed out in the HBR article- noise may be difficult to identify and may be observable much later when very little can be done. For instance a re-insurance company may only come to know about a particular risk (that it didn't want to keep in the first place) only after a large claim comes knocking at the door; or a Bank may come to know about an inconstant credit risk calculation only after the furious prospect calls customer service.


Having a blockchain alone is not enough unless managers can extract insights from it.

A self-auditing network, with codified autonomous business rules may reduce the need for noise audits (alluding to the article again) and improve the overall quality of decisions made by managers; amplifying efficiency in the entire value chain. That makes a case for Reporting and Analytics on Blockchain.


Daniel Kahneman is psychologist who was awarded the Nobel prize in economics in 2002 for his contribution in behavioral economics. 

HBR article referred in this Blog is, "Inconsistent decision making is a huge hidden cost for many companies. Here is how to overcome what we call Noise" by Daniel Kahnemal, Andrew M. Roenfield, Linnea Gandhi and Tom Blaser, published in HBR October 2016 edition.


Post a comment

(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)

Please key in the two words you see in the box to validate your identity as an authentic user and reduce spam.