Assessing whether it is warranted.
WHAT TO ASK FIRST
Who are the participants? Do they trust each other? Is there an acceptable central party? Does the data need to be shared? Does it need to be immutable?
WHAT THE DECISION USUALLY REDUCES TO
Whether a trusted intermediary exists and is acceptable.
WHAT IF ONE DOES
A database is simpler, faster, cheaper and correctable.
WHAT IF ONE DOES NOT
A ledger may be warranted.
WHAT OTHER CONDITIONS SUGGEST IT
Several parties writing to shared records Disputes about what happened A need for verifiable history Value transferred between parties without an intermediary
WHAT TO BE SCEPTICAL OF
Claims of transparency, where a published database would do Claims of security, which depend entirely on design Claims of cost saving, without accounting for operation Proposals whose benefit is being blockchain-based
WHAT QUESTIONS EXPOSE WEAK PROPOSALS
What happens when data is entered incorrectly? Who decides what the rules are? Who pays transaction costs? What happens if the network becomes expensive? How does data from the real world get in?
WHAT THE LAST ONE USUALLY REVEALS
A trusted party doing the actual work.
WHAT TO RECOMMEND HONESTLY
Frequently, something simpler.
