Consortium ledgers do not fail on technology. They fail on governance
We have now delivered enough permissioned networks to say with confidence that the hard problem is never consensus algorithms. It is the question of who is allowed to add a member, change a schema, or reverse an entry, and what happens when two members disagree.
The projects that stall are the ones that leave this to a side agreement, signed by legal teams, stored in a shared drive, and interpreted differently by every member within six months. The projects that ship encode the rules as chaincode, with an explicit voting threshold, and treat a governance change as a deployment.
Eleven companies can agree on a data model in a fortnight. Agreeing who may change it has taken some consortia two years.
That has a pleasant side effect: once governance is code, it is testable. You can demonstrate to a regulator exactly what would happen if a member attempted an unauthorised change, rather than describing the policy that says they should not.
If your consortium cannot answer “who approves a schema change, and by what majority” in one sentence, that is the work to do before anyone writes a smart contract.
Leave a Reply