We are asked to build blockchain systems more often than we build them. The assessment we apply is narrow and it disqualifies most candidates: does the requirement involve multiple organisations that must share a record, where no single one of them can be trusted to hold it, and where reconciliation between their separate records is currently a real operational cost?
If a single organisation controls the record, a conventional database with append-only semantics, cryptographic audit logging and strong access control delivers the same integrity properties at a fraction of the operational cost. It is also far easier to hire for, to back up, to query, and to explain to an auditor.
Where the consortium condition genuinely holds — multi-party settlement, shared provenance across a supply chain, asset registries with several independent participants — a permissioned ledger removes reconciliation as a category of work rather than automating it. That is a substantial prize. It is simply not the situation most organisations asking for a blockchain are actually in, and saying so is more useful to a client than building what was requested.