When does a blockchain actually make sense?

Most blockchain projects that fail do so for a simple reason: the problem never needed one. A distributed ledger is slower, more complex and more expensive to run than a well-designed database. It is worth that cost only in specific conditions.

Before we recommend one, we work through five questions with the client.

1. Are there several parties who need to write to the same record?

If a single organisation owns the data and everyone else only reads it, a conventional database with good access control is almost always the better choice. Ledgers become interesting when a farmer, a cooperative, an exporter and a certifier all need to add to one history.

2. Do those parties lack a trusted intermediary?

Where a respected central body already exists — a regulator, a clearing house — and everyone is happy to rely on it, adding a ledger mostly adds cost. Where no one wants to hand control to a competitor, a shared ledger can be the neutral ground.

3. Does tamper-evidence matter to someone outside the system?

Buyers, auditors and regulators increasingly want to verify claims about origin, impact or process rather than take them on trust. If proof has value to an outside party, an immutable, time-stamped record starts to earn its keep.

4. Can the physical world be linked reliably?

A ledger only guarantees that data has not changed since it was written. If the data going in is weak — a hand-typed weight, an unverified photo — the ledger simply preserves the mistake. We spend as much time on sensors, identity and field processes as on the chain itself.

5. Is there an economic model that keeps it running?

Somebody has to pay for nodes, integration and support after the pilot funding ends. We look for the party that captures the most value — often a buyer paying a premium for verified supply — and design around them.

If the answer to the first two questions is "no", we recommend a database. It is the cheapest good advice we give.

What this looks like in practice

For a traceability project, three or more "yes" answers usually justify a pilot. We then start with the smallest useful slice: one commodity, one corridor, a handful of participants, and a success measure such as reduced time to produce compliance evidence.

If you are weighing up a ledger for your industry, we are happy to run this test with you. Get in touch.