Many projects reach for a blockchain because it is new. NIST's advice is to understand the technology first, see where it fits, and use it only where it is appropriate. These questions come from NIST's overview and the decision guides it points to.
Information, not advice.
A blockchain is for a record that many participants add to, often in different places. If one organisation writes all the records, an ordinary database does that job.
If everyone already trusts one keeper, such as a bank, a registry or a government department, that keeper can hold the record. A blockchain is for when the parties do not want to rely on a trusted third party.
Do the parties each need the full history, every past change, and to know that none was quietly altered? A blockchain keeps the whole history, shared among them.
If the record is about passing something from one party to another (money, goods, ownership), that also fits how a blockchain works.
On an open network the data is generally public. On any blockchain, records cannot be removed. Personal data, or anything a law may require to be deleted, may not belong on one.
If the record is about the real world, such as goods, sensor readings or people, someone still has to put it in truthfully. The blockchain cannot check it.
Blocks on many open networks take seconds, sometimes minutes. Can the job wait that long for each record to be written?
NIST points to several published guides. One, “Do you need a Blockchain?” by Karl Wüst and Arthur Gervais of ETH Zürich, gives a flowchart. NIST notes that most of its paths lead to “no”, and only a few to “maybe”. NIST also reproduces a flowchart from the US Department of Homeland Security.
Their shared advice: look first at whether existing technology solves the problem better, and use a blockchain only where it fits.
The Wüst and Gervais flowchart itself was not read for this page: the archive refused to open the paper's PDF on 21 Oct 2026, so it is described here in NIST's words.