We build blockchain systems for a living. Most companies that ask us for one do not need it, and we tell them so in the first call. This article is that first call, written down.
The one property blockchain actually sells
Strip away the vocabulary and blockchain offers a single property no ordinary database has: several parties who do not fully trust each other can share records that none of them can quietly alter. Every claimed benefit (transparency, immutability, verification) is that one property wearing different clothes.
That property is expensive. You pay for it in infrastructure, in slower writes, and in design constraints. So the entire decision reduces to one question: does your problem involve multiple organisations that need shared, tamper-evident records?
When the answer is no
If the records live inside one company, the answer is no. Your inventory, your CRM, your accounting, your internal approvals: these need a well-designed database, access control, and backups. A blockchain would add cost and remove flexibility in exchange for a guarantee nobody asked for, because you already trust yourself.
A useful test: name the second organisation. If you cannot name another party who needs to trust your records without trusting you, close the topic and spend the budget on the actual bottleneck. We have redirected clients from a blockchain request to a database redesign that cost a fraction and fixed the real problem.
When the answer is yes
Credentials are the cleanest example. An employer checking a certificate should not have to trust the issuer's database, and the issuer should not be able to quietly edit history. We built a digital certificate platform on Hyperledger Fabric running on AWS Managed Blockchain for exactly this reason: more than 50,000 certificates issued, each one independently verifiable through a public explorer, with 99.9% uptime.
Other fits with a genuine second organisation:
- Supply chains, where manufacturers, shippers, and buyers each hold a piece of the story and disputes are settled by the record
- Asset registries, such as tokenised land titles, where ownership transfers must be provable across parties. Our team has built an Ethereum smart-contract layer for a decentralised real-estate platform: land-title NFTs, on-chain KYC, and multi-signature approvals
- Multi-party audit trails, where a regulator or consortium needs confidence no member edited the log
Private or public chain
If the parties are known organisations (universities, insurers, cooperative members), a permissioned network like Hyperledger Fabric is usually right: no gas fees, controlled membership, data visibility you configure. If the system must be open to strangers and survive without a gatekeeper (public asset tokens, open verification), a public chain like Ethereum is the tool. We build both, and the choice falls out of the trust model in about ten minutes of discussion.
The honest summary
Blockchain is a specialised tool for a specific trust problem. When the problem is real, it works, and we can show you production systems that prove it. When it is not, the best blockchain advice money can buy is "use a database," and any agency unwilling to say that wants your budget more than your outcome.
Unsure which side your idea falls on? Describe it to us. The answer usually takes one conversation, and we will give it to you straight either way. The full picture of what we build is on our blockchain development page.