Should you build or buy Enterprise Blockchain Platform?
Enterprise blockchain platforms give consortiums, banks, and market infrastructure firms a permissioned ledger, ordering and consensus layer, membership and identity services, and a smart-contract runtime for keeping shared records in sync across participants without one central database. Buyers include financial market utilities, trade networks, and multi-party groups that need a common, auditable ledger.
Investigate before you renew; the runtimes are open source, the production networks that matter run them in-house, and the managed vendors are selling node operations and governance tooling rather than anything you cannot run yourself.
Build it, buy it, or bridge?
When building makes sense
Running your own permissioned network makes sense once you have production precedent to point to and the operations staff to match it. DTCC, J.P. Morgan, and Swift all run their tokenization and settlement networks on open-source runtimes rather than a vendor's managed platform, and IBM itself ended its managed Blockchain Platform, now selling support for customers who run Hyperledger Fabric themselves. If your organization already runs Kubernetes at scale and has staff who can own key management, certificate authorities, and consensus configuration, self-hosting Fabric or Besu costs little beyond engineering time. What self-hosting doesn't remove is HSM operations, member onboarding, and the governance work a consortium still owes regardless of who runs the nodes. Build only after pricing that operations burden against a managed vendor's fee.
When buying makes sense
Buying still earns its place for a network member that wants to join without hiring dedicated blockchain infrastructure staff. Kaleido, AWS Managed Blockchain, and Oracle price by the node-hour, so entry cost stays low and the vendor absorbs patching, uptime, and certificate rotation. That fits members who touch the network occasionally rather than the institutions running its core. R3 Corda buyers get a packaged runtime plus support instead of bare open-source code, which suits teams that don't want to own Fabric or Besu configuration at all. The case weakens at scale: as node count and transaction volume grow, per-node-hour pricing gets expensive relative to running the network yourself, and a managed vendor still can't take on your consortium's own governance and member-onboarding work.
The desk read
Build-versus-buy analysis for Enterprise Blockchain Platform is being written. In the meantime, the framework that drives every B4 call is on the B4 Index page.
Frequently asked
Why isn't Enterprise Blockchain Platform a clear buy?
The runtime itself is undifferentiated and buildable on open-source code, but IBM's retreat to a support-only model and the real operations burden of self-hosting mean neither path is obviously right. Price both before renewing a managed contract or committing to build.
Can we self-host a permissioned blockchain without an internal blockchain team?
Not realistically. DTCC and J.P. Morgan's Kinexys run on open-source Besu, but each is operated by an institution with its own infrastructure organization. There's no managed shortcut to that scale of in-house operations.
What do vendors like Kaleido or R3 Corda actually sell if the runtime is open source?
Node hosting, key management, consortium governance tooling, and support contracts layered on Fabric, Besu, or Corda. You're paying someone else to run the operations, not for code you couldn't get elsewhere.
Is Hyperledger Fabric still supported if we're already running it?
IBM ended its own managed Blockchain Platform on April 30, 2023, and now directs customers to run Hyperledger Fabric themselves, with IBM providing paid support rather than a hosted platform.