For business · 2026-07-17 · 7 min read
Ask a finance team what their cloud storage costs and they’ll quote you a per-gigabyte figure that sounds reasonable. Ask an engineering team what it would cost to leave, and you’ll get a much longer, more uncomfortable answer. That gap — between what storage costs to keep and what it costs to move — is the real economics of the modern cloud, and it’s worth understanding before it becomes your problem.
Where the lock-in actually lives
It’s rarely one big fee. Lock-in accumulates quietly, in ways that each seem minor at signup and add up to a moat later.
- Egress charges: storing data is cheap; moving it out is billed per gigabyte, so the more you accumulate, the more leaving costs — exactly when you’d most want to.
- Proprietary surfaces: bucket policies, access models, and provider-specific APIs that your code grows around until “just switch providers” means a rewrite.
- Operational gravity: monitoring, permissions, and integrations all wired to one vendor’s console, so the switching cost is organizational, not just technical.
- Pricing power: once moving is expensive, price increases are easier to pass on. You’re not really free to shop around anymore.
None of this makes the big clouds villains — the tooling is genuinely excellent, and for plenty of workloads it’s the right call. But it concentrates control and pricing power with one company, and it’s worth naming that clearly rather than discovering it during a renewal negotiation.
What “not locked in” looks like
A decentralized network changes the shape of the relationship. On a network like Walrus, your data isn’t held in an account that a provider administers — access is bound to keys you control, and the data is spread across many independent operators, none of whom can hold it hostage. There’s no single company to raise egress fees, suspend your account, or decide your workload is no longer welcome. If you want the honest, non-tribal comparison of how this stacks up against S3, IPFS, and Arweave, we laid it out in Decentralized vs. Centralized Cloud Storage.
The pricing model is different in a way that matters for planning: you pay the network for the space and time you use, in its WAL token, directly — not a provider’s margin, and not a penalty for reading your own data back. For teams that have watched a storage bill balloon with usage they can’t easily reverse, that predictability is the point. You can put real numbers against it on the pricing page.
The right question isn’t “how much does this cost to store?” It’s “what will it cost me to change my mind?”
A low-commitment way to evaluate it
You don’t need a migration project to form an opinion. Pick one dataset — a backup, an archive, a set of public assets — and store it on the network yourself. Walrus Drive lets you do that straight from a Sui wallet, with the cost shown before you approve anything and no account to create. If your team is weighing this for something specific — research data, media, compliance archives — the solutions overview walks through where it fits and where it doesn’t, and Cold Archives covers how retrieval and minimum-duration rules change the arithmetic on long-term data. The freedom to leave is the feature. It’s worth testing before you need it.