This page is general information, not legal advice. An escrow agreement is a contract; have a lawyer review it.
How it works
Three parties are involved. The depositor is the vendor. The beneficiary is the customer. The escrow agent is the neutral third party that holds the material.
- The vendor and the customer sign an escrow agreement that names the release events.
- The vendor deposits the source code, with the build instructions and anything else needed to rebuild the product.
- The vendor updates the deposit as the software changes, on a schedule set in the agreement.
- The agent may check the deposit.
- If a release event occurs, the agent hands the material to the customer, who may use it to keep the software running under the terms of the agreement.
One provider’s description fits this. It says escrow is the deposit of source code with a third-party agent, protecting the vendor’s intellectual property and the beneficiary’s ability to maintain the software. It says escrow is typically requested by beneficiaries to safeguard business continuity, that code is released only if the depositor is unable to support and maintain the software, and that deposit and release mechanisms are defined and negotiated in the agreement. That provider runs a file integrity test on every deposit and accepts deposits from Git platforms or by SFTP and S3.
Types of agreement
The same provider lists several forms. A single-beneficiary agreement has one vendor and one customer. A multi-beneficiary agreement lets a vendor add customers to a master agreement by appendix, so each new customer does not need its own contract. Others cover SaaS and bespoke work owned by the customer.
Verification
A deposit is only useful if it can be rebuilt. The provider above lists verification as a service separate from the deposit itself. In general terms, checks range from confirming that the files are present and intact to confirming that a fresh build matches the product shipped to customers. Ask what a quote includes, and test a build from your own deposit once.
A worked case
An illustrative one. A company pays you $30,000 a year for a tool it uses in daily operations. Its procurement team asks what happens if you close. You add the customer as a beneficiary to a multi-beneficiary agreement, deposit the code quarterly, and name release events such as insolvency or ceasing to support the product. The customer gets continuity on paper and you keep the code private until one of those events.
Software escrow versus license continuity
Escrow protects access to the code. It is not the same as keeping licensed copies running. If a licensing service went away, tokens already on customers’ machines stay valid until their refresh_by, which is 14 days after the token was issued, and an app can choose to keep trusting an expired window. Every license can also be exported as CSV. That keeps paying customers running, and gives nobody the source.
How Clavem handles it
Clavem does not provide source escrow. Its continuity story concerns licenses: each response carries a token signed with a key that belongs to your product, and your app verifies it with the public key you ship, so a valid license works offline and without Clavem. See offline verification and the signed license token. Escrow for your own code stays a matter between you, your customer and an escrow agent. For business buyers more generally, read selling software to businesses.
Questions
Who asks for software escrow?
Mostly business customers that depend on your software and cannot afford to be stranded if you stop, such as larger companies buying through procurement. Individual buyers rarely ask.
Does escrow mean the customer gets my source code?
Only if a release event defined in the agreement happens. Until then the code stays with the escrow agent, and the customer has no access.