A low server bill can look attractive until someone has to restore a backup or fix a failed update. Include that person and their time in a self-hosting comparison.
For a website integration, the choice is partly about cost and partly about who operates the system. Write down those responsibilities before deciding where it should run.
Put ownership in writing
A self-hosted workflow needs an owner for updates, backups, credentials, logs and recovery. Name that person and the expected response when a workflow stops. A cheap server is not a support agreement.
A managed platform takes on some infrastructure responsibilities, but your team still owns workflow configuration, connected-account access and the consequences of incorrect data. Read the service's current support and retention terms.
Compare a representative month
Estimate successful runs, retries, scheduled checks and external API calls. Add operator time and the cost of recovery. There is no universal execution count or team size where self-hosting becomes cheaper.
For n8n specifically, check its current pricing and Sustainable Use License. Self-hosting availability does not imply unrestricted redistribution or free use of every feature.
Test recovery before launch
Restore a backup in a safe environment. Expire a test credential and verify that the responsible person sees the failure. Confirm that replaying an event does not send duplicate messages or create duplicate records.
Document how to pause a broken workflow and continue the underlying work manually. The office should not need to understand the hosting platform to find a missing inquiry.
Choose based on the actual constraint
Managed hosting may suit a team without an infrastructure operator. Self-hosting may suit a team with a specific deployment requirement and the ability to support it. Neither is inherently a compliance guarantee.
Use the website technical-scoping stage to resolve these responsibilities before committing to an integration.
