Does a Make EU data center satisfy GDPR transfer rules for your workflows?
An EU data center in Make reduces questions about where scenario data is hosted, but it does not settle who can access that data. For an EU-regulated team, the honest answer is that a Make EU zone is a starting point, not a conclusion. Self-hosted n8n is the main option in this group that keeps the runtime, credentials and execution data inside your own infrastructure, so the transfer analysis rests on your own access paths rather than on a vendor's hosting choice.
GDPR Chapter V governs transfers of personal data to third countries, and the EDPB's Guidelines 05/2021 (version 2.0) state that remote access from a third country counts as a transfer, even when the data is only displayed on a screen for support or administration. Storage in a cloud outside the EEA also counts. That means a workflow stored in an EU zone can still be a transfer if a support engineer, a sub-processor or a parent company outside the EEA can view the data. The EDPB guidance also notes that complying with a third-country authority request would be a transfer, so the exposure depends on the vendor's legal footprint and its access procedures, not only on the region label. Ask Make in writing which entities can access customer scenario data, from where, and under which legal process.
Work through the answer scenario by scenario. List the workflows that move personal data, the systems they touch and who in the vendor chain can reach execution logs and mapped values. Note that a Make blueprint export excludes connections, so credentials must be recreated in whatever platform runs the scenario. For each flow, record whether the data leaves the EEA at any point, including through support access. Where the answer is unclear, a self-hosted n8n instance in your EU infrastructure lets you control those paths directly, though you still need your own access controls, logging and a documented transfer assessment. Confirm the license edition you need before you deploy, because some controls depend on it.
Which requirements a workflow platform must meet
A workflow platform touches every system it connects, so its hosting, access model and exit path decide how much personal data leaves your control. The cards below set out the questions to answer for Make's EU zone and for self-hosted n8n before you choose.
Where processing and storage happen
Make lets you pick an EU data center when you create the organization, and that choice cannot be changed later. Ask Make which components of a scenario run in that zone, including execution logs, data stores and any sub-processors that handle payloads. With self-hosted n8n, the main instance, workers and PostgreSQL database all sit in infrastructure you choose, so you can map every location in your own architecture diagram.
Who can access data in the clear
Personal data in a workflow is readable by whoever can open an execution, inspect a mapped value or export a scenario. Confirm which Make staff roles can reach customer data during support, and whether access is logged. With n8n, your own administrators control access, but support engineers and vendors only see data if you give them access. Check that exported blueprints and workflow JSON files do not carry personal data into unmanaged storage.
Assessing third-country authority requests
EDPB Guidelines 05/2021 (Example 12) say that using an EU processor subject to third-country law is not itself a transfer, but complying with a request from a third-country authority would be one. EDPB Recommendations 01/2020 add that where the processor needs data in the clear and the foreign authority's powers go beyond what is necessary and proportionate, no effective technical supplementary measure can be envisioned. Ask your provider how it would respond to such a request and document your assessment before you rely on any EU zone.
Audit logging and access control
Audit trails help you show who changed a workflow, who viewed credentials and when data was accessed. Confirm which audit log features Make includes in your plan, since they are not part of the standard tier. In n8n, the Community edition has no SSO or log streaming, while the Business and Enterprise plans add SSO and LDAP, environments with Git and log streaming. Plan how logs will be kept, retained and reviewed in your own SIEM.
Credential encryption and key custody
Credentials stored by the workflow platform are protected by an encryption key, and the key decides who can read them. In n8n queue mode, the main instance's encryption key must be shared with every worker and webhook processor so they can decrypt credentials. Treat that key as a high-value secret and store it in your own secrets manager. For Make, ask how credentials are encrypted and whether you can control key rotation.
Portability and exit
Your exit path determines how much a future change of provider will cost. Make blueprints export scenario modules, settings and mapped values, but connections are not included and must be recreated after import. Zapier exports are JSON files available only on Team and Enterprise plans. The EU Data Act lets customers switch to another provider or to on-premises infrastructure, with a maximum two-month notice period and a 30-day transition, so ask your provider how it will support that switch.
What the Make EU zone does and does not settle
Make lets you choose an EU data center zone when you create an organization, and that choice cannot be changed afterward. The zone tells you where the platform processes your scenarios, but it does not by itself show every place data passes through during a run, storage or support.
- Zone fixed at creation: The EU zone is set once. If the wrong zone is chosen, the practical fix is a new account and a rebuild of every scenario and connection.
- Scope of the zone: Confirm in writing whether scenario data, execution history, logs and backups all stay in the EU zone, and where any supporting services run.
- Self-hosted placement is yours: With n8n on your own EU infrastructure, you choose the region of the application servers, PostgreSQL, Redis and backups.
- Queue mode placement: In n8n queue mode, the main instance, workers and webhook processors all reach shared PostgreSQL and Redis, so every component should sit in the jurisdiction you have assessed.
Check whether any component in the Make stack, such as an email, monitoring or error-notification service, is located outside the zone you selected.
The GDPR question is less about where storage sits and more about who can reach personal data. The EDPB's Guidelines 05/2021 treat remote access from a third country as a transfer, even when staff only view data on a screen for support. An EU zone does not settle this if administration or support staff work outside the EEA.
- Remote access counts: Any support engineer or administrator outside the EEA who can view personal data is a transfer under Chapter V, so ask Make which staff and locations can open tenant data.
- Foreign legal orders: EDPB Guidelines Example 12 says an EU processor subject to third-country law is not a transfer by itself, but complying with a third-country authority request would be. Article 48 limits recognition of foreign disclosure orders to those based on an international agreement.
- Reach of US law: The CLOUD Act provision at 18 U.S.C. 2713 applies to covered providers regardless of where data is located. Confirm whether Make is subject to that provision and what it would do in a request.
- Transfer frameworks: The EU-US Data Privacy Framework adequacy decision covers only organizations on the DPF List, and the General Court judgment upholding it is under appeal as Case C-703/25 P, so contracts should not rest on it alone.
- Self-hosted scope: With n8n in your infrastructure, the vendor holds no tenant data, but your cloud or colocation provider and any support engagement still need the same review.
Ask Make for its sub-processor list, its support access paths and the legal basis it relies on for any third-country access, in writing.
Audit evidence and exit terms decide whether a GDPR position can be shown to an assessor and kept over time. Make's blueprint export preserves scenario structure but not the credentials that connect it to other systems, so moving away means rebuilding connections. Self-hosting shifts log custody to your own team.
- Blueprint scope: A Make blueprint holds modules, module settings and mapped values, but connections are excluded and must be recreated after import.
- Logs under your control: With self-hosted n8n, execution and infrastructure logs stay in your environment. n8n's pricing page places log streaming in the Business and Enterprise plans, so confirm which edition you need to forward logs to your SIEM.
- Retention terms: Confirm with Make how long execution data and platform records are kept, where they are stored and how deletion is evidenced.
- Switching rights: The EU Data Act, applicable since 12 September 2025, gives customers the right to switch to another provider or to on-premises infrastructure, with a maximum two-month notice period and a 30-day transition. Switching charges are banned from 12 January 2027.
Confirm the export format, retention period and deletion terms in the agreement before you rely on either route.
Self-hosted options that fit an EU data constraint
This table compares self-hosted n8n, Activepieces, Node-RED and Windmill against an EU data constraint, covering how each keeps processing in your infrastructure, its license limits for commercial use, and which edition holds audit logs and SSO. Read it as a checklist: the license and edition columns decide what you can run in production, while the data location column shows where your team controls the runtime.
| Option | How it meets the constraint | Trade-offs and license limits | Audit log and SSO by edition |
|---|---|---|---|
| Self-hosted n8n | Runs in your own EU cloud account or data center, so workflow data and credentials stay where you place them | Sustainable Use License allows internal business use but not reselling; files with .ee require an Enterprise License. Confirm your use case with n8n | SAML SSO and log streaming are listed for Business and Enterprise only; confirm audit log coverage for your edition |
| n8n queue mode | Main instance, workers and PostgreSQL stay inside your network, keeping execution data in your EU infrastructure | All worker and webhook nodes need the main instance's encryption key, so key handling must be designed with your security team | Multi-main high availability is restricted to Enterprise self-hosted deployments |
| Activepieces | Self-hosted deployment keeps flows and connections on infrastructure you choose in the EU | MIT core is permissive, but directories under packages/ee and the ee API path use a separate commercial license | Confirm which SSO, audit log and Git features sit in the commercial ee code before relying on them |
| Node-RED | Flow engine runs on your own servers, so event data is processed only where you deploy it | Apache-2.0 license permits commercial use, but it is a flow tool without a built-in multi-user governance layer | Confirm whether your SSO and audit requirements need add-on modules or custom work |
| Windmill | Self-hosted via Docker Compose or Kubernetes Helm charts, with external PostgreSQL in your EU environment | Builds compiled without the enterprise flag are AGPLv3 and Apache-2.0; prebuilt Community images include proprietary code free for internal use only | Managed services or reselling need a commercial license; confirm which SSO and audit features each edition includes |
| Make EU zone (reference) | Processing location is chosen at creation and cannot be changed, so the zone is fixed for the account | Third-country access by Make's parent or support staff must be assessed under Chapter V and Article 48 of the GDPR | Confirm audit log and SSO availability for your Make plan with the vendor |
How to deploy n8n in your EU infrastructure
This sequence moves a Make scenario estate into a self-hosted n8n runtime that your team operates inside EU infrastructure, with licensing, encryption, identity, logging and transfer checks settled before production traffic flows.
Confirm license edition and internal use fit
Legal or procurement confirms that the planned use falls under n8n's Sustainable Use License, which allows use for your own internal business purposes, and that no part of the deployment is a managed service or resold offering. The team decides which edition it needs: the free Community edition runs without a license key, while SSO, LDAP, Git environments and log streaming are only included in the Business and Enterprise plans. Do not deploy any file with .ee in its name or directory unless a valid n8n Enterprise License is in place. The output is a written decision record naming the edition and the license basis, signed off by the data protection officer and the procurement owner.
Choose EU-located compute and PostgreSQL
The platform team selects compute and a PostgreSQL database in an EU region that your organization already contracts with, and documents the provider, region and account owner for each. Confirm with the cloud provider where backups, logs and snapshots are stored, because processing location and storage location both matter for transfer analysis. Confirm also who holds administrative access to the hypervisor, the database and the container host. The output is an architecture record that lists every component, its location and its operator, which later feeds the record of processing activities.
Deploy n8n with Docker Compose or a cloud install
The operations team deploys n8n using one of the documented installation methods, such as Docker Compose or a documented install on a major cloud provider, pointed at the external PostgreSQL instance from the previous step. Store the configuration in version control, keep secrets out of the repository, and place the editor and webhook endpoints behind TLS. Confirm that the instance starts, connects to PostgreSQL and serves the editor only to authorized staff. The output is a reproducible deployment that a second engineer can rebuild from the repository and documentation.
Configure queue mode when scale requires it
If execution volume justifies it, the platform team switches to queue mode, where the main instance handles triggers and webhooks, Redis carries executions to a pool of workers, and the workers read and write workflow data in PostgreSQL. Load test representative scenarios to set worker counts against your own volumes, since no sizing figures are supplied here. Note that multi-main high availability is restricted to Enterprise self-hosted deployments. The output is a documented topology with measured throughput for your workloads, or a recorded decision that single-instance operation is sufficient.
Store and rotate the encryption key securely
The security team generates the n8n encryption key, stores it in a dedicated secrets manager with restricted access, and records who can read it. In queue mode the same key must be shared with every worker and webhook processor so they can decrypt credentials in the database, so the distribution path must be automated and audited. Write and rehearse a rotation procedure that covers re-encrypting stored credentials, because losing or changing the key without that step can make stored connections unusable. The output is a tested backup of the key held apart from the database and a rotation runbook with a named owner.
Enable SAML SSO on Business or Enterprise
The identity team connects n8n to your corporate identity provider using SAML single sign-on, which the Business and Enterprise plans include, with the setup guides for Okta Workforce Identity or Azure AD as a starting point if those are your providers. Test sign-in with a pilot group, confirm that deprovisioned staff lose access, and confirm which roles map to which permissions. If you stay on the Community edition, document the local account controls instead and accept the gap. The output is a tested sign-in flow and a role mapping that the access review process can audit.
Set up log streaming on supported editions
On the Business or Enterprise editions, the operations team configures log streaming so that execution and audit-relevant events leave the n8n instance and reach your central log platform inside the EU. Confirm exactly which events are streamed for your edition, since audit log scope is an Enterprise matter and should be checked against the vendor's current documentation. Set retention in the destination according to your own policy. The output is a documented event inventory, a working test event visible in the central platform, and a retention setting that the compliance team has approved.
Rebuild scenarios from blueprints and recreate connections
Engineers read each Make blueprint .json export to inventory modules, module settings and mapped values, then rebuild each scenario as an n8n workflow. Connections are not included in blueprints, so each credential must be recreated in n8n by the workflow owner, using a service account where possible, and stored in the encrypted credential store. Mark each rebuilt workflow with its source scenario and owner. The output is a rebuilt workflow inventory with a one-to-one mapping to the original scenarios and a list of recreated connections that the owners have signed off.
Test transfers to each external API
For every external API a rebuilt workflow calls, the data protection team and the workflow owner record what personal data is sent, where the receiving service processes it, and whether that processing occurs in the EEA, in a third country or under an adequacy basis. Confirm the current status of any transfer mechanism with counsel, because the EU-US Data Privacy Framework is subject to a pending appeal. Then run each workflow with test data and check the payload and destination logs. The output is a transfer register with one entry per external API and a dated test record for each, which keeps the Make-to-n8n decision defensible under Chapter V of the GDPR.
Evidence a DPO or auditor will ask for
A data protection officer or auditor will test whether the EU data center choice is true in practice, not only on a vendor page. The cards below list the records that show where personal data sits, who can reach it, and which license terms apply, whichever route you take. DeployOpen prepares this evidence, but auditors and certification bodies, not DeployOpen, reach any conclusion about it.
Records of processing and data location maps
Article 30 records of processing activities must name each workflow that touches personal data, the categories of data, the recipients and the storage locations. For Make, the question is which EU zone each scenario runs in and whether logs, execution history and error payloads stay there too. For self-hosted n8n, the map should show the host, database, Redis queue, backups and any worker nodes in your own EU network. Ask Make to confirm in writing where execution data and logs are stored, and do not assume the zone choice covers every copy.
Transfer assessments and sub-processor lists
Even with an EU zone, a provider's support staff, sub-processors or cloud operators may access data from outside the EEA. EDPB Guidelines 05/2021 treat remote access from a third country, even only displaying data on a screen, as a transfer. Request the current sub-processor list and each location, then document a transfer impact assessment for every third-country path. Where the processor is itself subject to foreign law, record the exposure to government access requests before you rely on the provider.
Access logs and admin change records
Auditors ask who opened a workflow, changed a credential, exported data or altered a schedule, and when. Make's audit trail depends on the plan, and n8n's log streaming and audit features are limited to the Business and Enterprise plans, so confirm which tier you need before you commit. In a self-hosted install, you own the logs: define retention, forward them to a central store you control, and test that a sample change can be traced end to end. Confirm the retention period in writing with the vendor or in your own policy.
Key management and credential encryption records
In n8n queue mode, the main instance's encryption key must be shared with every worker and webhook processor so they can read stored credentials. Auditors will ask where that key lives, who holds it, how it is rotated, and whether a copy exists in backups. Document the key custody procedure, the access list for the key store, and the steps for rotation. For a managed Make scenario, ask the provider how connection credentials are encrypted and who at the provider can access them.
License and edition evidence
Compliance files should show which n8n edition runs in production, because features such as SSO, multi-main high availability and log streaming depend on the license. n8n's core code is under the Sustainable Use License, which is not an OSI open-source license, and files with .ee in their name need a valid Enterprise License. Keep the license key record, the edition in use, and a legal view on whether your use is internal business use under the license. Confirm the terms with your legal team, since this page does not give legal advice.
Questions buyers ask before choosing a route
Is an EU data center enough for GDPR compliance?
An EU data center is one factor, but it does not settle GDPR on its own. Under Article 44, every transfer of personal data to a third country must meet Chapter V conditions, and the EDPB's Guidelines 05/2021 treat remote access from a third country as a transfer, even when staff only view data on screen for support. Storage in a cloud outside the EEA is also a transfer. The US CLOUD Act provision at 18 U.S.C. 2713 applies to providers' data in their possession, custody or control, wherever it sits. Ask the vendor who can access your data, from where, and through which sub-processors. Self-hosting n8n in your own EU infrastructure removes the vendor from the processing path, but you still need a lawful basis, records of processing and a review of any cloud provider you use. Confirm the conclusions with your data protection officer and counsel.
How much does self-hosted n8n cost compared with Make credits?
Start with n8n's published pricing. The Community edition is free to self-host, Starter and Pro cloud plans are 20 and 50 euros per month billed annually, the Business plan is 667 euros per month billed annually and is currently self-hosted only, and Enterprise is custom priced. n8n states that all plans include unlimited users, workflows and integrations, and it counts an execution as one run of a whole workflow regardless of steps. The cost components differ from Make's. With n8n you pay for the license tier, the servers or cloud capacity, database and backup storage, staff time for upgrades and monitoring, and the one-time rebuild. Make's cost recurs with credit consumption. SSO and LDAP, Git environments and log streaming sit in the Business and Enterprise tiers, so those needs drive the license line. Ask for a quote built from a written scope.
How long does rebuilding Make scenarios in n8n take?
No fixed duration applies to every team, so ask for a timeline only after the scenarios are inventoried. The main drivers are the number of scenarios, how many modules and routers each one uses, custom code, error handling and the number of app connections that must be recreated. Make's blueprint export includes modules, module settings and mapped values, but not connections, so credentials are recreated in n8n. Iterators, aggregators and routers must be mapped to n8n nodes and tested against the same data. Scenarios that call AI steps or write to regulated systems usually need more review. Testing volume and the cutover method (parallel runs or a timed switch) also set the schedule. A fixed-scope written plan, agreed before work starts, is the usual way to pin these down.
What does the client team need to provide for a rebuild?
The client team supplies access, decisions and knowledge that no outside party can supply for them. Plan for these inputs:
- A named business owner for each scenario and an approver for data flows that include personal data.
- Export access to the Make scenarios and blueprint files, plus a list of which app accounts each scenario uses.
- Credentials for every app connection, recreated in a controlled vault during the rebuild.
- Decisions on hosting location, network paths to internal systems, identity integration and backup targets.
- Test data and a sign-off process for each cutover.
- Staff time from the teams that own the source and destination systems.
If you run n8n in queue mode, the encryption key must also be shared with worker and webhook processors, so plan how that key is stored and rotated.
Can our internal team run self-hosted n8n without outside help?
Many teams can run n8n alone, but only if the skills are already in place. The operations work includes container or Kubernetes deployment, PostgreSQL administration, Redis for queue mode, TLS, backups, upgrades, monitoring and incident response. Queue mode adds Redis and worker pools to operate. Multi-main high availability is restricted to Enterprise self-hosted deployments. On licensing, the Sustainable Use License permits use for your own internal business purposes, but it does not permit offering n8n as a hosted service to others, and files with .ee in their names require an Enterprise license. If the internal team lacks capacity or a skill, DeployOpen can plan, implement, customize, harden, back up, upgrade, document and support self-hosted platforms inside your infrastructure, under a fixed scope or a retainer. Each engagement starts with written scope.
How do the Data Act switching rights affect exit planning?
The EU Data Act has applied since 12 September 2025. Under Articles 25, 29 and 50 of Regulation (EU) 2023/2854, customers can switch from a data processing service to another provider or to on-premises ICT infrastructure. The notice period is at most two months, the transition period is 30 days, and switching charges are banned from 12 January 2027. For exit planning, check your Make contract for notice terms, export formats and data return steps, and do not assume the rights cover every dataset. Keep current blueprint exports and an inventory of connections so the exit does not depend on a single export. Ask counsel how the Act applies to your specific service and contract, because its scope depends on the service's classification.
How does queue mode change credential and key management?
In queue mode, the main instance hands executions through Redis to a pool of workers that read and write workflow data in PostgreSQL. Because workers and webhook processors must decrypt stored credentials, the main instance's encryption key must be shared with all of them. That makes the key a high-value secret. Store it in a secrets manager, restrict which hosts can read it, define a rotation procedure and back it up with the database, since credentials cannot be used without it. Limit database access to the n8n services, log administrative changes and test restore procedures before go-live. Confirm the exact key handling for your n8n version in its hosting documentation before you finalize the architecture.
What each route costs over time
Make's cost is almost entirely recurring, because the subscription and credit allowance are billed on a cycle for as long as scenarios run. Credit consumption is the component that moves: scenarios with many steps, polling triggers or AI steps draw down credits faster, so the bill grows with activity rather than with the number of workflows. Ask Make for the current plan tier that includes the EU zone you need, the credit allowance it carries, what happens when credits run out, and whether the EU data center choice carries any price difference. Confirm these against Make's current price list in writing, because this page does not state Make's figures.
n8n's software cost depends on the license edition, and the Community edition is free to run. The n8n pricing page lists the free self-hosted Community edition, a Business plan at 667 euros per month billed annually that is currently self-hosted only, and Enterprise at custom pricing. The Business plan is the only self-hosted tier that includes SSO, LDAP, Git environments and log streaming, and audit logs require Enterprise, so the edition you choose sets the recurring license line. Execution counting is per whole workflow run, so step-heavy scenarios do not raise the license price the way credits do. Remaining costs are one-time and recurring infrastructure, a PostgreSQL database, monitoring, backups, staff time for upgrades and support, and any retainer you buy for testing and updates. These costs rise with queue mode, which needs Redis and worker capacity, and with high-availability requirements, which need Enterprise multi-main setup.
Ask for a written, line-by-line quote from each route before comparing totals. For n8n, request the edition and license terms, the hosting estimate for your own cloud or data center, and a scope covering migration, backups, monitoring and upgrades, with who pays for each item and what recurs. For Make, request the plan, credit allowance and renewal terms. Then add internal staff time, which is often the largest line in a self-hosted route and is easy to omit. Your quote should state the fixed scope, deliverables and timeline, and who holds responsibility for uptime after handover.
How DeployOpen deploys n8n under an EU data constraint
DeployOpen plans, implements, hardens, documents and supports self-hosted n8n inside your own EU cloud accounts, data centers or private networks. The work is scoped around where personal data is stored, where processing runs and who holds the encryption keys, so the GDPR questions are answered in the architecture before the first workflow is rebuilt.
How we scope
- Written scope maps every Make scenario to its data categories, source systems and destination systems before any build starts.
- Scope records which EU region, data center or private network hosts the main instance, workers and PostgreSQL database.
- Scope lists who administers the platform, who can read execution logs and which third parties could receive any data.
- Engagements are fixed-scope, with deliverables and timelines agreed in writing before work begins.
How we build
- Deploy n8n in your own infrastructure using the documented Docker Compose or queue mode setup, with configuration kept in version control.
- Plan the encryption key handling for queue mode so workers can read stored credentials without exposing the key outside approved hosts.
- Rebuild scenarios from blueprint exports, then recreate connections inside the new instance under your own accounts.
- Test each workflow on sample data and document which execution paths keep data inside the agreed EU boundary.
How we hand over
- Deliver architecture diagrams, data flow records and a list of every system that touches personal data.
- Provide runbooks for backup, restore, upgrades and credential rotation, written for your operations team.
- Document the license edition in use and which features, such as SSO, audit logging and log streaming, depend on it.
- Prepare evidence for your assessors, while independent auditors and certification bodies issue any audit reports or certifications.
How we support
- Ongoing support, upgrades and continuous testing are available as retainers after handover.
- Retainers can cover n8n version upgrades, patching of the underlying host and periodic review of data flows as scenarios change.
- Ask in the quote which items your team keeps and which ones the retainer covers, including after-hours responsibilities.
- Confirm with DeployOpen during scoping how the retainer handles changes to the license terms or edition you choose.
