Back openDesk Edu for a sovereign, open-source education β every vote counts.
Vote nowSave products you love by clicking the heart icon.
My long-term vision for a sovereign, open, and human-centered digital workplace β and a phased roadmap showing how we move from vendor lock-in today to a federated, AI-companioned workplace by 2055.
Many organizations face a decision: adopt Microsoft 365 or not. At first glance, MS365 seems like the easy choice β complete package, integration, low initial costs. But without proper planning, the decision for MS365 is a decision for long-term dependency.
The problem: Exit strategies are often developed after implementation. By that time, the organization is already deeply embedded Microsoft ecosystems, data architectures, and workflows. Leaving becomes a technical, organizational, and financial challenge.
This article argues for a different principle: Exit strategies must be developed in parallel with MS365 implementation. MS365 should not be viewed as an end state, but as a reversible operational state with exit-by-design. Only this maintains digital sovereignty β control over your data, identities, and systems.
Microsoft 365 offers a complete ecosystem: Teams for communication, SharePoint for collaboration, OneDrive for storage, Exchange for email, Power Platform for workflows and automation. All services are integrated, identities run via Entra ID (formerly Azure AD), and applications connect seamlessly. However, this convenience is also dangerous.
What's often overlooked: Each year without exit planning increases an organization's technical debt. This accumulates as:
| Debt Type | Monetary Impact | Operational Impact |
|---|---|---|
| Data Migration | Exponential growth with data and time | Downtime, data loss |
| Training Potentials | Employees trained on MS360+alternatives | Productivity loss |
| Process Redesign | Redesigning all workflows | User resistance |
| Technical Refactoring | Re-architecting interfaces, APIs | Project risk |
| Contract Bindings | Renewals, previously planned termination dates | Negotiation power loss |
Practical Example: A mid-sized organization with 500 employees introduced Microsoft Teams and SharePoint without an exit concept. After four years:
A later exit would cost an estimated β¬1.5ββ¬2.5 million for consolidation, migration, and training alone β an amount that could have been reduced to 30β40% with parallel exit planning.
An exit strategy can and should be developed in parallel with MS365 implementation. The key idea: don't implement MS365 as an end state, but as a reversible, controlled operational state.
This means: Technical, legal, and organizational measures are taken during implementation so organizations can later extract services, data, and processes from MS365 β without chaos, data loss, or complete redesign.
Even during implementation, these questions should be clarified:
The exit strategy then becomes not a later special project, but part of MS365 governance.
Organizations can follow different target images. Complete exit is not the only option.
MS365 remains for certain functions, such as Excel, Word, or an Exchange migration. File storage, project work, and collaboration are increasingly replaced by open-source services. The goal is not anti-Microsoft, but pragmatic usage and parallel development.
MS365 is used only for less sensitive or highly standardized scenarios. Sensitive projects, administrative processes, critical collaboration, or personal data reside preferentially on own or federated infrastructure. The "where" decision is driven by data classification, not convenience.
MS365 is gradually replaced over several years via openDesk, Nextcloud, Collabora, Matrix, OpenProject, and other services. This is a multi-year project, but by preparing during implementation from the start, the path remains open.
Important: Organizations don't need to decide whether to exit entirely from day one. But they should remain exit-capable from the beginning.
There should be a joint program for implementation and exit-capability, not two separate projects. Participants should include:
This council decides not just MS365 configurations, but also:
An exit strategy only works if it's clear which data can go where. The following classification is a typical model for organizations:
| Data Class | Examples | MS365 Usage? | Target Strategy |
|---|---|---|---|
| Public | Website content, public materials | Possible | Uncritical |
| Internal | Working documents, protocols | Restricted possible | Keep export-capable |
| Personal | Customer data, employee data | Only after review | Prefer sovereign services |
| Particularly Sensitive | Health data, sensitive project data | Possibly not | Local/federated services |
| Financial Data | Invoices, balance sheets, accounting | Critical | DMS/Documents Management System |
| Strategic Data | Business plans, R&D results | Critical | Local/federated services |
This classification should be translated into guidelines, training, and IT policies already during MS365 implementation.
Avoid certain lock-in traps already during baseline configuration.
A particularly important point: Entra ID must not become the central identity system if exit capability is desired.
Better is:
Organizational IAM / LDAP / AD / Shibboleth / Keycloak
|
Keycloak or Federation Service with MFA
|
MS365, openDesk, other services
Goal:
This is one of the the most important technical prerequisites for any later exit.
While MS365 is implemented, a sovereign target platform should be evaluated in parallel. A compelling model:
| MS365 Function | Parallel Evaluation Alternative |
|---|---|
| Teams / SharePoint Workspaces | openDesk-CE |
| OneDrive | Nextcloud / ownCloud |
| Office Online | Collabora Online / OnlyOffice |
| Planner | OpenProject, Taiga |
| Forms | LimeSurvey |
| Teams Chat | Matrix/Element, Mattermost |
| Video Calls | BigBlueButton, Jitsi Meet |
| OneNote / Knowledge Base | XWiki, HedgeDoc |
| Exchange | SOGo, Open-Xchange, Postfix/Dovecot |
openDesk-CE is particularly well-suited as a parallel evaluation platform because it integrates several of these functions out-of-the-box: file storage, online office, project management, knowledge work, communication, and identity integration.
Parallel to MS365 implementation, organizations should enshrine rules that limit future dependencies rule.
Examples:
These rules prevent organizations from binding themselves more deeply than intended during implementation.
An exit strategy in parallel to implementation doesn't mean building two entirely separate worlds. Better is a controlled coexistence:
MS365: Short-term standard platform available
openDesk/OS Services: Strategic sovereign target platform
Organizational IAM: Common identity basis
Data Classification: Decides storage location
Governance: Controls usage and migration
Sample configuration:
Already during implementation, an "exit brief" should exist for each service.
Parallel to implementation, contracts and internal procurement guidelines should be adjusted.
Key requirements:
Open-source alternatives should also consider operations, support, and exit-capability contractually.
Possible roadmap:
| Timeline | MS365 Introduction | Exit / Sovereignty Measures |
|---|---|---|
| 0-3 months | Project kick-off, Tenant planning | Governance, Data classification, Exit guidelines |
| 3-6 months | Pilot MS365 | openDesk-CE/Nextcloud pilot, IAM decoupling |
| 6-12 months | Rollout of selected MS365 services | No-new-lock-in rules, export tests, training |
| Year 2 | Broader usage | Standard for new sensitive projects on OSS platform |
| Year 3 | Consolidation | Reduction of Teams/SharePoint new sites, migrate first areas |
| Year 4+ | License optimization | Possible partial or complete de-provisioning of individual MS365 services |
Thus no contradiction between Rollout and Exit emerges: MS365 is used in a controlled way while alternatives grow.
A sensible formulation might be:
The organization only integrates MS365 under the condition that identities, data, processes, and document formats are designed such that a later migration to open or sovereign platforms remains possible. In parallel, openDesk-CE and other OpenSource services are built up as strategic target environments and preferred for suitable usage scenarios.
This vision creates clarity: MS365 is not prohibited, but usage occurs under conditions that ensure exit capability can be maintained.
Brief summary:
The most important point: An organization should not implement MS365 "blind" and then think about the exit later. It should use MS365 from the start in a way that makes later exit technically, organizationally, and legally possible.
This article reflects experience from digital sovereignty implementation across various organizational sizes from 2023-2026.