Deploying an ERP without an integrator: it’s a promise that has been circulating in founder and tech communities for years. Odoo Community is free. Dolibarr runs on €5/month hosting. YouTube is full of “ERP in 2 hours” tutorials. Why pay €20,000 to an external partner?
The reality is more nuanced. Self-implementation is sometimes the most rational decision an SME can make. And sometimes it’s the mistake that ends up costing twice as much as the integrator it was trying to avoid. This guide lays the groundwork for an informed decision: profiles, real costs, and tipping-point thresholds.
Why Self-Implementation Attracts (and What Gets Underestimated)
The Economic Argument: Visible Savings
The appeal is real. An Odoo project delivered by a certified partner costs, at the low end of the market, between €10,000 and €20,000 for an SME of 20 to 50 employees with a standard scope (accounting, sales, procurement, inventory). For an extended scope (manufacturing, e-commerce, multi-site), the range reaches €30,000 to €40,000. Dolibarr is cheaper to implement, but a serious configuration by a consultant still falls in the €5,000 to €15,000 range. Avoiding these amounts by doing it yourself is an immediate and visible saving — which is also what creates the bias.
Hidden Hours: Configuration, Testing, Training, Maintenance
What teams forget to account for is the cost of internal time. A successful Odoo self-implementation requires, based on consolidated practitioner feedback, 3 to 6 months at 0.5 FTE from a senior technical profile. A Python developer or an experienced systems administrator — not an intern. At the loaded cost of a senior developer in Western Europe (€90,000 to €120,000 per year), six months at half-time represents €22,500 to €30,000 in real cost — comparable to what an integrator would have charged, but without the guarantee of outcome or the accumulation of domain expertise.
On top of that come the systematically underestimated costs: end-user training, internal procedure documentation, fixing configuration bugs in production, managing backups and security updates. None of these tasks take two hours.
The Optimism Bias: “We Have In-House Developers”
This is the most common trap. Having competent web developers does not guarantee success on an ERP project. These are two different crafts: the developer who is fluent in React does not necessarily master Python, PostgreSQL, Odoo’s modular architecture, the business logic of a chart of accounts, or the intricacies of a multi-warehouse inventory setup.
The other trap is availability. An in-house developer working on an ERP project alongside production responsibilities will always deliver more slowly than estimated, with less focus, and will be the first person pulled in when production incidents occur. The ERP project takes a back seat, and the configuration sits half-finished for weeks.
Odoo Community Self-Deployment: Profile of a Successful Implementation
Real Technical Prerequisites
Odoo Community (LGPL-3 license) is technically demanding to deploy and maintain. The real prerequisites for a self-implementation without an integrator:
- Linux server: Ubuntu 22.04 LTS or Debian 12, with proficient system administration (Nginx, SSL, Fail2ban, service management)
- Python 3.10+: familiarity with virtualenv, pip, and system dependencies
- PostgreSQL 14+: basic administration, backups, slow query optimization
- Docker (optional but strongly recommended for isolation): proficiency with volumes, networking, and image updates
- Git: to manage custom modules and updates without regressions
Without these skills in-house or from a trusted freelancer, an Odoo self-implementation is not a saving — it’s an operational risk.
Out-of-the-Box Modules vs. Modules Requiring Development
Odoo Community covers without development: basic accounting, CRM, sales, procurement, inventory, project management, scheduling, and a supplier portal. That’s already substantial for many service or trading SMEs.
What requires custom development or goes through OCA (Odoo Community Association) community modules:
- Local payroll: the official Odoo payroll module is reserved for the Enterprise edition. Community builds rely on OCA modules (e.g.,
l10n_fr_hr_payrollfor French payroll), but with no official SLA or update guarantee. - Country-specific fiscal exports: OCA accounting modules cover local chart of accounts and statutory exports, but their integration requires careful configuration by someone who knows local accounting rules.
- Advanced e-commerce: the Community eCommerce module is functional but basic. Marketing automation, upsell, and abandoned cart recovery are reserved for Enterprise.
The 3 Paths and Their Real Costs
Self-hosted Odoo Community: reduced infrastructure cost (€15 to €80/month depending on the server), but 100% of the maintenance burden falls on your team. Major version upgrades (Odoo 17 to Odoo 18) represent 1 to 3 weeks of development work, with a risk of regression on customized modules.
Odoo.sh: Odoo’s managed cloud platform, available with the Custom plan. Hosting, backups, and infrastructure are handled by Odoo. Development and configuration remain your responsibility. Check current pricing directly on odoo.com/pricing as rates vary by user count.
Odoo Enterprise (Standard or Custom): per-user subscription. The Standard plan is listed at €19.90 per user per month (source: odoo.com/pricing), all apps included, hosted on Odoo Online. The Custom plan (€29.90 per user per month) adds Odoo Studio, multi-company management, and external APIs, with access to Odoo.sh or on-premise. Both plans include official support and updates. For an SME with 20 users, the Standard plan represents approximately €4,800 per year — less than the cost of a failed self-implementation.
Dolibarr Self-Deployment: Who Is It Really For?
The Ideal Use Case
Dolibarr (GPL-3 license) is the most accessible open-source ERP for self-deployment among mature solutions. A standard installation on a PHP/MySQL server or via Docker takes under an hour for a junior web developer or an intermediate-level systems administrator.
The profile that gets the most value from a Dolibarr self-implementation:
- Micro-businesses under 10 employees with simple needs: invoicing, CRM, basic inventory management
- Freelancers, consultants, and independent professionals who invoice clients and track projects
- Associations managing members, dues, and treasury
- SMEs that want a lightweight management tool alongside a specialized vertical industry application
The Dolibarr community is active and multilingual, with strong localization for multiple countries (France, Belgium, Spain, etc.). For European micro-businesses, this is a significant advantage over ERPNext.
What Breaks at 20, 50, and 100 Users
Dolibarr’s limits emerge as the load grows and processes become more complex.
At 20 users: performance degrades if the server is undersized. Permission management by user profile becomes tedious to administer. Community modules in the DoliStore vary widely in quality — some haven’t been maintained in 2 or 3 years.
At 50 users: the absence of MRP logic, production management, and capacity planning becomes a blocker for industrial SMEs. Multi-dimensional cost accounting doesn’t exist natively. Customization requires PHP development, and major Dolibarr upgrades regularly break customized third-party modules.
At 100 users: Dolibarr is no longer fit for purpose. At this scale, migration to Odoo, ERPNext, or a mid-market ERP is inevitable. The cost of this unanticipated migration (data to clean, retraining, process re-modeling) can exceed the cost of a properly-scoped ERP project done right from the start.
The 3-Tier Rule
Tier 1: Fewer Than 10 Users, 1 Entity, 1 Business Process
Self-deployment is viable, under conditions. Dolibarr or Odoo Community on a basic VPS will do the job. Risk is proportional to complexity: small scope, low risk. If the person who deployed the system leaves the company, an external partner can take over within a few days.
What you still need to plan for: an annual maintenance contract with a trusted freelancer (€500 to €1,500 per year), automated daily off-site backups, and documentation of critical configurations in a shared space — not locked in one person’s head.
Tier 2: 10 to 50 Users
Hybrid approach recommended. Initial self-implementation supplemented by an on-call support contract with an integrator. The “integrator on retainer” model offers the best cost-to-risk ratio at this stage.
At this point, processes start to intersect. A configuration error in the inventory module can propagate to accounting. The cost of a production incident (corrupted data, ERP downtime on a month-end closing day) quickly exceeds the cost of well-calibrated preventive support.
Tier 3: More Than 50 Users or Multi-Entity
An integrator is non-negotiable, not optional. From 50 users onward, or whenever multiple legal entities, currencies, or countries are involved, self-implementation becomes a risk that is hard to justify. The complexity of intercompany transactions, consolidation accounting, and entity-level access rights requires expertise that few internal teams possess without having already run several ERP projects.
A failed project at this scope doesn’t just cost technical remediation — it costs months of lost productivity, data that must be manually re-verified, and potentially regulatory penalties if tax filings were incorrectly generated.
What a Year of In-House ERP Management Really Costs
HR Cost: 0.3 to 1 FTE Depending on Complexity
The administrative burden of a self-hosted ERP doesn’t disappear after go-live. It typically breaks down as follows:
- 0.3 FTE for a stable Dolibarr installation with fewer than 15 users and limited customization: handling user tickets, minor updates, backups.
- 0.5 FTE for Odoo Community with 20 to 40 users and third-party OCA modules: system administration, module updates, internal functional support.
- 1 full FTE for heavily customized Odoo Community with custom modules, more than 40 users, or multi-site: bug fixes, major version upgrades, incident management.
Infrastructure, Backups, Major Upgrades
Infrastructure cost is often the only one teams calculate upfront. An intermediate VPS (2 vCPU, 8 GB RAM) at Hetzner, OVH, or Scaleway costs between €15 and €50/month, or €180 to €600/year. A dedicated server for large databases: €80 to €200/month.
But major upgrades are the invisible cost line. Migrating from Odoo 16 to Odoo 17 typically represents 1 to 3 weeks of developer work to test custom modules, fix regressions, and retrain users on interface changes. This cycle repeats every 2 to 3 years.
The Key-Person Risk
This is the risk no one quantifies until it materializes. If the person who deployed and maintains the ERP leaves, they take with them the knowledge of customizations, specific configurations, and maintenance scripts. Onboarding an external partner under those conditions costs systematically more than a fresh project — because they first need to understand the existing setup before they can improve it.
The mitigation: document non-standard configurations, maintain an up-to-date technical reference, and put a support contract in place with an integrator even if you only activate it once or twice a year.
The Right Compromise: Initial Self-Implementation + On-Demand Support Contract
Choosing a “Standby” Partner
The most effective model for tech-savvy SMEs at Tier 2: deploy yourself, then contract ad-hoc support with a certified Odoo partner or an experienced Dolibarr consultant.
This standby partner steps in for specific cases: major version migration, development of a critical module, resolution of a production incident. They are not on a continuous retainer. They respond within 48 to 72 hours and intervene 2 to 6 times per year. The cost of such a contract ranges between €3,000 and €8,000 per year for a package of 20 to 40 hours. That’s the cheapest insurance against critical incidents.
Key Clauses to Negotiate
In an ad-hoc support contract, three clauses are essential:
- Guaranteed response time: 24 hours for blocking incidents (ERP inaccessible, corrupted data), 5 business days for non-urgent changes and requests.
- Pre-configured environment access: the partner must be able to access your server and database without you needing to create access credentials on a Friday evening. A secure, documented, and tested SSH or VPN access must exist before any incident occurs.
- Documentation clause: require a technical dossier updated at each intervention. You must be able to switch partners without losing knowledge of your own system.
Final Recommendation: 3 Profiles, 3 Clear Answers
Profile A: Tech startup or SME (fewer than 20 employees, in-house dev team) Self-implementation is viable. Budget 3 to 4 months at half-time from a developer, a standby support contract as a safety net, and plan the migration to Odoo Enterprise when you exceed 25 to 30 active users. The question isn’t whether you’ll need to move to Enterprise — it’s when.
Profile B: Non-tech SME (20 to 80 employees, no in-house IT profile) Do not attempt self-implementation. The cost of an integrator is lower than the cost of a failed project followed by remediation. Request proposals from certified partners and use a scoring grid to compare them objectively.
Profile C: Very small business with simple needs (fewer than 10 employees) Dolibarr self-deployment is the logical answer. Installation in one day, two-week learning curve, zero license cost. Budget €1,000 to €2,000 for a day of configuration with a consultant if you have specific needs (historical data import, cost accounting, sector-specific modules).
To go deeper before making your decision, read our full open-source ERP comparison (Odoo, Dolibarr, ERPNext, Axelor), our guide to choosing an ERP integrator with a scoring grid, and our article on customizing an ERP without a developer via low-code.
