The matter began on July 2. On that day, SAP and IBM jointly announced that four global enterprises (JYSK, GBM, DIFARE Group, Plastilene Group) had chosen to migrate to the cloud with SAP Cloud ERP Private on IBM Power Virtual Server. On the same day, German media Handelsblatt reported that SAP had tightened new hiring and travel rules — CEO Christian Klein's goal was to extract funds from traditional businesses to invest in AI. On the other side, Odoo's latest data from July showed that its paying users exceeded 2.11 million, with 67% supported by its partner network. Odoo 20's native MCP Server has reduced the barrier to AI adoption for small and medium-sized enterprises to zero.
Large ERP vendors are tightening their belts, while open-source ERP is opening up interfaces. This is no coincidence—AI is rewriting the cost structure of the ERP industry.
Where did SAP's money go
Let's first look at SAP's moves. A report from Handelsblatt shows that SAP has tightened its rules for new hiring and travel approvals — although there is no full hiring freeze, recruitment quotas have been strictly limited to "AI core positions." CEO Christian Klein has expressed the same message on multiple occasions: AI is SAP's core strategy for the next decade, with legacy businesses needing to slim down and new businesses requiring investment.
Bloomberg previously reported that SAP is evaluating its restructuring project launched 18 months ago, which involves reconfiguring thousands of positions, cutting jobs in slow-growth areas, and expanding in high-potential sectors. SAP is also studying replacing 15-20% of its traditional application workloads with AI agents.
The logic is actually quite straightforward: for SAP, with an annual revenue of 34 billion euros, every extra euro spent on non-AI legacy business means one less euro invested in AI. And on the AI path, no one knows when large-scale monetization will be achieved. SAP's own "2026 AI Value Report" states frankly: only 4% of enterprises are ready for intelligent agents.
So SAP is tightening its belt while still spending money — it just completed the acquisition of Dremio on July 6, and launched People Intelligence on July 7. The money isn't less; it's being taken out of old pockets and stuffed into new ones.
On the same day (July 2), SAP and IBM jointly announced that four global enterprises have chosen SAP Cloud ERP Private on IBM Power Virtual Server: Danish home furnishings retailer JYSK (3,600 stores worldwide), Ecuadorian pharmaceutical giant DIFARE Group, Colombian soft film manufacturer Plastilene Group, and Central American IT service provider GBM. These four companies share one commonality—they are all migrating their on-premises SAP systems to the cloud.
IBM Institute for Business Value data shows that companies embedding AI into ERP systems have a 27% higher ROI. This figure is not exaggerated, but in the current economic environment, 27% is enough to get the CFO to sign off.
What logic is Odoo running on the other end
Unlike SAP's "slim down and invest in AI" approach, Odoo takes a different path.
According to the latest data from Odoo official and the partner community: Odoo paid users have reached 2.114 million, of which 1.422 million (67%) are supported by the partner network. The growth curve of this number accelerated significantly last year — Odoo's revenue in the European DACH region increased by 133%, and the number of partners increased by 70%.
More importantly, Odoo 20's native MCP Server. In early July, the Odoo R&D team announced that Enterprise users can directly access Odoo data through the /mcp endpoint via AI assistants, without any middleware. User reactions are quite interesting—many people commented, "I've been waiting for this feature for a long time." To be honest, this kind of response is uncommon in the ERP community.
Comparison of AI paths between Odoo and SAP (July 2026):
| Dimension | SAP path | Odoo path |
|---|---|---|
| AI data access | Acquire Dremio to build a unified data foundation | Native MCP Server, zero-middleware integration |
| Data Governance | Business Data Cloud + People Intelligence | Community Module + MCP Security Layer |
| Customer Group | Large/very large enterprises, central and state-owned enterprises | Primarily small and medium-sized enterprises, penetrating upward |
| Cost Structure | Freeze non-AI positions, cut travel expenses | Open source community collaboration, partner network |
| User scale | 40,000+ enterprise clients (including subsidiaries) | 2.11 million paying users (personal accounts) |
What is the difference between Odoo's logic and SAP's logic? SAP builds a complete data governance system within the platform before implementing AI; Odoo opens up data capabilities at the interface level, allowing AI to come in on its own. One is like building a highway, the other is like installing a universal socket—it's not that one is better than the other, but they target completely different user groups.
I recently came across a mid-sized manufacturing company that is migrating from SAP to Odoo. Their CIO made a very practical remark: "SAP's AI features are great, but so great that we can't afford to use them. Odoo's AI capabilities are 'good enough,' but the key is that the cost of integrating them is nearly zero."
Three Structural Signals in the ERP Industry
Putting together the three events of SAP freezing hiring, Odoo surpassing 2 million users, and IBM moving four clients to the cloud reveals three interesting changes in the ERP industry.
First, AI is becoming a cost watershed for ERP vendors. SAP is freezing non-AI positions to free up funds for AI investment—this shows that the scale of AI investment can no longer be covered by "supplementary budgets" and requires tapping into existing resources. Odoo can run at low cost on the other end because the open-source community shares a large portion of development costs, and AI integration follows a standardized route like the MCP protocol. In the future, competition in the ERP industry may not be about big fish eating small fish, but rather "who has a lower unit cost in AI."
Second, cloud adoption and AI are becoming increasingly intertwined. JYSK, DIFARE, Plastilene, and GBM—these four companies chose SAP Cloud ERP Private on IBM, not just for cloud migration, but to run AI workloads. IBM's Power Virtual Server provides enterprise-grade elastic scaling, security, and compliance. Without the cloud, SAP's Dremio data lakehouse and People Intelligence simply cannot run. This, in turn, is driving SAP customers' cloud adoption decisions: AI is forcing you to move to the cloud.
Third, the criteria for SMEs to choose ERP are changing. Among Odoo's 2.11 million users, SMEs are the absolute mainstay. Their AI selection logic is not "the most powerful features," but "the lowest integration cost, out-of-the-box usability." The Odoo MCP Server, which allows AI assistants to directly read Odoo data, is for many SMEs the first time they can truly use AI capabilities within ERP—without needing data scientists or custom integrations.
Practical advice for CIOs
I have organized three questions worth spending time thinking about this quarter.
Recalculate the "total cost of ownership" of your current ERP's AI. Don't just look at software licensing fees; also consider how much custom development you've done for data governance, integration, and security compliance to run AI. SAP's BDC + Dremio model may reduce AI data costs for large enterprises, but customization needs still exist. Odoo's MCP model may lower the AI entry cost for small and medium-sized enterprises to zero, but can it handle the complex scenarios of large enterprises? There is no standard answer, but you need to calculate it clearly.
AI-driven cloud migration decisions can't wait any longer. The logic behind IBM and SAP's joint promotion of Cloud ERP Private is clear: the AI data foundation relies on cloud infrastructure, and on-premises deployment won't cut it. If your ERP is still on-premises, the man-hours and costs of AI projects will be 30-50% higher than on the cloud. It's not a must, but the gap is widening.
Focus on ERP vendors' AI cost structure, not their AI feature list. SAP freezes hiring to invest in AI, Odoo uses open-source community to dilute AI R&D costs—two cost structures determine two product pricing models. SAP's AI will be in the high-priced enterprise edition, while Odoo's AI is a standard feature with no additional charge. When selecting a system, asking "How do you price your AI?" is more useful than asking "How many agents do you have?"
A sentence to sum up my recent feelings: The AI competition in the ERP industry is not about who has the better model, but about who can integrate AI into daily ERP operations without raising prices. SAP is tightening its belt to do this, while Odoo is doing it through open source—each has its own approach. For users, this is a good era—more choices and lower prices.
