A business does not need more excitement about software; it needs to know whether a new release reduces friction or creates new dependency. With Odoo 20, the useful question is not “should we upgrade?” in the abstract. It is “which processes can we move now without breaking the way we operate?”. Odoo announced version 20 in September 2026 during Odoo Experience, and its release notes already document the changes. Odoo also distinguishes between features that are released and others that depend on context, localization or module choice. (odoo.com)
The right question is not whether Odoo 20 has more features, but what operational decision it lets you make today.
What actually changed: a version aimed at less friction
The most visible shift in Odoo 20 is not cosmetic. Odoo’s own launch material frames the release around instruction-based automation, a new way to build websites, and improvements to everyday business work. The underlying idea is simple: reduce the number of manual steps between a business signal and the system action that should follow. (odoo.com)
That matters for concrete decisions. A sales order can trigger a follow-up action, a low-stock situation can raise an alert, and a repetitive task can stop depending on someone remembering it from a spreadsheet. This is not magic; it is fewer handoffs between systems and more explicit rules. That is valuable if your pain is repetition, not if your pain is poor data quality. (odoo.com)
There is also a practical point for distributed teams or sites with unstable connectivity. Odoo 20’s offline mode documentation says users can continue working in certain views and records when the internet drops, and changes sync automatically once the connection returns. But Odoo is explicit that this is for short interruptions, not for full or long-term offline operation. (odoo.com)
What is released, what is previewed, and what you should not assume
This is where careful readers should slow down. Odoo 20 was announced and released in September 2026. That does not mean every new capability applies equally to every company or is active in every configuration. The release notes organize changes by area, and the Spanish release notes show that some updates are tied to localization or version-specific conditions. (odoo.com)
That matters because many software decisions fail through overconfidence. A manager sees a demo, assumes the automation will work the same way in their environment, and then discovers that their real process includes exceptions, approvals or local tax rules that change the result. The mistake is not wanting automation. The mistake is confusing a demonstrated capability with operational readiness for your case. (odoo.com)
A simple example: a distribution company wants confirmed orders to create tasks automatically. Fine. But if some orders are partial, urgent, or changed by phone at the last minute, the design must decide which events trigger automation and which stay under human review. If that boundary is not defined, the tool accelerates mistakes as well as work. That is the line leaders need to watch. (odoo.com)
The uncomfortable mistake: automating before cleaning the process
The temptation with a new release is to use it to hide an already messy operation. You see this in companies that have accumulated informal rules: “Ana always checks that one”, “if the customer is big, skip the flow”, “we fix it later in accounting”. Odoo 20 can help formalize work, but it does not fix a process nobody has wanted to describe. (odoo.com)
In a scenario similar to many SMEs, a business with disconnected sales, purchasing and warehouse teams turns on automations out of fear of missing out. The first result is frustration: more alerts, more exceptions and more dependence on the internal team to stop fires. The lesson is not “don’t automate”. The lesson is that automation should follow operational judgment, not replace it. First define what the system should decide; then let the new release execute it. (odoo.com)
The decision worth making this week
If you are evaluating Odoo 20, do not start with the feature list. Start with three questions: which process repeats too often, where do you lose control when work moves from one team to another, and what exception would force you to roll back. That is how you separate real news from decorative novelty. (odoo.com)
Three sensible next steps for this week:
- Pick one repetitive process and write down its real exceptions, not the theoretical ones.
- Check which part of that flow could be automated without changing permissions, taxes or approvals.
- Ask the team which manual task would disappear first if the new version worked as intended.
If you want to discuss this on LinkedIn, the better question is not “do you like Odoo 20?”, but: which process in your business should be automated first without losing control? That is the kind of conversation we like to have at Codefuente when a platform changes.