This week I attended an Odoo Business Show. During the partner presentation, by my own estimate around 60% of the questions were about AI.
That makes sense. If AI lets you build software faster and faster, what does that mean for a package like Odoo? Why would you still pay for standard software when you can build exactly what you need?
Whether you should start building is a question we answered at length in building your own software with AI. This piece is about the layer underneath: the assumption the whole argument rests on.
Because my view is that Odoo in particular holds up well here. And that has less to do with what you can build than with who else has AI.
The assumption almost nobody says out loud
The case for building your own often sounds as though only the individual entrepreneur has access to AI.
That is where the flaw sits. The same technology is available to Odoo, to partners and to the community.
That difference is not small. If you improve something with AI, your company benefits. If Odoo improves something with AI, it lands with a great many customers at once, and all those companies carry the development cost together.
I expect that scale to weigh heavily. An individual company can achieve a lot with AI. A platform can deploy that same technology to keep developing a product that thousands of companies pay for.
On top of that, the public Community code and the available documentation offer a concrete basis for giving AI context during development and implementation. You start with an existing platform that already holds years of work, not with a blank screen.
The option to build your own is not new
What stands out in this discussion is that Odoo customers have been making this trade-off for years.
Odoo has an open-source Community edition, with the licensed Enterprise edition on top of it. So there has long been a choice between freely available software and a paid product with additional functionality and services. Odoo explains the differences here, and we worked through the trade-off in Odoo Community or Enterprise.
Yet our customers choose Enterprise. And they do not do that because they forgot Community exists.
They choose it because for business-critical processes, such as accounting, they want to build on a professionally maintained product. Because there is an organisation behind it that keeps developing. Because they value support, training material and a professional partner network. And because hosting and backups can be properly arranged under a suitable hosting choice.
The ability to continue with a different partner later also has value. That is easier when your environment stays recognisable to someone who did not build it themselves.
In other words: the choice between arranging it yourself and buying a maintained product existed here long before AI made building cheaper. And for business-critical processes, that choice consistently fell the same way.
Meanwhile, a lot is going to change for partners
If I see pressure building anywhere, it is on my own profession.
I already see it at our own company. AI helps us analyse customer situations, run fit-gap analyses and assess solution directions. That lets us get to the core faster: what does this business need, what fits within the standard, and where is a deliberate deviation justified?
I expect this to become visible throughout the whole implementation in the coming years. From configuring and preparing migrations to documenting and building extensions. How that shift played out for us is in how AI changed our company.
If a large part of your added value consists of knowing where to click, you need to think about that seriously.
That pressure existed already, by the way. Technical work could be outsourced before AI as well. AI only makes that competition more broadly accessible.
For me, the value of a good partner increasingly lies in understanding the business. In guiding choices. In recognising when a deviation is necessary, and when an organisation has mostly grown used to a convoluted way of working.
And sometimes in saying: we can build this, but I would not do it.
Precisely when building gets easier, that judgement matters more.
Business software becomes a framework
I also expect business software to become more and more a framework within which AI helps build business logic. First under supervision, later increasingly in maintaining it and proposing improvements too.
In that case you want a solid base: clear data, well-configured access rights, coherent processes and an environment someone else can understand.
That point is often read the other way round. People expect AI to make the setup matter less, because you can have anything built anyway. I think the opposite: the more you let AI work inside your processes, the more it matters whether those processes cohere and whether your data is right.
So what would I do if my business software were outdated?
I would choose an open and broadly supported platform. With a good partner or a strong internal implementation team that understands how far you should go. Including customisation, including when AI can write that customisation in an afternoon. That customisation is now within reach for far more companies is in fact true: we worked that out in custom software has become affordable for the mid-market.
But Odoo drinks Red Bull too
That reminds me of that Red Bull commercial. A tortoise applies for a delivery job where speed is the whole point. He produces a can of Red Bull to show he can keep up just fine.
Only, the hare drinks Red Bull too.
For me, that is the heart of the AI story. The question is rarely whether a new technology makes you faster. The question is whether it makes you faster relative to everyone who also has it.
Of course you can build software faster with AI.
But Odoo drinks Red Bull too.
Frequently asked questions
Will AI make standard software obsolete? I do not expect so. The case for building your own silently assumes only you have AI, while the vendor, the partners and the community hold the same technology.
Why would a vendor benefit from AI faster than I would? Not faster, but wider. An improvement at Odoo lands with a great many customers at once, and they carry the development cost together.
What does AI mean for the role of an Odoo partner? Work that consists of knowing where to click is under pressure. What remains is understanding the business and guiding choices.
Why choose Enterprise when Community is free? Because for business-critical processes you want to build on a professionally maintained product, with support, training material, a partner network and properly arranged hosting and backups.
Does AI change how you should set up your software? Yes, and the other way round from what many expect. The more AI works inside your processes, the more it matters whether those processes cohere and whether your data is right.
Not sure whether a deviation in your process is genuinely necessary? Book a free Odoo scan. We look with you, and we also say when you are better off not building something.
Read more: Build your own software with AI? Why the smartest entrepreneurs still buy · In three months, AI changed our company · Odoo Community or Enterprise