When not to
I make a living implementing Odoo, so this page works against me, and that is why it exists. A client who bought what they did not need ends up blaming the system, the vendor, and the whole idea of getting organised. I would rather say it up front. These are the six cases where I would turn the work down.
Three, four, five people, no stock and no production, and the pain is issuing invoices and keeping records for the accountant. A full ERP there is buying a lorry to fetch the bread.
What to do instead: an electronic invoicing tool for your country, which costs about what lunch costs per month, and bookkeeping through your accountant. When stock starts to hurt, come back.
Clinics, hotels, workshops, pharmacies, laboratories, construction firms. There are vertical systems that already solve ninety per cent of what is particular to the trade. Odoo can do it, but that ninety per cent would have to be built.
What to do instead: look at the vertical first. If it handles the operation and only falls short in accounting or purchasing, then the conversation is about integrating, not replacing.
You need to track sales opportunities, or control stock in one warehouse, or run payroll. One problem, well defined.
What to do instead: a tool built for that. An ERP earns its keep when the value is in the pieces talking to each other; with one piece there is nothing to connect.
An ERP is not an appliance. Someone in the company has to create products, review settings and decide when something changes. If nobody will have those hours, the system degrades within months.
What to do instead: decide who that person is and free up their time in writing. If that cannot happen, the project is not ready, whatever system you buy.
A sale, a merger, a change of business model, a partner leaving. Implementing during that means configuring for a company that is about to stop existing.
What to do instead: wait until the change is settled. In the meantime, sorting out the processes costs little and holds its value afterwards, with any system.
If the money covers the software but not the work of implementing it, migrating the data and training people, what you are buying is an expensive file. The licence is the small part.
What to do instead: wait a quarter and start with the cheap part, which is sorting out the processes. A half-finished project costs more than not starting.
Odoo fits when the value is in connecting: when the order has to move stock, the invoice has to come out of the delivery, and accounting has to find out by itself. When several teams write the same thing in different places today. When custom code or free modules can cover what is particular to the business without reinventing it.
There I am good. In the six cases above I would rather say so and lose the project.