The short answer: the right ERP is the one you will actually finish setting up. Feature lists and vendor comparisons matter, but they are rarely where small businesses lose money. The expensive mistakes happen after the choice — defaults nobody checked, modules nobody switched on, and forms that never reached the system.
When I first wrote this guide in 2024, it was a list of selection criteria. Those still hold, and they're below. But since then I've run Odoo as the operating system for my own businesses — one database holding the CRM, the projects, the knowledge base, and the websites for my coworking space, my consultancy and my own name. That taught me more about choosing an ERP than any comparison chart did, so I've added what I learned.
The seven questions to ask before you choose
| Criterion | The question to ask | What it costs you if you skip it |
|---|---|---|
| Requirements | Which three problems must this solve in the first ninety days? | You buy for features you will never configure |
| Scalability | What happens when we add users, a location or a new line of business? | A migration in two years instead of a setting change |
| Integration | Does it connect to the tools we already depend on — or replace them? | Data re-typed by hand between systems, forever |
| Vendor and partner support | Who helps when something breaks, and how fast? | Your operations wait on a support queue |
| Total cost of ownership | Licenses, implementation, customization, training and maintenance — over three years, not one? | The cheap option becomes the expensive one |
| Usability | Will the people doing the work actually use it, or work around it? | The system holds data nobody trusts |
| References | Who in a business like ours is running it today, and what would they do differently? | You learn their lessons at full price |
Ask for a demo or a trial before deciding, and run a real task through it — not the vendor's script.
What I learned running Odoo across my own businesses
These are the three lessons I wish someone had given me before I started. None of them shows up in a feature list.
Blank does not mean "none"
In an ERP, an empty field often means everything, not nothing. When I audited my own setup, I found four published course channels showing up on websites they didn't belong to — because the field that assigns them to a website had been left blank, and blank meant "all of them." I also found two marketing automation campaigns with an empty audience, which in Odoo means every contact in the database. Neither had been switched on. If either had been, it would have emailed everyone.
Before you trust a default, find out what blank means.
Installed is not running
When I audited my marketing automation, there were eight campaigns built and not one had ever been switched on. I had paid for a capability and was using none of it.
This is the hidden half of total cost of ownership. Count your ERP by what is live, not by what is installed. An unused module is a cost with no return, and it is easy to miss because nothing breaks.
The pipes matter more than the pick
For four years my website forms emailed an inbox instead of landing in the CRM. Every lead had to be copied over by hand — which means some weren't. I spent that period trying to fix my strategy, rewriting positioning and pricing, when the actual problem was that nothing could reach a customer through pipes that weren't connected.
Strategy is a luxury for businesses with working pipes. Whatever ERP you choose, connect the intake first: forms, email and payments landing in the system without a human retyping them.
If you're a small operation
Most ERP guides are written for companies with an IT department. If you're a founder, a team of a handful of people, or running more than one small business at once, the order matters more than the platform:
- Write down how the business actually runs — roles, processes, what gets tracked — before you configure anything. Software encodes a process; it can't invent one.
- Connect intake before you build reports. A dashboard over incomplete data is worse than no dashboard.
- Switch on one module at a time and confirm it is live before adding the next.
That's the approach I packaged as FoundERP — an operating system for solo founders, built from running my own businesses this way. You can read how I built it, or start with the template kit if you'd rather have the documents than the story.
The bottom line
Pick on the seven questions above. Then spend your real effort where small businesses actually lose money: the defaults, the switches, and the pipes. A modest system that is fully connected will outperform a powerful one that is half set up.