Most ERP regret is a selection problem, not a software problem. The platform was chosen from a demo, the demo showed the parts that work beautifully, and the parts that would matter in eighteen months were never tested. What follows is the way we run a selection, and the honest trade-offs between the five platforms we implement.
Start from the processes that are hard, not the ones that are common
Every ERP handles invoices, purchase orders and stock movements. That is not where the decision is made. The decision is made on the two or three processes that are specific to how you earn money: progress billing on a construction contract, recipe costing across branches, tank reconciliation at a fuel station, batch traceability on a production line. Write those down first, and make every vendor demonstrate them with your data rather than their sample company.
Odoo
The broadest application range of the five, and the easiest to grow into: you can start with accounting and inventory and add manufacturing, field service, a website and a point of sale later without changing platform. Customisation is quick, which is a strength and a trap, because heavy customisation is what makes version upgrades expensive. It suits companies that want one system covering many departments and have someone able to own it internally.
ERPNext
Open source with no per-user licence, which changes the arithmetic for companies with many light users: warehouse staff, site supervisors, shop floor. Manufacturing and inventory are genuinely strong. The trade-offs are a smaller partner ecosystem, fewer ready-made localisations, and reporting that usually needs building rather than picking from a list. It rewards companies willing to own their system technically.
Zoho
The fastest of the five to get running, and the strongest on the commercial side: CRM, quotations, customer portals and support desks. It works best for sales-led and service businesses. The caution is that Zoho is a suite of separate applications rather than a single core, so the seams between them become your integration work, and deep manufacturing or project costing is not its ground.
Microsoft Dynamics 365
The right answer when the company already lives inside Microsoft: Teams, SharePoint, Excel as the reporting habit, Azure for infrastructure, and an IT department that thinks in those terms. Business Central covers mid-sized operations and Finance & Operations covers genuinely large ones. The cost is higher and the implementation is more dependent on the partner, so the risk is buying an enterprise tier for a mid-sized problem.
Oracle
Chosen for scale, regulatory depth and multi-entity consolidation rather than for convenience. It earns its place in large groups with several legal entities, currencies and audit obligations. For most companies asking us which ERP to pick, it is more system than the problem requires, and we say so.
The question most selections forget
Who will own this system after we leave? An ERP is not a purchase, it is a responsibility that continues: new staff, new processes, new reports, a version upgrade every couple of years. The platform that fits is the one your team can actually run. That single question changes the shortlist more often than any feature comparison.
We implement all five, and we are paid to implement, not to resell licences, which is why this note reads the way it does.