How do I choose an ERP system?

Work requirements-first, not vendor-first. Document your core workflows and turn them into a written must-have and nice-to-have requirements list, weight the criteria that matter (industry fit, functional coverage, size fit, deployment, total cost of ownership, and lock-in), model five-year TCO, narrow to a shortlist of three to five systems, then evaluate demos against your own data before deciding.

How long does ERP selection take?

For a small or mid-sized business, a disciplined selection runs four to twelve weeks: one to two to document requirements, one to build a shortlist and model cost, and two to four for scenario-based demos and reference checks. Rushing selection is the most expensive way to save time, because the cost of the wrong system dwarfs the weeks saved.

What is an ERP requirements checklist?

It is a structured list of the capabilities your business needs from an ERP, grouped by function (finance, inventory, manufacturing, CRM, projects, reporting, integrations) and marked must-have or nice-to-have. It is the yardstick you grade every vendor against, and it prevents you from buying features you were shown rather than the ones you need.

Should I hire a consultant to choose an ERP?

A good independent advisor earns their fee by keeping the process requirements-first and calling out lock-in and hidden implementation cost. Be wary of advisors paid by the vendors they recommend. Even a single honest scoping call, with someone who will tell you when their own product is the wrong fit, can save a six-figure mistake.

GuideERP SelectionAugust 9, 2026By Octura Solutions

How to Choose an ERP System
A practical selection process, from requirements to signed decision

Q: What are the main ERP selection criteria?

Six criteria carry most ERP decisions: industry and process fit, functional coverage of your must-have list, company size fit, deployment model (cloud, on-premise, or open-source self-hosting), total cost of ownership over five years, and customization and lock-in. Weight them before you see prices or demos so an irrelevant strength cannot distort the result.

Most ERP projects that fail were lost at selection, not implementation: the wrong system was chosen for the wrong reasons, usually a slick demo and a discount that expired on Friday. This guide is the process we walk clients through, requirements first, vendors last. It is deliberately vendor-neutral, even though we implement Odoo for a living, because a system chosen to fit the sales cycle rather than the business is the most expensive mistake on this page.

Start with requirements, not vendors

The single most common selection mistake is shopping before you have written down what you need. You end up grading vendors on the features they chose to demo, not the processes that actually run your business. Reverse it. Before you watch a single demo, document your core workflows, the ones that make or lose you money, and turn them into a written requirements list grouped by function: finance and compliance, inventory and fulfilment, manufacturing, sales and CRM, projects and services, reporting, and the integrations you cannot live without. Mark each as a must-have or a nice-to-have. That list, not a vendor's brochure, is the yardstick every later step measures against. We keep a free, structured ERP requirements checklist you can run through in ten minutes to produce exactly this document.

The selection criteria that actually matter

Once requirements are written, weight the criteria you will judge systems on, because they are not equal. Six carry most of the decision: industry and process fit (does it serve your vertical without heavy customization), functional coverage (how much of your must-have list it covers out of the box), company size fit (systems built for 20 people and systems built for 20,000 are different animals), deployment model (cloud, on-premise, or open-source self-hosting), total cost of ownership over five years, and customization and lock-in (how hard is it to change, and can you leave). Decide the weighting before you see prices or demos, so a strong performance on a criterion you barely care about cannot distort the result.

Compare total cost of ownership, not the sticker price

The per-user monthly figure on the pricing page is the smallest part of what an ERP costs you. The real number over five years is licenses plus implementation, data migration, integrations, training, hosting, support, and the upgrades you will owe down the line. Two systems with identical license fees can differ by six figures once implementation complexity and renewal increases are counted, and the cheapest sticker is often the most expensive system by year three. Model it properly before you shortlist: our total cost of ownership calculator gives you a five-year figure you can put next to each vendor's quote.

Build an honest shortlist of three to five systems

With weighted criteria and a real cost model, narrow the field to three to five candidates before you book demos. More than five and evaluation collapses under its own weight; fewer than three and you have not really compared. Match your profile against how each system is actually positioned rather than how it markets itself: a mid-market manufacturer, a micro service firm, and a multi-entity distributor should end up with different shortlists. To do this in a couple of minutes, our ERP Selection Tool scores Odoo and nine major alternatives against your answers and returns a ranked shortlist, and it tells you plainly when Odoo is not your best fit. For head-to-head detail, we keep honest breakdowns of Odoo against 25+ systems.

Evaluate demos against your data, not theirs

A scripted demo is designed to succeed. Neutralize it by controlling the script. Send each shortlisted vendor the same three or four scenarios drawn from your requirements list, your quote-to-cash flow, your month-end close, your worst product with its real bill of materials, and ask them to demo those with sample data you provide. You are no longer watching a sales reel; you are watching your business run on their software. Score each demo against the weighted criteria from step two, involve the people who will use the system daily, and write down where each candidate needed customization to do something your list marked must-have. That is where the real implementation cost hides.

Decide, then de-risk the implementation

The winner should fall out of the scores, not the last conversation with a sales rep. Before you sign, check references in your own industry and size, confirm what the quote excludes, and pin down who implements it, because with most systems the software and the implementer are separate decisions, and a great system with the wrong partner still fails. Agree a phased rollout rather than a big-bang cutover where you can, and make sure you own your data and can export it. Choosing the system well is most of the battle; choosing how you roll it out is the rest.