How to Choose an ERP: A Practical Selection Framework

To choose an ERP, start with your processes and requirements, not with vendor demos. Define how your business needs to work, shortlist a few platforms that fit your industry, size and budget model, then score each one against weighted criteria using scripted demos built around your own scenarios. The right ERP is the one that fits your operations and your ability to run it, not the one with the longest feature list.

I help businesses through ERP selection as an independent consultant, which means I have no commission riding on which product wins. Below is the process I use, including the scorecard I give clients.

Why ERP selection goes wrong

Most poor ERP choices come from a handful of predictable mistakes:

• Demo-driven decisions. Vendors show their strongest features with clean sample data. If you let them set the agenda, you evaluate their pitch, not your fit.
• No agreed requirements. Without a written, prioritized list, each stakeholder judges on different criteria and the loudest voice wins.
• Licensing cost over total cost. Subscription price is visible; implementation, integration, customization and internal time are not. The cheapest license can be the most expensive project.
• Ignoring the partner. The quality of the implementation team often matters as much as the software.
• Buying for a future that never arrives. Paying for complexity you will not use for years makes adoption harder today.

Step 1: Understand the business before the software

Before you speak to any vendor, map your core end-to-end processes: order-to-cash, procure-to-pay, inventory and fulfillment, production or project delivery, and record-to-report. Note who does each step, which tools they use, and where things break.

This is not paperwork for its own sake. It tells you which processes are standard (and should follow the system's best practice) and which are genuine differentiators that the ERP must support well. If you need help here, this is the core of my business process consulting work.

Step 2: Write prioritized requirements

Turn the process map into requirements, and rank each one:

• Must-have: the business cannot operate without it (for example, multi-currency invoicing, batch tracking, local tax compliance).
• Should-have: important, but a workaround is acceptable for a period.
• Nice-to-have: useful, but should not drive the decision.

Keep requirements specific and testable. "Good inventory management" cannot be scored. "Track stock by batch and expiry date across multiple warehouses, with FEFO picking" can. My ERP requirements checklist gives a module-by-module starting list, and for a formal document see how to create an ERP BRD.

Step 3: Build a realistic shortlist

Narrow the market to three or four candidates using a few hard filters:

• Industry fit. Does the platform handle your core model (manufacturing, project-based, distribution, services) natively, or only through heavy customization?
• Company size and complexity. Number of entities, currencies, countries, users and transaction volume.
• Deployment model. Cloud subscription, self-hosted, or open source, and any data-residency constraints.
• Local compliance. Tax rules, e-invoicing and statutory reporting in each country you operate in.
• Ecosystem. Availability of experienced implementers and support in your time zone.

For many SMB and mid-market businesses, the shortlist includes platforms like Zoho, Odoo and ERPNext, alongside larger suites such as NetSuite, Microsoft Dynamics or SAP for more complex groups. Each has a different philosophy on configuration, hosting and extensibility, so the "best" one depends entirely on your requirements.

Step 4: Run scripted demos

Give each vendor or partner the same set of real scenarios and ask them to demonstrate those, ideally with a sample of your data. Good scenarios follow a full process, such as:

1. A customer orders a mix of stocked and made-to-order items in a foreign currency, with a partial delivery and a credit note.
2. A purchase request goes through two approval levels, is converted to a PO, partially received, and matched to the supplier bill.
3. Month-end: close the period, run a consolidated P&L across two entities, and drill into a variance.

Score during the demo, not afterwards from memory. Note anything the presenter says "can be customized," because that is a cost and a maintenance commitment, not a feature.

Step 5: Score with a weighted scorecard

A scorecard turns opinions into a comparable result. Agree the weights with leadership before the demos start, so nobody adjusts them to favor a preferred product. Score each criterion from 1 (poor) to 5 (excellent), multiply by the weight, and add up the results.

The weights above are only an example. A manufacturer might weight industry fit higher; a group with entities in several countries might weight compliance higher. What matters is agreeing them up front.

Two rules I apply to every scorecard:

• Must-haves are a gate, not just a score. If a platform cannot meet a true must-have without risky customization, it is out, whatever its total.
• Close totals mean look at risk. If two products score within a few points, the decision should come down to implementation risk, partner quality and total cost.

Who should be involved in the decision

ERP selection should not be run by IT alone or by finance alone. A small selection team works best:

• Executive sponsor: usually the CEO, COO or CFO, who owns the decision and breaks ties.
• Process owners: one person each from finance, sales operations, purchasing, warehouse or production, and projects, depending on scope. They write and score the requirements for their area.
• Power users: people who do the daily work. They spot usability problems that managers miss.
• IT or systems lead: to assess integration, security, hosting and data questions.
• Independent advisor (optional): someone without a stake in which product wins, to run the process and challenge vendor claims.

Keep the team small enough to make decisions, and make sure each member actually attends the demos they are scoring.

Step 6: Understand total cost of ownership

Compare total cost over several years, not first-year subscription fees. Include licenses or subscriptions, implementation services, data migration, integrations, customization, training, hosting (if self-managed), ongoing support, and the time your own staff will spend on the project. I break these down in ERP implementation cost.

Step 7: Check the implementation partner

Ask who will actually work on your project, what similar businesses they have implemented, how they handle scope changes, and how they test before go-live. Ask to speak to their past clients directly. A strong partner will challenge your requirements; a weak one will agree to everything.

Step 8: Decide, then validate

Before signing, run a short proof of concept or a paid discovery phase on your highest-risk requirement. Confirm the contract terms: user types, data ownership, exit options, and what support covers. Then make the decision formally, record why, and move into planning with the ERP implementation guide.

Next steps

If you want an independent view on your shortlist, or help building requirements and scripted demos, I run ERP selection as part of my ERP consulting services. Get in touch to discuss where you are in the process.

Frequently Asked Questions

How many ERP vendors should be on a shortlist?

Three or four is usually enough. Fewer leaves you without a real comparison; more spreads your team's time thin and makes scripted demos hard to run properly. Use hard filters such as industry fit, deployment model and local compliance to narrow the field first.

What is a scripted ERP demo?

A scripted demo is one where you give every vendor the same real business scenarios, ideally with your own sample data, and ask them to walk through them end to end. It lets you compare products on your processes rather than on each vendor's preferred highlights.

Should we choose the ERP or the implementation partner first?

Evaluate them together. A good product with a weak partner can fail, and a strong partner can make a reasonable product work well. Ask who will actually staff your project, what similar work they have done, and speak to their past clients.

Is the cheapest ERP subscription the best value?

Not necessarily. Subscription fees are only one part of the cost. Implementation, customization, integrations, data migration, training, support and internal staff time often outweigh license costs, so compare total cost of ownership over several years.