How to Choose a Manufacturing ERP (Without Getting Burned)

Why ERP selection goes wrong

Most shops choose an ERP based on a demo. The salesperson shows you the best-case scenario with clean data, perfect workflows, and features you didn't ask for. You sign. Six months later, you're halfway through implementation, the system doesn't match how your shop works, and the vendor says customization is extra.

The problem isn't that ERPs are bad. It's that the selection process is backwards. Shops start with features when they should start with fit. The best ERP for your shop isn't the one with the most capabilities — it's the one your team will actually use.

Start with your pain, not their features

Before you look at any software, write down the three things that cost you the most time or money right now. For most small shops, it's some combination of:

- Quoting takes too long and margins are a guess - Job status is tribal knowledge — you walk the floor to find out what's running - Scheduling is a whiteboard that's always wrong - Invoicing is disconnected from production - Quality records live in binders and nobody can find anything

These are your evaluation criteria. Any ERP you consider should solve your top three problems in the first month — not after six months of configuration.

Red flags in the sales process

Walk away if you see any of these:

No published pricing. If the vendor won't tell you what the software costs without a "discovery call," the price will be higher than you expect and negotiated based on your budget rather than the product's value.

Required implementation partner. If the software needs a third-party consultant to configure, it's too complex for a small shop. You'll spend $10K-100K before you track a single job.

Per-user pricing above $100/month. At $150/user, a 10-person shop pays $1,500/month — $18,000/year — and every new hire increases the cost. Per-company pricing scales better.

No data export. If you can't get your data out in a standard format (CSV at minimum), you're locked in. Ask this question early.

The demo uses their data, not yours. A good vendor will let you trial the system with your actual customers, parts, and routings. If they won't, the demo is hiding how painful data entry really is.

The evaluation checklist

Score each ERP you consider on these ten criteria:

1. Can you quote a real job within your first hour? 2. Can an operator clock into a job without training? 3. Does scheduling show you machine conflicts at a glance? 4. Does job costing compare quoted vs actual automatically? 5. Can you generate an invoice from a completed job in under 2 minutes? 6. Is pricing published and predictable? 7. Can you set up the system yourself without a consultant? 8. Can you export all your data in CSV format? 9. Does the vendor have manufacturing experience (not just software experience)? 10. Can you talk to a real reference customer in your industry?

Any system that scores 7+ out of 10 is worth a deeper evaluation. Below 5, move on.

The 30-day test

Never commit to an annual contract without a real trial. The right approach:

Week 1: Import your customer list and part master. Quote three real jobs. If quoting doesn't work well, nothing else matters — stop here.

Week 2: Track five real jobs through production. Have operators log time. See if the data is useful or just noise.

Week 3: Generate invoices from completed jobs. Run a margin report. Does the number match reality?

Week 4: Make your decision based on real data, not demo impressions.

If the vendor won't give you 30 days to try the system with real data, that tells you something about their confidence in the product.

The total cost calculation

The sticker price of an ERP is never the real cost. Calculate the total cost of ownership:

Software: Monthly subscription × 12 months × years you expect to use it. Implementation: Consultant fees, data migration, training time. For enterprise ERPs, this is often 2-5× the first-year software cost. Productivity loss: How many hours will your team spend learning the system instead of making parts? At $75/hour shop rate, two weeks of disrupted production across 10 people is $60,000. Ongoing costs: Per-user fees as you hire, upgrade costs, support costs, add-on modules.

A system that costs $300/month but your team uses in a week has a lower total cost than a system that costs $200/month but takes three months to implement.